Reed's Ruminations: A Blog by Dan Reed

Upgrading the Future: House Science Committee Testimony

N.B. I am extremely grateful to the National Science Board Office (NSBO), and its dedicated and talented cadre of professional staff, who helped prepare my testimony.  These include, but are not limited to, Elizabeth Jeffers, Elise Lipkowitz, Nadine Lymn, Andrea Rambow, Amanda Vernon, and John Veysey.

In my role as chair of the U.S. National Science Board, which oversees the National Science Foundation (NSF), I recently testified to the U.S. House Committee on Science, Space, and Technology, along with the NSF Director, Sethuraman Panchanathan.  The hearing was entitled “An Overview of the National Science Foundation Budget Proposal for Fiscal Year 2024.”

These hearings typically follow a standard script, with initial statements from the committee chair and the ranking member.  You can read Chair Lucas’ statement here, including his important observation:

I strongly believe that the nation that leads in science and technology will shape the world order for the next century. I’d like that nation to be ours, and I’d like for emerging technologies to be developed with our values of transparency and fairness.

Opening statements from the committee chair and ranking member are then followed by oral statements from the witnesses, typically five minutes each, with a subsequent question and answer period.  There, each member of the committee is allotted five minutes to ask questions.  The fact that the hearing lasted two and a half hours, with thoughtful questions, speaks to the level of interest from both political parties. You can find the hearing details here, including the written statements from NSF Director Panchanathan and me (Download Reed Testimony), as well as a video of the oral statements and the subsequent question and answer period. 

I have spent decades talking to U.S. Senators, Representatives, Congressional staffers, and policy groups. (See Working the Hill.) Despite the rancor that often exists in the U.S. Congress, the House Science Committee continues to operate in a bipartisan way, strongly supporting investment in basic research.  I am grateful to both Chair Lucas and Ranking Member Lofgren for their strong support for the CHIPS and Science Act, in both the previous Congress and in this one.  This hearing was a continuation of that support.

My Written Testimony

My extended written testimony, entitled Upgrading the Future: Realizing Better Dreams, outlines the issues in greater detail, but the gist of my argument is simple. The United States is losing ground in science and engineering.  The country needs to increase investment in basic research – both people and leading-edge research infrastructure – expand the numbers and diversity of the STEM workforce, while continuing to be a magnet for global talent, and more aggressively translate research ideas into practice. The latter is the focus of the NSF’s new Technology, Innovation, and Partnerships (TIP) Directorate.

I also called for a more coherent, strategic plan for U.S. science and engineering, one that engages “all of government” in a collaborative, coherent way. To be clear, I was not calling for a top-down government plan, with all its attendant bureaucratic burdens, but rather one that lays out the key components our system needs and rewards systems thinking. I hope the upcoming Quadrennial Review, called for in the CHIPS and Science Act, moves the nation in this direction. I know the National Science Board is eager to partner with the Office of Science and Technology Policy (OSTP) on this project.

In my written testimony, I also drew an analogy between operating systems and our science and engineering ecosystem as an operating system. As we have become more complacent in our approach, we have resorted to patching “holes” in the code. As any computing expert knows, it is expedient in the short term, but in the long term, it is a risky strategy. Fixing one piece of code can break another, and a multitude of patches creates a brittle system, making it difficult to add desirable new features or respond to new threats.

Simply put, we are running a patched, 20^th century innovation ecosystem in a 21^st century world. We need to upgrade our entire operating system: expanding and diversifying the STEM talent pipeline, accelerating the delivery of research benefits, upgrading our research infrastructure, and elevating our commitment to basic research.

With that backdrop, what follows is my opening statement, as delivered, with a few additional annotations for context. Again, you can watch the video here.

My Opening Statement

Chairman Lucas, Ranking Member Lofgren, and Members of the Committee, thank you for the opportunity to speak with you as Chair of the National Science Board. Thank you as well for the CHIPS & Science Act and its bold blueprint for a brighter future.

One of my favorite science fiction authors, Neal Stephenson, once wrote, “if we want to create a better future, we need to start with better dreams.”  The CHIPS & Science Act is the stuff of better dreams, and it arrives at a critical juncture for our nation.  (See For Science and Society, The Future Begins with Better Dreams.)

To ensure future breakthroughs and innovations are “Made in America,” we must continue translating the Act’s vison into action – advancing scientific frontiers, developing STEM talent, expanding the geography of innovation, and delivering benefits to society.

Fully funding the Administration’s FY24 budget request will help make the vision a reality. But, we must to do more.

Let me be clear: U.S. leadership in science and technology is in peril. China is charging ahead, and absent further action, it is not a question of if but when the United States loses its STEM leadership, with deep consequences for our country.

Here’s why we must act – now.

FIRST. China’s announcement that it is ramping up government investment in basic research as “the only way to build a world scientific and technological power” both validates our strategy and highlights our challenge.  

It’s time for us to “double down” – expand investment in basic research and cutting edge scientific instruments – nationwide – and unleash American innovation.

SECOND. China continues to invest heavily in building its homegrown talent and now produces more STEM PhDs than the United States.

Meanwhile, we face a STEM talent crisis.  Students at all levels and all backgrounds are struggling in STEM – and COVID made it much worse. This crippling situation is even more acute for students from lower socioeconomic standing or underrepresented backgrounds. We are simply not producing enough STEM workers – at all levels – to meet the needs of a 21st century economy, and we are leaving millions of talented individuals behind.  (See Missing Talent, Missing Opportunities, and Losing Ground.)

For our STEM workforce to be representative of the U.S. population in 2030, the number of women must nearly double, Hispanic or Latinos must triple, Black or African Americans must more than double, and the number of American Indian or Alaska Natives must quadruple.

Forty-eight years ago, I was a poor, first-generation college student from the Arkansas Ozarks, studying computer science. I was fortunate to graduate debt free, thanks to academic scholarships, a Basic Educational Opportunity Grant (now Pell), and my paltry savings.  While I am extraordinarily grateful, I am alarmed that my educational path is no longer widely available. (See Transforming Lives via Public Higher Education.)

To grow our STEM talent base, we must do more to make higher education affordable and STEM graduate work more financially viable.  Hence, NSF’s FY24 budget request emphasizes broadening participation in STEM education and turning STEM career dreams into realities. (See Public Higher Education: A New Social Compact for Innovation.)

THIRD AND FINALLY. We must deliver research benefits to society.

NSF’s new Directorate for Technology, Innovation and Partnerships – TIP – will accelerate innovation by supporting use-inspired research and translation, building institutional and regional innovation capacity, and enhancing academic, government, and industry partnerships.

At the end of World War II, the compelling rationale for federal government research investment was to advance the national health, prosperity, and welfare; and to secure the national defense.

Almost eighty years later, it still is.

I dream historians – and our children and grandchildren – will mark now as the time we not only embraced better dreams, but put aside our differences, committed to our common goals, and acted – with compelling vision and unwavering resolution – to create a better future – for our country and for the world.

Thank you. I look forward to your questions.

Concluding Perspectives

While we in the United States continue to debate and dither, the time for concerted action grows ever shorter and the stakes grow ever higher. Having watched and learned, other countries have been investing heavily in their own innovation ecosystems, cultivating human talent, expanding their knowledge workforce, and constructing the advanced infrastructure needed to facilitate discovery, economic growth, and defense capabilities.  

It’s time for the United States to relearn and heed Vannevar Bush’s important lesson, delivered in his seminal essay, Science: The Endless Frontier, at the end of World War II:

… since health, well-being, and security are proper concerns of Government, scientific progress is, and must be, of vital interest to Government. Without scientific progress the national health would deteriorate; without scientific progress we could not hope for improvement in our standard of living or for an increased number of jobs for our citizens; and without scientific progress we could not have maintained our liberties against tyranny.

Science – it’s about a better future – for everyone.

On WSPRs, Geeks, and Luftballons

A few weeks ago, a large, high altitude balloon carrying a large payload of Chinese surveillance equipment was shot down by a U.S. Air Force F-22 Raptor using an AIM-9X Sidewinder missile.  The F-22 was scrambled from Langley Air Force Base in Virginia, the downed balloon landed offshore in South Carolina, and an intensive intelligence salvage operation began immediately.  This was a large, likely multiple ton, balloon with a substantial array of solar cells, navigational and directional capability, and electronic packages. Because my wife and I own a beach house in North Carolina, near the South Carolina border, I received local warnings to report and deliver any floating debris to the authorities.

In response, Secretary of State, Antony Blinken, canceled his planned trip to Beijing, and Sino-American geopolitical relations, already strained, sank even further, with public accusations and recriminations flying fast and furiously. The resulting consternation ignited political posturing by pundits across the political spectrum, which, in turn, led to strident calls for a public statement from the U.S. President.

How could this happen?  Why wasn’t the balloon detected earlier?  Why was it allowed to drift across the continental United States before being shot down? Members of the U.S. Congress wanted answers, and they got them via classified national security briefings. Meanwhile, NORAD increased its aerial and near-space surveillance, given heightened national awareness and concern about other objects in the sky. 

International law is clear that satellite overflights of a country (i.e., from space) are allowed.  Conversely, countries have the complete right to control the airspace over their territory, though international law is a bit uncertain about the rules at very high altitudes. As a practical matter, might makes right; if you can shoot it down over your territory, you have the right to do so. (Remember the history of the U-2 and SR-71 spy planes and note that the aerial photograph of the balloon above was taken from a U-2.) Relatedly, radio transmissions are controlled by International Telecommunications Union (ITU) treaty, as part of the United Nations.

Putting aside the very real and important national security questions behind the Chinese balloon and the calculations and mis-calculations that led to its path over the United States, the public hysteria and political ennui brought to mind the West German band Nena’s song, Neunundneunzig Luftballons (99 Air Balloons), which became an international hit in 1983 as an anti-war protest. It’s a poignant and melancholy Cold War story of a group of air balloons accidently triggering a 99-year global war, leaving only devastation in its wake.  (You can find it on YouTube, and the German lyrics are just as haunting today as they were forty years ago.)

            Original German                                          English Translation
            Neunundneunzig Jahre Krieg                        Ninety-nine Years of War
            Ließen keinen Platz für Sieger                       Left no room for winners
            Kriegsminister gibt's nicht mehr                    There is no longer a Minister of War
            Und auch keine Düsenflieger                        And also no jet planes
            Heute zieh' ich meine Runden                      Today I'm doing my rounds
            Seh' die Welt in Trümmern liegen                 See the world in ruins
            Hab' 'n Luftballon gefunden                          Found a balloon
            Denk' an dich und lass' ihn fliegen               Think of yourself and let him fly

An English version of the song, entitled 99 Red Balloons, was released shortly thereafter, with somewhat different lyrics, changed to scan with the original melody, and with a more American-centric story. However, many felt it lacked the punch and ennui of the German original, and it failed to achieve the same chart success. My always limited German language skills have atrophied with time, but I am inclined to agree.

But Wait, There Are More!

In the wake of the balloon shootdown, questions were asked; fingers were pointed; answers were demanded. Were there other, heretofore undetected, balloons floating over the United States and Canada? It turned out the answer was yes, and UFOs, literally – unidentified, flying objects – were back in the news. Soon, an unidentified balloon was reported to be drifting over Alaska and heading into the Canadian Yukon, followed by another over Lake Huron, the latter described as “small, metallic balloon with a tethered payload below it.”

After diplomatic consultations between the U.S. and Canada, both were shot also down by the U.S. Air Force. (Technical aside:  One should be impressed at NORAD’s surveillance capability, which can identify extremely small objects in the large airspace over North America.)

Were aliens in flying saucers involved?  Was North America under threat?  Was there a new Cold War “balloon gap,” analogous to the infamous “missile gap?” Inquiring minds wanted to know.

In a “you can’t make this stuff up” farce, a comedic mashup of fiction and reality – Dr. Strangelove (“You can’t fight in here.  This is the War Room!”), Neunundneunzig Luftballons, and national security fears – the news was soon filled with speculation that the United States may have used a $100M advanced fighter jet and $500K missile to shoot down a $250 amateur science project, a pico-balloon carrying a ~10 gram, extremely low power amateur “ham” radio transmitter. (See Hobby Club’s Missing Balloon Feared Shot Down By USAF.) 

The group potentially responsible for a balloon traversing America’s airspace is – wait for it – the Northern Illinois Bottlecap Balloon Brigade (NIBBB), a group of pico-balloon amateur radio  and science enthusiasts whose club was named in homage to the 2009 animated movie Up. If you are thinking these people hardly look like a threat, you would be right; the NIBBB notes that the club’s youngest member is only eleven years old.

One of their balloons, with amateur radio call sign K9YO-15 (balloons use the FCC call sign of the owner and a distinguishing number), was launched from Libertyville, Illinois on October 10^th, 2022 and had circumnavigated the Earth multiple times before the club lost contact with it.  The balloon’s last reported position was at 38,910 feet off the coast of Alaska on February 10^th, 2023.  The NOAA HYSPLIT wind model projected the balloon to be over the Yukon on February 11, 2023.

On that same day, NORAD downed an unidentified airborne object in the Yukon. The Pentagon issued a press release, and the Canadians searched briefly for the downed fragments but found nothing. 

However, given the small size of the balloon and the remote environment, it is unlikely that any debris will ever be found, leaving it uncertain if K9YO-15 was indeed the object shot down.  The NIBBB responded with equanimity as well, noting:

As has been widely reported, no part of the object shot down by the US Air Force jet over the Yukon territory has been recovered. Until that happens and that object is confirmed to be an identifiable pico-balloon, any assertions or claims that our balloon was involved in that incident are not supported by facts.

Even as the klieg lights of sudden and unexpected fame can be harsh and unforgiving, the NIBBB website shares the group’s passion for pico-balloons and amateur radio, including detailed construction details and a bill of materials.

That’s quite the story, with lots of jargon.  Let’s deconstruct this just a bit, starting with pico-balloons.

That’s My Science Project

What’s a pico-balloon, you say?  As the name suggests, it is a very small balloon, such as the Mylar kind filled with helium one might purchase from a hobby store or a florist shop. As anyone who has brought one to a birthday or graduation party knows, they can be purchased for a few tens of dollars, and are at most a few feet in diameter.

Most such balloons are soon discarded, though a few escape, intentionally or unintentionally, and float into the sky, never to be seen again. How high can they go?  The answer depends on the characteristics of the balloon material, gravity, and the Archimedes principle.  Scientifically, a balloon rises until it reaches neutral buoyancy (i.e., the balloon’s average density equals that of the surrounding air and the buoyant force balances gravity).

Simple Mylar party balloons are unlikely to rise to altitudes beyond 6-9 kilometers (20,000-30,000 feet), limited by either helium leakage or material stresses. In contrast, hardier balloons made of polyethylene or latex are neutrally buoyant up to fifteen kilometers (~50,000 feet). Bear in mind that these pico-balloons are much, much smaller than weather balloons ,and are often less than one meter in diameter. As such, their payload is also quite small, only 10+ grams (0.02 pounds for those of us in non-metric countries), and they are easily affected by weather, even frost.  For this reason, experienced pico-balloon aficionados prefer balloons that can rise to higher altitudes above most weather.

Under Federal Aviation Administration (FAA) 14 CFR Part 101, such small balloons can be launched by anyone, anywhere in the United States, without the need for FAA approval or reporting.  The only requirements are that the balloon’s payload weighs less than six pounds, and the balloon is launched in an unrestricted area. Given the low weight of pico-balloons, the jet stream can easily carry a balloon around the world, and there are many tracked and reported cases of such circumnavigation.

Why might anyone want to do this?

The truth is that people do such things because they are possible and they are cool. After all, humans did not build the Gossamer Albatross to usher in a new era of human powered aviation. They did it because building a human powered airplane that could cross the English Channel was the union of technological prowess and human inspiration; it was the epitome of cool. (See Science: It’s About the Wide-Eyed Wonder.)

Amateur radio pico-balloon enthusiasts are no different. Yes, the technical answer is that amateur “ham” radio enthusiasts often use pico-balloons to study the atmosphere and radio propagation characteristics, but the joy is in the construction and the operation. In such cases, the payload consists of a GPS tracker for location, a WSPR or APRS radio transmitter, an inexpensive micro-controller such as an Arduino to manage the transmitter, and a few solar cells to power everything. A complete package, including the balloon, can be assembled for at most a few hundred dollars.

What’s WSPR, you say?  It’s an acronym for Weak Signal Propagation Reporter (WSPR). The protocol was designed by Nobel Laureate and avid ham radio operator, Joe Taylor, and is intended to test propagation paths on the medium frequency (MF) and high frequency (HF) amateur radio bands using very low (and I do mean very low) power transmissions. A typical WSPR message contains the licensed amateur radio operator’s call sign, the Maidenhead grid locator of their rough location, and the transmitter power in dBm.  The entire message fits in 50 bits. Transmitting such a 50-bit WSPR message takes 110 seconds at 1.46 baud.

Okay, let’s now reconstruct the phrase “pico-balloon carrying an extremely low power amateur ham radio transmitter.”  There are several design options of varying price and sophistication, but here are the elements of a globe-circling amateur balloon, all within the expertise of amateurs and even (with supervision) high school science enthusiasts

• One small balloon (~$50 from Yokohama Balloon or ~$150 for a slightly more expensive research version from Scientific Balloons)
• Helium canister for balloon charging (~$75)
• Lightweight (~10 gram) payload, custom designed or purchased (e.g., ~$90 ZachTek WSPR controller or ~$130 QRPLabs LightAPRS-W tracker) h containing
  • GPS tracker
  • Microcontroller (e.g., Atmel 8-bit AVR microcontroller)
  • WSPR and/or ARPS transmitter
  • Solar cells for electrical power
• Dangling 0.005-inch wire dipole antenna of resonant length for the desired frequency (~33 feet for the 20m radio band)
• Open source software
  • Microcontroller configuration (call sign, Maidenhead locator, …)
  • Flight path prediction (e.g., NOAA HYSPLIT model or com)
  • Pico-balloon tracking collaborations (e.g., https://amateur.sondehub.org/ or http://lu7aa.org/wsprset.asp)

I would have been thrilled beyond measure to have had the opportunity to participate in such an activity as a student. Such projects are precisely the kind of experiential science education that can excite students, filling them with the realization that science is not a desiccated recitation of facts, but rather is an endeavor filled with passion and excitement. We need many more such experiences if we are to attract the millions of additional STEM workers this country needs to remain globally competitive. (See For Science and Society, The Future Begins with Better Dreams.)

My Geek Bona Fides

During the COVID-19 pandemic, my days (and often my nights) were consumed with managing the evolving crisis and ensuring the academic enterprise of the University of Utah continued with as few disruptions as possible.  Given the stress and uncertainty, I regularly counseled others to spend some time away from work doing something fulfilling and emotionally satisfying, lest they be overwhelmed by uncertain and rapidly changing events. (See The Simple Things Matter, Most of All Now.)  Knowing I should heed my own advice, I sought solace and distraction in my own happy place – science and technology. 

My first stop was my digital astrophotography hobby, which though always rewarding, proved uncertain, given the vagaries of Utah weather.  More practically, the late nights were incompatible with my day job due to sleep deprivation. After that, I turned to writing (See My Balm in Gilead) and, finally, to my tried and true friend, computing technology.

First, I developed an end-to-end suite of software for low-power, wireless sensors, using an inexpensive ESP32 microcontroller, LoRaWAN low power wireless transceivers, environmental sensors, and open source software tools.  All of this served as a citizen science complement to the larger SAGE project of which I am a part. (See Cyberinfrastructure: Lots of Room at the Edge.)

With the sensor suite complete, I studied for and passed the three FCC-mandated examinations – technician, general, and amateur extra – to become a fully licensed amateur radio operator.  The study materials gave me an opportunity to refresh my memories of electricity and magnetism (E&M) from undergraduate physics and also to delve more deeply into antenna and radio transceiver design.  The geek in me wished the American Radio Relay League (ARRL) examination materials had been a bit more mathematical, as many of the concepts are best understood via their roots in Maxwell’s equations.

All of which brings me back to pico-balloons and the “because we can” excitement.

First, armed with an ICOM-7300 radio and a trusty PC, I fired up WSJT-X and transmitted WSPR messages (radio callsign KK7EUJ, maidenhead grid locator DP40cp) at one milliwatt on the 20M HF band (14.0956 MHz) using my house’s gutter as an end-fed-random-wire (EFRW) antenna. Reflecting the effectiveness of WSPR encoding in a high noise, low power environment, these WSPR messages have been received by amateur stations across the entire continental United States, as shown at the right.

Second, I have written extensive code for the ESP32 and its smaller, 8 and 16 bit cousins, the ESP8266 and Arduino, the latter based on the Atmel 8-bit AVR microcontroller. Just as the proliferation of powerful, open source software has enabled construction of complex toolkits, so too have open source libraries of hardware components and powerful, open source design tools (e.g., KiCAD) facilitated design of inexpensive, custom hardware boards. As a complement to the wireless sensor library, I have also specified the design of custom printed circuit boards (PCBs), with GPS, sensors, and ESP32, which have been inexpensively fabricated in varying quantities by one of many fabrication firms.

Put another way, I realized I had followed the same path as my pico-balloon compatriots, using open source software and easily designed open source hardware to build new and interesting projects, ones that in the past would have been only within the reach of nation-states.

Coda

Sino-American relations remain deeply strained, and not just because of the surveillance balloon. The global battle for 21^st century geopolitical hegemony is underway, waged via economic competition, trade sanctions, diplomacy and political rhetoric, and military posturing. There seems little doubt that we are entering a new political era, one fraught with uncertainty.

In the midst of all this, I was on Capitol Hill just days after the downing of the Chinese balloon, discussing the importance of greater U.S. research investments and developing the domestic STEM workforce. I have been working with the National Science Board (NSB) on the intersection of dual use technologies, basic scientific research, economic competitiveness, and national security. In addition to a variety of meetings on Capitol Hill, I have invited the White House National Security Advisor, Jake Sullivan, to join the NSB in discussions.

Meanwhile, no alien UFOs have been confirmed, unless the definition of alien includes a loose affiliation of geeks (young and old), all fascinated by the combination of electronics, software development, atmospheric sciences, mathematics and coding theory, ionospheric electromagnetic propagation, and small balloons, and all wishing to share their scientific passion with others.  Could they be?  They might be. The signs and omens are all there. Yes, they are.  I am proud to say those are my people.

I am a geek, albeit an aging (and balding) left-handed, myopic one, some of those adjectives necessarily being implied by the geek nom de guerre.  I will try not to trigger an international incident by launching any pico-balloons of my own, but I will not rest in encouraging the passion of discovery in young minds.

Nena WSPRed rightly about the Luftballons:

        Man, who would have thought
        That it will come to this
        Because of ninety-nine balloons

Public Higher Education: A New Social Compact for Innovation

N.B. This is a long post, discussing both higher education futures and the economic challenges of STEM graduate education. 

As Lewis Carroll wrote in The Walrus and the Carpenter,

"The time has come," the Walrus said,
"To talk of many things:
Of shoes, and ships, and sealing wax -
Of cabbages and kings,
And why the sea is boiling hot,
And whether pigs have wings."

I speak, of course, of the complex, often eroding, and – to any outsider, seemingly nonsensical – financial models of most U.S. public higher education institutions. Some of this necessarily reflects their manifold and expanding mission, in varying parts social service agency, hospitality service operator, real estate developer, performing arts center, museum complex, sports franchise, and healthcare system, alongside the enduring and core missions of education, research, and community service.

Even as public colleges and universities struggle financially, the rising cost of higher education has put a college degree beyond the financial reach of an increasing fraction of students, with many increasingly questioning the value of a degree, as the post-World War II notion of the American Dream fades with it.  Meanwhile, the U.S. leadership of the global academic research enterprise is slipping away.

What happened?  How did college become unaffordable for so many? How did we lose touch with so many of society’s pressing needs? Where did we lose our way? More importantly, can we speak honestly and forthrightly about the problems and their solutions?

The Brave and Uncertain New World

Today’s colleges and universities are so much more than curricula and classrooms, social networks, and athletic teams. Universities all too often find themselves not just educating, but also helping impoverished students find a place to sleep overnight or even find a meal to eat.  They may be providing essential mental health services for students in crisis or addressing intimate partner violence and sexual assault.  With the public social safety net increasingly limited, universities are now often the helper of last resort, with resource demands and service expectations unknown a generation ago.

All of this is occurring amidst lagging state and federal support, and the limited backfill afforded by philanthropy. Consequently, university finances have become a byzantine web of cross-subsidies and interdependencies, with a surfeit of needs and a deficit of resources.  In turn, this has created competing agendas and incentives, an expanding backlog of deferred building maintenance, and increasing and worrisome addictions to rising tuition and cheap labor.  It’s a Faustian bargain at best, and the future of our public colleges and universities now hangs in the balance, measured in dreams deferred and opportunities lost.

At their best, U.S. public higher education institutions have been a transformative societal force, uplifting millions into higher socioeconomic classes via inculcated knowledge, skills, and acumen, and creating the new knowledge that has illuminated our world, transformed our economy, advanced our national security, and improved our health.  The generational impact of the G.I. Bill in educationally and economically uplifting millions of returning veterans is undeniable, as have been the benefits of the National Defense and Education Act’s investments in higher education and science, technology, engineering, and mathematics (STEM).

Pointedly, the cracks in the façade of U.S. higher education excellence are increasingly obvious and concerning, even as global competition in the knowledge economy intensifies. Although U.S. higher education institutions remain the envy of the world, other global universities, notably in China, are increasingly rivaling or eclipsing the U.S. in research citations and impact, as those countries invest in higher education.

Make no mistake, an educated and empowered populace of a community, a state, a region, or a nation remains the only truly renewable resource, the key to public health, national security, and economic competitiveness. A clarion call for reckoning now echoes loudly throughout our society, as the past speaks urgently to the present about the exigencies of future. It whispers about what was, what is, and for better or for worse, what might be. (See Academia:  Who We Are and Why It Matters and Innovation Imperiled: Golden Eggs No More.)

How Did This Happen?

Some of U.S. higher education’s challenges are a consequence of our fractured and litigious society, the pitched cultural battle between the left and right over perceived ideological purity, and, concomitantly, declining societal support for education as a universal public good rather than a perceived individual and private benefit. It’s been further exacerbated by the discordant, self-reinforcing echo chambers and ever shrinking sound bites that now pass for public debate. (See From Broadcast to Narrowcast: Finding Signal in the Noise and Contemplative Reflection and Instantaneous Communication.)

Nor are we in higher education without culpability, having been complicit in protecting the ivory tower, slow in responding to societal needs, and smug in our defense of often hidebound traditions and byzantine bureaucracy. There are also substantive ways to reduce costs judiciously and prune non-essential services, while retaining the core educational mission, and these should be pursued assiduously. However, such changes are unlikely to substantively change the cost structure of higher education.  Even with automation and expanded online classes, education remains a labor-intensive endeavor.

In public higher education institutions, these pernicious realities have led to a decline in inflation-adjusted state support, compensated only in part by rising student tuition.  The latter, of course, has made college increasingly unaffordable and precipitated a student debt crisis.  Concurrently, colleges have found themselves in the crosshairs of the national political battle, critiqued by some for indoctrinating students with liberal values. It's a wicked problem, a catastrophe of potentially existential proportions.  (See On Catastrophes and Rebooting the Planet.)

Not surprisingly, the result has been increased questioning of the cultural and economic value of higher education, even among those predisposed to believe its enduring importance.  As an example, a recent survey by the Association of American Universities showed that roughly half the respondents felt a college degree from one of the leading research universities in the United States was valuable but not worth the cost. Without doubt, a college degree remains an economically and culturally wise choice for many, when coupled with appropriate career counseling, but we have not done nearly enough to address rising education costs, respond to market demands, meet students where they are, and prepare them for a lifetime of learning and periodic retraining. (See Higher Education in the 21^st Century.)

Conventional wisdom says the first step in any addiction treatment and recovery program is personal honesty – admitting that one has a problem, then taking determined action. Alas, I fear we in higher education are still in denial. (See Pinball, War Games, and Universities.)

Our addiction takes many forms. On the staff side, it includes an increasing number of poorly paid adjunct lecturers, often euphemistically called career-line faculty, who teach a large fraction of our undergraduates. It includes the underpaid support staff who maintain our buildings and grounds, and others who advise and support students.  It also includes many of our faculty outside the professional schools, particularly in the liberal arts. All struggle to afford housing and find child care, and some are forced to avail themselves of federal assistance programs. Even more troubling are those contract teachers whose work is spread across multiple institutions; many lack health insurance.

On the student side, it includes well-documented undergraduate tuition and affordability challenges, as well as mental health support, and food and housing insecurities. Convolved with these is an ugly truth – fewer than half of all college students will graduate within four years, leaving them saddled with debt, but lacking the intellectual “union card” – a college degree – that creates differential market value.  

Even there, the near-term financial return on investment (ROI) differs markedly based on the type of degree earned and the institution from which it was received. Meanwhile, high paying, intellectually satisfying jobs in the skilled technical workforce go unfilled, ones for which training is much less expensive and time consuming than a college degree.

It also encompasses an issue little discussed outside academia, namely the sad financial state of graduate students and post-doctoral research associates (post-docs).  Finally, it is reflected in our focus on academic rankings and prestige, often at the expense of the communities and states who need our help.

Before delving into these dynamics, it’s worth remembering how much the world of higher education has changed in my professional lifetime.  Herewith is a personal story, which I will use to illustrate shifting academic finances and affordability.

A Personal Tale

Sometime after the K-T extinction, when mammals began filling the newly vacated ecological niches, I started college.  The year was 1975, Saigon had just fallen, Nixon had resigned the year before, and platform shoes and polyester leisure suits were all the rage. I will begrudgingly admit to owning the former, but I vehemently deny ever owning or wearing the latter – I may have been poor, but I did have some sartorial principles.

As a first-generation college student and poor scholarship kid from the Arkansas Ozarks, I was able to attend and graduate from college without debt, thanks to a combination of academic scholarships (National Merit Scholarship Program and university), Basic Educational Opportunity Grants (now Pell) for low income students, and the small amount of money I could save working at summer jobs; my parents were unable to help me financially. I lived like a pauper, with a social life limited to free lectures and cultural events, and my wallet was empty, save for my student ID card and a driver’s license.  The latter was of about the same practical value as a launch authorization for a Mars mission, as I lacked both a working heavy lift launch vehicle and an automobile.

In the process, I became intimately acquainted with student dormitory housing and the culinary delights of fixed menu cafeteria food – eat what was served or go hungry.  Finally armed with an aging and cantankerous Ford Galaxie 500 – the green monster – I lived in a trailer park mobile home for a bit, where I learned to appreciate a variety of nutritional price performers, including “four for a dollar” Kraft mac-and-cheese (Pro tip: It’s better with milk, even powered milk, but it’s edible even with water). Nevertheless, in my mind, I was rich beyond imagining. (See Transforming Lives via Public Higher Education.)

With an undergraduate degree in computer science now within reach, I told my father that I had higher ambitions. I planned to attend graduate school, rather than get a job, one that would have paid substantially more than my father earned as a sawmill laborer.  In response, he looked at me sadly and said, “Son, you know we can’t ever afford that.”  I just smiled and said, “Dad, here’s the amazing thing. It won’t cost anything.  In fact, they will pay me to go to graduate school!” Already poor, I did not fully realize how many more years of deferred financial reward I had unknowingly embraced. Free can prove extraordinarily expensive.

When I began graduate school in January 1979, annual inflation in the U.S. was hovering near 14 percent, proving the truth of the hoary guns versus butter economic adage and highlighting the hidden costs of the Vietnam War, which begat the failed Whip Inflation Now campaign. Because I completed college early, I began graduate school “off cycle” in January, with a fellowship that did not start until September.  That spring, I was a teaching assistant (TA), lecturing on the joys of FORTRAN programming to engineering students at 7:30AM as I watched the sun rise over the snow-covered Purdue campus.

As a TA, my monthly stipend was a princely $368/month, which covered food and a room in a graduate student residence hall, with perhaps ~$20/month of discretionary money. I felt rich when my graduate fellowship began that fall, for it was a munificent $400/month, about $1,400/month in today’s inflation adjusted dollars.  Alas, with no cost-of-living adjustment (COLA), the $400/month stipend was quickly eroded by inflation, moving me from poor to poorer. Then my old car died, leaving me an involuntary pedestrian, constrained to live in on-campus graduate student housing until I completed my Ph.D. in 1983.

Graduate school now behind me, I drove away to my new academic job in a cheap, subcompact car, financed on the strength of my future income, all my worldly possessions inside – some old clothes, a battered popcorn popper, a cheap guitar, some books and graduate school notes, and a bicycle. After my arrival at my new home, I began my career as an assistant professor, living in a one bedroom apartment.

Lest the de minimis nature of my material state remain in any doubt, my little apartment was soon burglarized and ransacked, but nothing taken, for I lacked anything worth stealing. I’d have been annoyed if my bicycle has been stolen, but outraged if my graduate school books and notes had been pillaged. The only other, potentially more devastating insult would have been if the burglar had left a $20 bill and a note, saying “You need this more than I do.”

Shifting Student Economics

What is the moral of my little story?  First, the 1970s and 1980s economics of higher education afforded me an opportunity to realize my dreams.  It was not easy, but like thousands of other students, I completed both undergraduate and graduate degrees and emerged with no debt. For that, I will be forever grateful.

Alas, it is increasingly clear that my path is no longer easily replicable. Undergraduate tuition at a state university in the 1970s was just a few hundred dollars per year, an amount covered by any reasonable combination of an academic scholarship, BEOG/Pell grant, summer job savings from a minimum wage job, and for families wealthier than mine, some parental contributions.

Today, undergraduate tuition at most public institutions exceeds $10,000, with comparable student room and board costs. Put another way, a four-year college degree at a public university can cost substantially more than $100,000, while the maximum Pell grant is under $7,000, and just recently increased by $500.  Further confounding the economics, four years was once the nominal time to complete a college degree. Today, we most often report six year graduation rates, as students work full or part-time jobs, delaying graduation as they struggle to cover college costs.

As for graduate school, I never paid graduate student tuition – I had no idea what it even was. It was once common to waive tuition for students employed as TAs or research assistants (RAs), and it was included in fellowship awards.  Such graduate tuition waivers are now less common, particularly in STEM disciplines, as universities seek to recover tuition from federal research grants and contracts.  In turn, this has exacerbated pressures on the federal research enterprise – more on that later.

Meanwhile, as data from the National Science Board and the National Center for Science and Engineering Statistics (NCSES) show, the percentage of graduate STEM degrees awarded to international students continues to rise.  The fact that the U.S. is a magnet for global talent is a tremendous competitive advantage. 

Not only do many of those international students pay tuition, many remain in the U.S. after school and contribute greatly to the vitality and economic competitiveness of the country. A recent National Bureau of Economic Research study highlighted the disproportionate positive contributions of immigrants on the entrepreneurial ecosystem.  But this begs an important question – where did the domestic STEM graduate students go?

Graduate School: Yes, It Is Different

If one is pursuing a terminal professional degree (e.g., an MS or MBA), graduate school may seem like an in-depth continuation of the undergraduate experience, filled with classes and projects, albeit focused on one’s chosen discipline. For those aspiring to a research Ph.D., the experience is strikingly different. (See Research: When There Are No Answers in the Back of the Book.)

Many Ph.D. students serve as teaching assistants (TA), research assistants (RA), or graduate fellows (i.e., they are supported by the university, either with institutional funds (TAs) or research grant or contract funds (RAs)). In STEM fields, RA support is common, though rare in the humanities. One is not quite staff, and certainly not faculty, but neither is one fully a student. 

TAs assist faculty, hold office hours, grade homework sets, lead discussion sessions, and, in some cases, as I did, even have full responsibility for a class. RAs and fellows are research apprentices, working in faculty-led research laboratories, often, but not always on projects related to their research, hoping to collect the requisite data and experimental validation needed to craft a Ph.D. dissertation. Post-docs labor in a different netherworld, hoping to build a resume worthy of an academic faculty position.

In a perfect world, all of these young scholars would be mentored, nurtured, supported, and treated as future members of the profession, whether as future faculty, industry or laboratory technical staff, or government experts.  In many cases, they are, but all too often they are treated – implicitly or explicitly – as cheap labor whose primary function is to further the career ambitions of the faculty advisor or to satisfy the institutional needs for inexpensive instructors. Amidst this class system, TAs and RAs battle the existential angst of graduate school, uncertain when – or if – they will graduate, dependent on the continuing good will of their supervisor and advisor for intellectual and emotional support, letters of recommendation, and certifications of academic progress.

Graduate school will always be intellectually challenging – pushing the boundaries of the known is never easy.  As Jimmy Dugan said in A League of Their Own, “It’s supposed to be hard.  If it wasn’t hard, everyone would do it. The hard is what makes it great.”

However, intellectually difficult, graduate education can and should be a nurturing and supportive environment.  Indeed, there are very good ideas for how to improve this process, reemphasizing the role of mentorship and respect, reducing the power inequities between advisor and advisee, and better educating students to translate their knowledge into practical impact. (See, for example, the recent NAESM study, Graduate STEM Education for the 21^st Century.) Sadly, there can be strong faculty resistance and institutional disincentives to implement these changes.

More Than Genteel Poverty

The stress of graduate school is made more difficult by the vow of poverty taken by graduate students.  Yes, one is paid to be a TA, RA, or fellow, but the compensation is low, and increasingly inadequate to meet the basic cost of living needs in many American communities. For graduate students with dependents, those from traditionally under-represented groups, and those with little intergenerational wealth, the financial challenges can seem bleak, especially with annual inflation in the U.S. now above six percent.  (See Missing Talent, Missing Opportunities, and Losing Ground.)

In response, graduate students are increasingly turning to strikes and unionization to increase their bargaining power. The recent strike by University of California graduate students and post-docs – the largest in U.S. history – brought salary and working environment issues into stark relief. The newly negotiated contract raises the minimum salary for nine months to just over $34.5K, with a slightly higher rate of $36.5K at UC-Berkeley, UCSF, and UCLA, in recognition of the high cost of living there. Contrast that with MIT’s Living Wage Calculator, which estimates a “living wage” for one adult without children at $50,463 in Alameda County, the home of UC-Berkeley.

Why has it been so challenging to raise graduate student salaries?  In part, it is related to the financial exigencies facing universities, but there is a complicating factor related to federally-funded research. Federal research grant sizes have also not kept pace with inflation, and raising RA stipends would necessarily reduce the number of funded RAs.  Moreover, it would create strong equity pressures to raise TA salaries, which are funded directly by universities, to comparable levels.

Again, let me illustrate the challenges with a personal example. My very first National Science Foundation (NSF) grant in 1984 was for ~$55K (~$155K in current dollars), enough to cover two months of summer salary for me for three years, fund travel and supplies, and cover the salary of one graduate student for three years, the standard NSF grant period.  Mind you, my nine-month academic salary was $30K, the then common starting salary for assistant professors in computer science. Moreover, the university waived tuition for funded research students, meaning tuition was not charged to the grant, and the cost of health benefits was much lower.

If one examines the history of NSF funding from the Merit Review Digest,, the median award size in 2020 was ~$150K, and it has remained largely stagnant since 2011. Any simple analysis shows that this number is now inadequate to support activities similar to those I pursued in 1984. The stipend, benefits, and tuition for a single graduate student over three years easily exceed this amount, leaving no funds for summer faculty salary, supplies, or research travel. This means faculty must scramble to write more proposals – to multiple funding agencies – to create a portfolio of awards with differing objectives, timescales, and funding levels to maintain a research laboratory.

To be sure, money – current or deferred – has never been – nor should it be – the primary reason to attend graduate school or become an academic. However, a stark and growing financial disparity between private sector remuneration and graduate student compensation can be an insurmountable social and economic barrier to many. 

Houston, We Have a Problem

We’ve now talked of many things.  The sea, though not boiling hot, is experiencing global warming due to climate change.  As for pigs, absent a lower gravity environment and deeply unethical genetic engineering that substantially exceeds our current capabilities, they really can’t fly.  But, U.S. higher education and its research enterprise face very real problems.  Let’s summarize the challenging calculus.

Public Higher Education Dynamics

• Society is increasingly questioning the value proposition of public higher education, both as a valuable return on family investment in undergraduate education and as a force for good – culturally and practically – in our cities, our states, and our country.
• The upcoming (2025) demographic cliff – a decline in the number of U.S. high school graduates – will create even greater competition for students and their tuition dollars. Some smaller institutions, particularly private ones, will be forced to close their doors, and many public institutions will be forced to retrench and eliminate programs.
• The student debt crisis is real, with the burden falling disproportionately on lower income, first generation, and traditionally under-represented students.
• We have made a college degree the default social objective for all high school graduates, even as high paying jobs in the technical trades go unfilled.
• Public colleges and universities continue to raise undergraduate tuition to offset declining state support. In most states, these form an “x curve,” with tuition now a larger contributor to college and university budgets than state appropriations.
• Financial pressures have triggered a shift toward lower paid, contingent faculty, who are hired only to teach, rather than conduct a mixture of teaching, research, and service.
• Aspirations are good, but mission creep has led many of our community colleges to aspire to become four-year institutions, our four-year institutions aspire to become research universities, and our research universities to aspire to flagship leader status. An appropriately differentiated — and interconnected — educational system serves different educational niches.

I believe it is time for us to collaboratively and thoughtfully rethink the social compact surrounding public higher education.  This includes defining the 21st century knowledge economy version of the public land-grant university, which was originally charged to

… teach such branches of learning as are related to agriculture and the mechanics arts, … in order to promote the liberal and practical education of the industrial classes in the several pursuits and professions in life.

It means admitting that we in higher education do not have all the answers, or even know all the right questions. It means a true conversation – of give and take and mutual respect – where all parties put aside preconceived notions about the best path forward. Concomitantly, it means using our intellectual expertise to deliver perceived and real value to our communities, regions, and states, addressing the social, economic, technical, and cultural issues that now bedevil our society.

It also means embracing differentiated higher education roles, recognizing that each type of institution is a critical part of our higher education ecosystem. Finally, it means better career counseling about expected outcomes from differing courses of study, including celebrating the skilled technical workforce (aka, the skilled technical trades) and importance of preparing graduate students for non-academic careers. Only by doing so will we change the rhetoric about higher education’s utility and the importance of investing in our children’s and our country’s futures.

Now, let’s turn to the other issue – the challenges facing our STEM domains. (N.B. Equally important, partially overlapping but also partially orthogonal, issues face the liberal arts and humanities.  Both are important, as transdisciplinary education is critical to addressing our societal challenges.)  (See Renaissance Teams: Reifying the School at Athens.)

STEM Research Dynamics

• Completing a STEM Ph.D. can take five to seven years, sometimes more, at low pay and long hours, with continued support dependent on the research supervisor’s ability to secure competitive research funding.
• In many STEM disciplines, four or more years of post-doctoral study are then expected, also at low pay and with long hours, if one intends to seek an academic position. During this process, all too often we fail to provide adequate mentoring and career development, as most post-docs take industry positions, rather than academic ones.
• Academic faculty salaries are often not competitive with those in the private sector, and one may never recoup the five to ten years of “lost income” during one’s graduate and post-graduate education.
• Except in growing STEM disciplines (e.g., computing), the competition for tenure track faculty positions is increasingly intense. Moreover, with few universities growing, the availability of such positions depends on faculty retirements and domain fit. This is exacerbated by current faculty deferring retirement and many tenure track positions being replaced with lower paid adjunct instructional faculty.
• U.S. leadership in STEM, once unquestioned, is increasingly in doubt as China builds world-class research universities and develops critical technologies such as AI, quantum computing, and biotechnology.
• Because the growth in federal research funding has not kept pace with workforce growth and inflationary pressures, the probability that a research proposal will be funded is now lower than in the past. Many NSF programs now have proposal success rates in the teens, and median award sizes have been stagnant for a decade.  Similarly, the mean age to receive a National Institutes of Health (NIH) R01 grant (the academic gold standard in biomedicine) is now well over 40 years.
• Universities conduct research to build prestige, attract high-quality faculty, and explore the future. However, in a narrow sense, research loses money, as the direct and indirect costs (overhead) from federal grants and contracts rarely cover the true costs of research facilities, libraries and serials, research computing, pre-award and post-award services, research compliance, and a host of other services. The deficit is paid by institutional funds.

The career opportunities enabled by graduate degrees vary widely across disciplines. In some, a Ph.D. is de rigueur; in others, it is optional, though the value is intellectually and economically demonstrable; and in some cases, it may not be worth the opportunity cost of delaying entry to the workforce.  In all cases, the personal financial costs of graduate education continue to grow.

In my own discipline, computer science, the opportunities and the resources available in industry often make it more attractive for domestic students to either skip graduate school entirely, or leave with an M.S. rather than a Ph.D. In part, these opportunities explain the relative dearth of domestic Ph.D. students in computing.  Conversely, graduate school provides an intellectual, cultural, and visa entry point (via optional practical training and H-1B visas) for international students.

Finally, we must never forget that humans pursue science and engineering research for the joy it brings, just as do artists and musicians. Discovery illuminates the great unknown and satisfies the human curiosity older than history. (See Science: It’s About the Wide-Eyed Wonder and The (Scientific) Good News.) In so doing, it provides the insights and ideas that advance our society, improve human health, and secure our economic future. 

It’s not always about the money, but neither should the lack of money close the door on talent. Nevertheless, the invisible hand of economics is real, tempered by market realities, career ambitions, and credential expectations. We must recognize and address the economic challenges that bedevil our domestic graduate students and our faculty, which limit both their numbers and their diversity.  It’s ethically and practically important.

Coda

There can be no doubt that our current higher education and research enterprise has both great strengths, but also deep flaws. Higher education is too expensive for too many, and too many of our public colleges and universities are struggling financially. Simultaneously, our research ecosystem is stressed and falling behind our global competitors.

As a country, I believe we must put aside our differences – at least long enough to recognize our shared fears, hopes, and dreams – and rethink our approaches and renew the social compact. Only then can we build the social consensus needed to invest in the future appropriately. In so doing, we can create an environment that not only redresses current university funding deficits and academic dynamics, but also responds to societal needs and expectations.  In the process, we can and must build a culture shaped by what Lincoln called the “better angels of our nature.” (See For Science and Society, the Future Begins with Better Dreams.)

Computing Futures: Technical, Economic, and Geopolitical Challenges Ahead

In February 2022, I wrote an essay entitled American Competitiveness: IT and HPC Futures – Follow the Money and the Talent, where I outlined a series of challenges facing the state of the global information technology ecosystem. These spanned the shifting locus of innovation and economic power, from those companies we have traditionally called computing vendors (Intel/AMD, HPE, and IBM) to the hyperscalers and social media giants (Amazon, Facebook, Google, and Microsoft, along with their Asian siblings), the influence of smartphone vendors (Apple and Samsung), and the artificial intelligence juggernaut driving them all.

I also discussed the end of Dennard scaling and the slowing/end of Moore’s Law, both of which have ended the “free lunch” cycle of faster and cheaper microprocessors every two years. Then, within this framework, I outlined the technical, economic, and cultural challenges bedeviling leading edge high-performance computing, where ever larger machines deliver increasingly small fractions of the peak hardware performance. (It is often the case that complex, multidisciplinary applications achieve single digit percentages of the peak hardware performance.) I ended by discussing the sad state of U.S. STEM education and the implications of these trends for U.S. global competitiveness and national security.

As I had hoped, the essay generated a great deal of federal policy and vendor introspection, as well as considerable Internet chatter. 

Based on many discussions and helpful feedback from several readers – including Doug Burger, Jim Larus, Tony Hey, and John Shalf, I asked my friends and colleagues, Dennis Gannon and Jack Dongarra (this year’s Turing Award winner), to help me formalize my blog observations in a March 2022 arXiv paper, entitled Reinventing High Performance Computing: Challenges and Opportunities. There, we expanded the high-performance computing history lesson and expounded on the rise of cloud computing and deep learning, something Dennis and I were privileged to see first-hand at Microsoft.  Later in March, I participated in an insideHPC podcast, where I discussed several of our observations.

We followed this with a standing room only discussion in November at SC22, with additional panel members Dorian Arnold (Emory), Jack Dongarra, Torsten Hoefler (ETH Zurich), and Kathy Yelick (UC-Berkeley). (For those of you who attended SC22 or are ACM/IEEE members, you can view the panel discussion here.)  Then in January 2023, John Shalf organized a panel discussion with Kathy, Neil Thompson, and me at the Exascale Computing Project (ECP) annual meeting, where we discussed the topic as well.

Finally, the Communications of the ACM has published an abridged version of our arXiv paper in the February 2023 issue.  The latest version of the article is entitled HPC Forecast: Cloudy and Uncertain, and you can listen to Dennis, Jack, and me discuss the maxims and their implications at this link.

A Few Maxims

In both our arXiv paper and our Communications of the ACM paper,  we outlined some maxims that we believe should guide any consideration of advanced computing futures.

• Maxim One: Semiconductor constraints dictate new approaches. The “free lunch” of lower cost, higher performance transistors via Dennard scaling and faster processors via Moore’s Law is at an end. (Most do not realize that the inflection point for transistor costs occurred near the 20 nanometer feature size.  Since then, chip feature sizes have continued to decline, with more transistors per unit area, but the per transistor costs began to rise due to yield challenges and rising fabrication facility costs.) This minimax problem of maximizing chip yields, minimizing costs, and maximizing performance has driven the adoption of the chiplet model, mixing and matching multiple chips via interconnects such as UCIe, which opens new opportunities for scientific computing optimization and domain-specific accelerators.

• Maxim Two: End-to-end hardware/software co-design is essential. Given current semiconductor constraints, substantially increased system performance will require more intentional end-to-end co-design, from device physics to applications. (See Advanced Computing: Integrative Thinking for the Future, for a discussion of end-to-end design.) There is also an important corollary about changes to procurement models at scale.
• Maxim Three: Prototyping at scale is needed to test new ideas. Across multiple axes – semiconductors, chiplets, AI hardware, cloud innovations – the computing system is now in great flux.
• Maxim Four: The space of leading edge HPC applications is far broader now than in the past. Domains dominated by complex optimization problems and time-dependent partial differential equations will always matter, but other areas of advanced computing are of high and growing importance, enabled by advances in deep learning. The AI revolution is also driving the hardware ecosystem in new ways.
• Maxim Five. Cloud economics and deep learning have changed the supply chain ecosystem. The largest HPC systems are now dwarfed by the scale of commercial cloud infrastructure deployments.
• Maxim Six. The societal implications of technical issues really matter, and a diverse and inclusive domestic and global workforce is critical. This speaks to our challenges around workforce development and attracting and retaining talent individuals.

Chips Shortages and Globalization

While we were sounding the alarm about the technical and economic challenges in advanced computing, the public debate on what became the U.S. Chips and Science Act continued apace. Ultimately, the act appropriated money to help “on shore” a larger fraction of U.S. semiconductor manufacturing and authorized, but did not fully fund, major increases in funding for the National Science Foundation (NSF) and the Department of Energy’s (DOE) Office of Science. (See Working the Hill for some perspectives on where my research interests and public policy roles – as chair of the National Science Board, which oversees NSF, and chair DOE’s Advanced Scientific Computing Advisory Committee (ASCAC) – intersect.)

How did we find ourselves in this situation, you might ask?  Simply put, via a series of locally rational and arguably, even short-term optimal decisions, companies focused on quarterly profits and shareholder value. They outsourced chip fabrication and device manufacturing, benefiting from lower offshore costs, and focused on design and integration of the chips into high margin products.  Meanwhile, Intel, the dominant U.S. semiconductor fabricator, struggled with the transition to extreme ultraviolet lithography (EUV) and ceded global semiconductor leadership to Taiwan Semiconductor Manufacturing Company (TSMC). Nevertheless, all seemed well until two things upset the global supply chain – COVID-19 with disruptions in global supply chains and the rising economic tensions and national security concerns involving China. Now, economic uncertainties, declining tech sector profits, and corporate layoffs have further exacerbated the challenges.

The pressure to maximize short-term gains obscured the longer term need to control the means of production, exposing the risks of what my friend Tren Griffin calls the power of wholesale transfer pricing (i.e., the bargaining power of company A that supplies a unique product XYZ to Company B which may enable company A to take the profits of company B by increasing the wholesale price of XYZ.) Put another way, it is very attractive to be a fabless semiconductor company, but only when your semiconductor fabrication partner is both inexpensive and highly reliable.

As Warren Buffet famously said about the hidden risks some companies take, “Only when the tide goes out do you discover who’s been swimming naked.” The U.S. has been swimming naked for quite a while, dependent on international sources for key semiconductor components. If you have tried to buy anything containing embedded electronics – and almost everything these days does, from automobiles to greeting cards – then you know there has been a global chip shortage, manifest in rising product costs and even scarcity.  Nor has high-performance computing been immune from these shortages. To complete construction of the Frontier supercomputer at Oak Ridge National Laboratory, the U.S. Department of Energy had to involve national security priorities to secure access to scarce chips.

How big is the U.S. dependence on Asian semiconductor foundries? As the chart shows, Taiwan has about 65 percent of the global market, with TSMC holding the largest fraction of this, followed by UMC, PSMC, and VIS.  Korea (Samsung) has about 18 percent, China (SMIC and HHGrace) has about 5 percent, and the United States is in the “other” category – the final 12 percent – along with Global Foundries.

Why don’t we just build more fabrication plants, you say? We are, stimulated by Chips and Science Act subsidies, but building new facilities takes time and serious money.  To put this in perspective, a state-of-the-art fabrication line for high-end microprocessors and graphics accelerators (GPUs) now costs at least $10B and is rising. Moreover, much of the chip shortage is not at the bleeding edge, but for older chip technologies (e.g., for car electronics) that are fabricated on fully depreciated semiconductor fabrication lines (e.g., 45-28 nanometer feature sizes).  Vendors rarely build new lines for older technologies, they simply extract cash from existing fabrication lines, and until the recent economic downturn, there was unprecedented demand for devices normally fabricated on these lines.  (The New York Times has a reasonable summary of the global semiconductor dynamics.)

National Security

Semiconductors have become more than just an economic engine; they are now critical to any country’s national security and global aspirations.  Indeed, semiconductor manufacturing and supply chain control have become proxies in the ongoing battle for global hegemony.  In a bit – but only a bit – of hyperbole, at the World Economic Forum event in Davos, Intel’s Pat Gelsinger recently called chips the new oil of geopolitics. The U.S. Department of Commerce recently tightened export controls around China, both to purchase and to manufacture certain high-end semiconductor chips used in military applications. (See the October 2022 Department of Commerce declaration.)

Meanwhile, the new U.S. National Security Strategy focuses on protecting semiconductor supply chains and continued advances in semiconductors, advanced computing, and next-generation communications, among others. The U.S. National Security Advisor, Jake Sullivan, has highlighted four pillars of U.S. strategy:

• Investing in the science and technology ecosystem,
• Nurturing top STEM talent,
• Protecting technology advantages, and
• Deepening and integrating alliances and partnerships.

All these goals emphasize three families of technologies, including (a) biotechnologies and biomanufacturing, (b) clean energy, and (c) computing-related technologies, including microelectronics, quantum information systems, and artificial intelligence.

Nor are these issues of concern only to the U.S.  Korean leaders have raised similar concerns. Whatever one might wish to be true, it is increasingly clear the world is moving rapidly into a post-globalization era, with the increasingly likelihood of a 21^st century non-kinetic Cold War, albeit one with very different dynamics than the one of the 1950s to 1980s.  The global economy – and especially that of the U.S. and China – are deeply intertwined in ways unthinkable during the days of the standoff between the Warsaw Pact and NATO. Meanwhile, the Russian invasion of Ukraine has highlighted the uncertain future of Taiwan and western dependence on TSMC.

Coda

How might the U.S. respond in a positive way?  The short answer is we must out-compete – invest more and continue to expand an ecosystem that attracts and retains the best and brightest talent the planet can produce, both domestic (The Missing Millions) and international. I have sought to summarize some of these objectives in another series of essays.  As I wrote in Not Alone: Leaders in Focus, let’s dream big and bright about a larger, more diverse and inclusive science and engineering ecosystem that continues to be a magnet for global talent. (See For Science and Society, the Future Begins with Better Dreams.)

Holiday Explanations: So, What Do You Do?

Any young researcher, whether in academia or a government laboratory, who spends time with childhood friends and extended family over the holidays, will undoubtedly be asked, “What do you do?” Red Alert! Shields Up! Brace for Impact!

This is not an academic question, posed by a colleague and potential collaborator, where an extended, deeply technical answer is both expected and desired.   No, this is a potential social trap, one that has ensnared more than one naïve academic in its steely jaws. With the exception of your parents, and maybe your therapist, the putative interlocutor does not really want to know; they are simply feigning interest to stimulate social discourse.

The Social Conundrum

Yes, you are excited and passionate about what you are doing, and you should be.  After all, the scientific process may be the most successful intellectual endeavor in human history, demonstrably capable of illuminating the three big and enduring questions:  matter and the universe, life and its processes, and the human condition.  It has been and always will be about the age-old desire to explore and to understand, and in doing so, make life better for those around us.

Alas, you are fighting ingrained stereotypes and misinformation.  Everyone knows what a plumber does, but few people understand how science is really conducted.  Their models are usually the (sadly) boring experiences of primary and secondary school science, which often made after-school detention look attractive by comparison.  Class, what are the steps of the scientific method?  What is DNA? What causes it to snow?  These are important and interesting questions, but they are desiccated recitations, lacking all the joy of curiosity and the excitement of discovery that motivate every intellectual explorer, whether aged 8 or 50, who doggedly and persistently keeps asking “Why?” and “How?” (Editorial comment: We must stop ingraining scientific disdain and dislike in generations of students. See The Epistemology of Science and For Science and Society, the Future Begins with Better Dreams.)

Alternatively, the scientific understanding of your family and friends may be shaped by the Hollywood stereotypes perpetuated on television and the movies.  You know the ones I mean.  There’s the hacker tapping on the keyboard for a few seconds, then exclaiming, “I’m in.” (Brute force attacks on public key cryptography systems are exceedingly time consuming and difficult. Oh, and there are fascinating mathematical topics at the intersection of quantum computing and homomorphic encryption. Stand down; your spidey-sense tingling – the social trap is near.) Then there’s Spock or Scotty conjuring a scientific miracle to save the Enterprise before the next commercial break. Or, and this one beggars the imagination, the crime laboratory that does precise facial recognition using nine pixels from a VHS video tape recovered from a convenience store’s ancient, analog surveillance system.

You know the reality is far different. Science involves long hours of toil, filled with dead ends – experiments that do not work, which can be even more frustrating than failing, and elegant theories that are crushed by contradictory data – confounded with balky instruments, uncooperative specimens, and buggy code – and often fueled by hot coffee and cold pizza. Conversely, the reward is something transcendent, unraveling a mystery older than history and being the first person to ever understand the answer to a vexing scientific question.  At its best, science is child-like wonder as a lifelong passion.  Why wouldn’t you be excited? After all, you are doing Cool Stuff^TM that solves nature’s mysteries and might change the world! (See Science: It’s About the Wide-Eyed Wonder.)

Yes, you are fighting stereotypes as you sit at the holiday dinner table. Among them is the hoary cliché, one every scholar resignedly learns at some point in their career:

You learn more and more about less and less, until you know everything about nothing.  Then, we give you a Ph.D.

The positive interpretation speaks to the specialization and depth required to truly understand a topic and extend, even marginally, the boundaries of human knowledge. The explosive growth of knowledge means it is now impossible – and even laughable – for even the most gifted, determined, and learned to make any credible claim to know all that is worth knowing, even in a narrow academic field. These are the days of Renaissance teams, not Renaissance individuals. (See Renaissance Teams: Reifying the School at Athens.) You are an expert, but in a necessarily limited area. You learned a lot, but experience, wisdom, and humility also taught you that compared to what there is to know, you know almost nothing.

The negative interpretation – that you are filled with knowledge but lacking in practical “smarts” – is the one you seek to avoid at all costs. There’s an even more colorful, though less polite version of the everything about nothing aphorism that has one learning B.S., then M.S., and until finally, it’s Piled Higher and Deeper.

If you are still a graduate student, it can be even worse, when you may be asked “the question.”  It’s the one that bedevils every Ph.D. student, “When are you going to finish?” (See Doctoral Comedy: Which Way Is The Door?) Few understand that completing a Ph.D. is a qualitatively different process, involving open-ended, original research, not just completing additional classes. Consequently, the question stirs angst and ennui in even the most accomplished Ph.D. student.  Will I finish and if so, when?  Or, will I die homeless and alone, an ABD? (That’s All But Dissertation (ABD) for those of you playing the game at home.)

Explaining What You Do

Despite your well-justified passions, resist all temptation to open your laptop and walk your uncle through the brilliant presentation on early galaxy formation you just gave at the departmental seminar.  Do not fetch the 3-D prototype you just built (still in your backpack) and explain additive manufacturing and laser sintering to your cousin. Avoid explaining the innovative laboratory protocol you are using to screen for SARS-CoV-2 vaccine resistance to your mother – she’s already proud of you. 

Do not ask your nieces and nephews if they have an opinion on the P=NP problem, and if you foolishly do, thank your precocious eight-year-old niece who delightedly tells you that they are equal when N = 1. (Note to self: while the other adults are telling stories about their lives, make sure to talk to her about the joy of science and mathematics. Encouragement really matters!)

Finally, whatever you do, do not expound on Tomasulo’s algorithm and pipelined microprocessor design at a holiday party.  (I am personally guilty of the latter. Trust me; it’s rarely an effective ice breaker. If it is, you are in a safe place, with “your people.” See The Magic behind the Curtain: Hardware and Software for why I was (and still am) so excited.)

When asked what you do, pause, take a deep breath, and think. Few in your family really want to know all the details, and certainly not at the depth and breadth with which you are tempted to reply. This is your time to shine, by channeling your inner Mr. Wizard or Richard Feynman, by being clear, thoughtful, and memorable, but most of all, funny, self-deprecating, and brief.  (See Understanding and Explaining: The Feynman Technique.)

Keep your inner Sheldon Cooper in check, and whatever you do, do not engage in an argument, especially if some of your family members think COVID-19 vaccines were a government conspiracy, the Apollo moon landings were fake, or that humans and dinosaurs were temporally co-resident. Do not bring a knife (logical arguments) to a gunfight (emotional belief). Even if you win (and you won’t), you lose. (See Just the Facts, Ma’am: Reasoning is Not Dead, Jim.)

If asked by family members what I do, I have learned to check my propellor beanie at the door and focus on the potential impact of what I am doing – designing and building cheaper and faster computers -- and how those faster computers help create new medicines or enable better weather forecasts.  How research improves lives is something people generally understand and appreciate; the technical details are rarely important to them.  If they ask and clearly want to know, I will explain but try to do so in ways connected with their experience.

This is my most important piece of advice to young scientists and would-be science advocates.  Do not try to impress people with your erudition, and do not use obscure, though technically accurate nomenclature.  Speak clearly and in ways everyone can understand. It’s about the societal outcomes, not the personal quest for academic success or even expanding the human knowledge base. (See Simplifying Communication and Research, Education, Engagement and Bovine Transport.)

Above all, do not be flippant.  I learned this from painful experience. When asked if I were going to teach during the summer, I explained that I did not teach during the summertime because the federal government paid me to think. Though vacuously true, this response was met with stunned silence, especially after I unhelpfully added that if I had a good idea, I was supposed to write it down, omitting all discussion of research proposal submissions and peer review. The facial expressions alone confirmed my extended family member’s worst fears about my lack of common sense and wasteful government spending! (See Research Summers.)

Trust me, few people really want to hear an extended holiday discourse on exascale algorithmic efficiency, deep learning models, decidability and Zermelo-Fraenkel set theory, CRISPR DNA editing, or Lambda-CDM cosmology models – endlessly fascinating though they all are.  Even when enlivened with wild gesticulations and impromptu diagrams drawn in the mashed potatoes and gravy, in-depth science explanations are enough to make even your beloved grandmother look longingly at the holiday sherry, and they will confirm everyone’s unstated but deeply held belief that you are a “special child.”

Eschew the obscurity; chew the turkey. Stick to football if you can, opine on religion or politics if you must, but try to avoid the treacherous intellectual waters of scientific explanations over the holiday cranberry sauce and sliced turkey.  If they really want to know, they will pull you aside during the football game with specific questions. Only then should you unleash a small glimpse of your inner geek.

SenseCam Experiences: What I Really Do

As an academic researcher, what do I spend my time? Sometimes I wonder myself. To find out, while at RENCI, I once spent several weeks wearing a Microsoft SenseCam on a lanyard around my neck, and the University of North Carolina even wrote a story about my experiences.

A SenseCam takes periodic, still photographs, triggered either by elapsed time (roughly a minute) or by motion.  The resulting images then form a stop motion movie of the wearer’s day. The technology has also found use in assisting patients with memory loss, by replaying their experiences to strengthen memory pathways.

The Microsoft SenseCam was an early example of wearable computing, part of the Microsoft MyLifeBits life-logging project, long before smartphones with digital cameras were available. In turn, MyLifeBits was an attempt to realize Vannevar Bush’s 1940s vision of a Memex, which he described in a seminal essay, As We May Think:

Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and, to coin one at random, "memex" will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.

My friend and future colleague, Gordon Bell, then at Microsoft, popularized the SenseCam for personal recording of his life experiences. Gordon is, of course, best known as a parallel computing pioneer, creator of the eponymously named Gordon Bell Prize, and the father of the Digital Equipment Corporation VAX 11/780, one of the first superminicomputers.

The New Yorker has a great 2007 article, entitled Remember This? about Gordon and his digital knowledge experiences with MyLifeBits. About the same time, I wrote about the coming library revolution, with digital archives and portable devices as potentially transformative educational forces, something I still believe, particularly when coupled with adaptive artificial intelligence.

The first sight of me wearing a SenseCam (see the photograph of a younger me above) triggered expressions of shock and concern for my health, as my colleagues initially feared I had suffered a near fatal heart attack and had been outfitted with an external pacemaker. After assuaging those fears, I explained the SenseCam’s functions, and I offered to put it in my pocket if they were concerned about being recorded during our meeting.  Some were, but most were simply amused by yet another manifestation of my geek disposition. Years later, the response was the same when I donned both a Narrative Clip and a Google Glass.  (See Though a Google Glass, Darkly.)

For a few months, I wore the SenseCam literally all over the world, from Australia to many trips across the United States.  Without exception, it prompted questions about the technology and operation (most common) and about privacy and intended uses (second most common).

So, what did I learn from my SenseCam experience?  Perhaps unsurprisingly, my professional life largely consists of three things: drinking prodigious amounts of coffee, reading a deluge of email, and participating in meetings. In this regard, the broad outlines of my professional life – and that of most other academics – differed little from that of most office workers. However, betwixt these prosaic activities were the distinguishing attributes of my professional life, the all too brief periods of thinking and writing – research papers, conference presentations, and that bane of all researchers, project reports and funding proposals.

Coda: A Safe Harbor

There is, of course, one safe harbor, where being a geek is valued at the holidays. I speak, of course, about the assembling, installing, configuring, and/or testing of any electronic devices the family received as gifts.  Here, family members will temporarily put aside their skepticism of your life choices and dubious assessment of your practical skills if you can connect the Wi-Fi or produce a needed USB hub from your backpack. If they are wondering why you have – at a minimum – a spare USB hub, as well as a set of HDMI and DisplayPort adapters, USB keys of varying capacity, and a microphone headset, then they’ve never been a road warrior. Bask in your momentary glory, then ask for a second helping of pecan pie.

Working the Hill

Creating bills, negotiating amendments, and shepherding them to passage as U.S. laws is a messy process, with a cacophony of voices, each with differing perspectives, all seeking to be heard and appreciated.  To observers – and often to the participants – it can be slow, confusing, frustrating, and convoluted, but that is how participatory democracy works.  It is a collective fumbling toward majority views on what are almost always complex issues with many discordant voices and few, if any, unambiguously obvious solutions.

As one program examiner from the Office of Management and Budget (OMB) once noted to me, when I was chairing a review of the Networking and Information Technology Research and Development (NITRD) program as a member of the U.S. President’s Council of Advisors on Science and Technology (PCAST),

Dan, I think you know this, but I feel obligated to say it.  Rarely does anyone come to Washington and say, “I’m dumb and I have too much money.  Can you help me fix my problem?”

There are always more ideas than there is money to support them. Consequently, successful legislation is by necessity a compromise, a selective and sometimes partisan amalgam of ideas, priorities, and projects.  Rarely does any individual or group get everything they want, even when a bill becomes law on a party line vote.

Sadly, the notion that compromises are often key to progress is increasingly anathema to many, as social media echo chambers spread misinformation and harden emotional battle lines, posing increased risk to the political process. (See From Broadcast to Narrowcast: Finding Signal in the Noise.) Fact-based reasoning matters, especially as it relates to matters of science and engineering. (See Nature Cannot Be Fooled.)

 It seems increasingly old-fashioned, but I still believe thoughtful advocates can and will disagree on one issue while also finding common ground on another. Put another way, one can have respectful disagreements while not demonizing the other participant; demonization forecloses the possibility of future collaborations. Most of the time, most people are trying to do what they think is right; it is important that we cling tightly to belief in innate goodness, as a shared commitment to a better future.

Nor does the legislative process always yield the best outcome, as what constitutes “best” lies in the eye of the beholder.  This too is part of the great and ongoing experiment of collective debate and shared governance. Unanimity is rare, for as E. B. White once wrote, “Democracy is the recurrent suspicion that more than half of the people are right more than half the time.” Therein lies one of the bedrocks of democracy -- the consent of the governed; accepting the current will of the majority, even as one continues to disagree and respectfully advocate for reconsideration and change.

There are, of course, always exceptions to the notion of compromise.  There are occasions when any leader, whether elected or appointed, recognizes that a decision of moral and ethical consequence is at hand.  Are the public statements mere rhetorical devices and political slogans, or are they derived from deeply held, carefully researched, and wisely considered beliefs and principles.

One is sometimes called to do the right thing, even when the personal and political costs are high. At these times, leaders are tested; some rise to the occasion, and some do not. We all instinctively know true leadership when we see it.

In these cases, not only does the weight of history rest heavy on one’s shoulders, but that most fierce critic and constituent – the one who stares back at you in the mirror each morning – will judge you each day for the rest of your life.  That is why it is always important, but especially important in these circumstances, to think deeply and deliberatively, seek facts and context, and welcome perspectives from both those who agree and those who disagree.

CHIPS and Science

I was reminded of all of this as I watched the U.S. Congress struggle this summer to reconcile differing versions of bills that would enhance U.S. competitiveness by funding – among many other things – new semiconductor manufacturing facilities and semiconductor workforce development.  What began as the Endless Frontier Act, a political nod to Vannevar Bush’s famous memorandum, then became the U.S. Innovation and Competition Act (USICA), then the COMPETES Act, the Bipartisan Innovation Act, CHIPS-Plus and, ultimately, the CHIPS and Science Act.

The CHIPS and Science Act appropriated $52 billion in subsidies for semiconductor manufacturing and associated research and development.  The overwhelming majority of that money is to help the U.S. semiconductor industry “reshore” semiconductor fabrication at new and upgraded facilities.  I will editorialize just a bit here and note that the U.S. lost its semiconductor manufacturing dominance in part due to a series of business decisions, each of which made near-term sense, but were not consistent with long-term national interests.

As passed, the CHIPS and Science Act also authorizes major budget increases for the National Science Foundation (NSF), targeting expanded investments in basic research, growing and diversifying the Science, Technology, Engineering, and Mathematics (STEM) workforce, and creating the Technology, Innovation, and Partnerships (TIP) Directorate for technology transfer. (See Missing Talent, Missing Opportunities, and Losing Ground.) As the current chair of the Presidentially-appointed National Science Board (NSB), which oversees the NSF, I have been heavily involved in discussions surrounding what ultimately became the CHIPS and Science Act and the still evolving implementation of its provisions.

Despite the inevitable compromises, the CHIPS and Science Act is a huge win for the science and engineering enterprise.  As science fiction writer, Neal Stephenson, once noted, “If we want to create a better future, we need to start with better dreams.”  The CHIPS and Science Act is a big dream, and we owe a debt of gratitude to all who helped make it a reality.

Making the Sausage

In my varied academic roles over the years, and as leader of the Technology Policy Group (TSG) while at Microsoft, I have worked the Hill and testified to both the U.S. House and the Senate on a variety of issues, mostly related to high-performance computing and how it advances scientific discovery and economic competitiveness, but also on topics as varied as weather forecasting, clouds (the compute kind), digital privacy, wireless spectrum policy, and transnational data flows.  (See, NITRD: Come, Let Us Reason Together and It is Fall: There Must Be Clouds, for example.) These hearings have a certain rhythm, with prepared written statements, a brief oral presentation by each witness, then questions from the committee or subcommittee members.  

Hearings are part of the sausage making that is the political and legislative process, and as important as they are, hearings are just the public theatre that complement the many private discussions with industry, academia, think tanks, policy and lobbying groups, Congressional staffers, and members of Congress. These one-on-one and small group meetings, policy briefings, and commissions, along with protests (though rarely in scientific circles) are how constituents, and we in science and engineering are definitely constituents, make their voices heard.

Most of these meetings are with Congressional staff members, who take notes and collect messages for their principals. Bear in mind there are staff who work directly for members of Congress and those who work for Congressional committees, each of which has responsibility for some aspect of government. Such Congressional committee staffers are often long-time experts and serve as the Congress’ institutional memory.

For example, the House Committee on Science, Space, and Technology has exclusive jurisdiction over the National Science Foundation (NSF), the National Aeronautics and Space Administration (NASA), the National Institute of Standards and Technology (NIST), and the White House Office of Science and Technology Policy (OSTP).   However, although the House Committee on Science, Space, and Technology has jurisdiction over NSF, the agency’s appropriations are ultimately determined by the House and Senate Committees on Commerce, Justice, Science, and Related Agencies.

The fact that NSF is but one among several agencies addressed by these committees speaks to both its relatively small size (~$8.8 billion) and the competing priorities the committees must juggle (e.g., NOAA and weather forecasting, NASA, community policing, civil rights, and economic development).  The 302(b) mark given to each subcommittee, in turn derived from the 302(a) mark given to the House and Senate Appropriations Committees, determines the total amount of money – the spending cap – available for allocation.  Hence, there are always difficult choices to be made.

In between all this, lies the U.S. President’s proposed budget, which outlines the administration’s priorities. It is developed by the Office of Management and Budget, working with the executive branch agencies.  These priorities may – or may not – be those of Congress, particularly if different parties control the White House and the Congress.  That’s why an old Washington aphorism applies, “The President proposes and Congress disposes.”  Congress gets the final word on both budget authorizations and appropriations.

NSF and the CHIPS and Science Act

Distinguishing budget authorizations (“you can and/or you shall”), as defined from appropriations (“take this money and do this”) is important.  You cannot spend an authorization; it and a few dollars will buy you a cup of coffee.

There is a long history of increased authorizations for NSF, including its proposed budget doubling in 2007. Alas, the poetry of authorization yearnings has rarely translated into the appropriations prose of major budget increases. Research discoveries satisfy our unrequited curiosity, what astronomer Edwin Hubble called “the urge older than history,” but all too often, transient exigencies – wars, disaster remediation, pandemic response – and the realities of discretionary spending limits have precluded major budget increases for NSF.

One experienced Washington hand once explained Federal budget politics by noting that the U.S. government is a social service agency (Medicare, Medicare, Social Security, and other services) with a police force (national defense) and some hobbies.  Science is one of the hobbies – a really important one, but, like hobbies, it is discretionary.

All of which is why it is important to make the case for appropriations in person and with “leave behind” evidence and talking points. This fall, I did just that, spending two days on Capitol Hill, meeting with Representatives, Senators, and staff members about the importance of increased investment in basic research, expanding and diversifying the STEM workforce – the Missing Millions of the NSB’s Vision 2030 – and funding the TIP Directorate’s Regional Innovation Engines (RIEs), which will broaden the geography of innovation by creating technology transfer engines across the United States.

Specifically, we asked Congress to

• Fund the National Science Foundation’s (NSF) FY23 budget at $11.89 billion, as authorized by the CHIPS and Science Act, and concomitantly
• Support the RIEs, lest the opportunity to broaden the geography of innovation be lost
• Embrace the rising global competitiveness challenges, as noted in the CHIPS and Science Act, and support growth of the STEM workforce.

It was a start.  There is more work to be done – more visits, more conversations, and more evidence.

Coda

It was great to be back, seeing the hustle and bustle, the young staffers, the excited interns, the wise and experienced veterans, the earnest conversations, the clock ticking down on votes, even the concerned protesters exercising their rights.  It’s democracy in action.  I’ve missed seeing it.

Before 9/11 and certainly before January 6^th, one could enter the U.S. Capitol and the U.S. House and Senate buildings with little fanfare, as everyone who visited was expected to exhibit a certain level of decorum and respect.  It is a much more complicated process today, with more security and in many cases, an escort required on entry.  However, once inside the buildings – Cannon, Longworth, and Rayburn (House) and Russell, Dirksen, and Hart (Senate) and even the U.S. Capitol – one is largely left to navigate on their own.  It is not at all uncommon to pass unaccompanied Representatives and Senators in the hallways. After all, this was and is the heart of American democracy.

At the end of two days of Hill visits, my compatriots and I retired to Bullfeathers, a bar in the Capitol Hill neighborhood, to assess and recap our Hill discussions.  The irony that I had once been trapped there on 9/11 for several hours was not lost on me. (See I Could Smell the Smoke.)

For Science and Society, The Future Begins with Better Dreams

Working with National Science Board (NSB) staff, I recently authored a post for Rafael Bras' new discussion group, Not Alone: Leaders in Focus. As the tagline says, it is a place for "unfiltered perspectives on global issues by research and academic leaders."  You can read my recent post, For Science and Society, the Future Begins with Better Dreams."

In addition, I co-authored an a science funding advocacy op-ed in The Hill with Dario Gil (SVP for Research at IBM).

Cyberinfrastructure: Lots of Room at the Edge

Twenty years ago, while director of the National Center for Supercomputing Applications (NCSA) at the University of Illinois, I gave a series of presentations on the future of cyberinfrastructure.  Looking through my presentation archives, I found this diagram, which emphasized building up (i.e., larger and more capable computing systems and scientific instruments) and building out (i.e., ubiquitous access to computing systems and instruments of all sizes, including distributed sensor/actuator networks).

At the time, we did not understand that fungal infections from chytridiomycosis were in part responsible for widespread death and extinction among amphibian species. I suggested with some seriousness that we instrument selected biological environments with MEMS smart dust and track flora and fauna movements in situ. Or, as I put it at the time, “Why not give every frog an IP address?” Of course, understanding ecosystem evolution eventually became the raison d'etre for the for the National Ecological Observatory Network (NEON) – more on that later. I also had several conversations with then National Science Foundation (NSF) Director, Rita Colwell, about the prospects for HPC-enabled in silico computational modeling of biological processes, something only now becoming possible. (See In Silico Whole Cell Modeling.)

With the perspective afforded by twenty years, how have we fared in realizing this integrated vision?  Building up, we have done well in most contexts, but we now face the twin realities of escalating costs and technology challenges at very large scales, both for computing platforms and for scientific instruments. It is unclear how many multibillion-dollar instruments or computing platforms we have either the political or economic will to construct. As for building out, it is a mixed story of broader access to midrange commodity Linux clusters and new instruments, but we are far from having created a ubiquitous scientific infosphere.

Building Up

In 2002, the first terascale computing systems had just become operational, including the NSF TeraGrid, which connected NCSA, SDSC, Argonne National Laboratory, and Caltech via a then extraordinarily fast 40 Gb/s transcontinental network. (See NCSA@30: The Revolution Continues.) At the time, we were still dreaming of petascale data storage systems, and research discussions about the path to petascale computing were well underway. The 1996 and 1999 Caltech Petaflops workshops at Bodega Bay and Santa Barbara were seminal in shaping the future that because petascale computing.  As this photograph shows, we have all aged a bit in the intervening twenty five years, and sadly, a few of our colleagues are no longer with us.

Today, terascale and petascale computing systems via GPU accelerators are widely available, spanning desktops to laboratory and university cluster environments. In May 2022, the latest incarnation of the TOP500 list of (some of) the world’s fastest computers was revealed at the ISC High Performance conference in Hamburg, Germany. There was much hype and chest thumping as the Oak Ridge National Laboratory (ORNL) Frontier system broke the exascale barrier based on the high-performance LINPACK (HPL) benchmark for dense matrix factorization.

I say “some of” the world’s fastest computers because submission to the TOP500 list is entirely voluntary, and some organizations chose not to submit entries, for a variety of reasons, some political and economic, some technical, and some related to national security. In particular, China has at least two exascale systems (OceanLight and Tianhe-3), but has not submitted either of them to the TOP500 list, likely to avoid further inflaming tensions with the United States.

Meanwhile, the very definition of exascale is itself now subject to debate, both because the dense linear algebra of the HPL benchmark is widely acknowledged as no longer a good predictor of multidisciplinary application performance and because mixed precision arithmetic is now common in many applications, particularly those with AI-accelerated components. When convolved with system and chip energy constraints, semiconductor fabrication challenges, and ecosystem shifts, a sense of quo vadis now pervades the field. (In a recent essay and associated article, my colleagues Jack Dongarra, Dennis Gannon, and I offered a few thoughts on strategic directions.)

In the world of computing, the locus of innovation has focused on the small and many (think smartphones and the Internet of Things) and the few and large (think commercial cloud infrastructure and leading edge HPC systems); scientific instruments are dominated by a similar dichotomy, though the large and few receives most of the attention and visibility.  As for large-scale scientific instruments, one need look no further than the aging Hubble Space Telescope, the recently upgraded Large Hadron Collider (LHC), the International Thermonuclear Experimental Reactor (ITER) being built by a global consortium, the soon to be operational Vera Rubin Observatory (nee Large Synoptic Survey Telescope) and James Webb Space Telescope (JWST), or the three proposed and under construction Extremely Large Telescopes. Biology has joined the physical sciences in large-scale instrumentation, spanning the Human Genome Project to the National Ecological Observatory Network (NEON).

Like leading edge supercomputers, these big instruments are the shiny objects featured in press releases and popular science stories, and well they should be, for they produce scientific results achievable via no other mechanism. Each involves hundreds, thousands, and sometimes tens of thousands of scientists and engineers; each requires billions of dollars, euros, or yuan to design, build, and operate; and produces torrents of data that demand petascale data archives and global content distribution networks. Thanks to these new instruments, the associated explosion of scientific data has been hugely democratizing, allowing many researchers to focus on the key intellectual questions rather than the mechanics of data acquisition. 

However, this brave new world and the shift from data paucity to plethora have also raised new questions about data preservation and triage, shared and individual responsibilities, FAIR access, and the costs of data retention, as mandated by government sponsors.  (See Research Data Sustainability and Access and My Big Scientific Data Are Lonely.) As the late Nobel Laureate Herbert Simon once noted,

What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention, and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.         

In my humble opinion, we have not yet fully internalized the social, political, economic, and technical implications of the shift from data scarcity to data ubiquity.  What data we keep and for how long, how we manage access, and how we distribute the very real costs of data management and preservation have not yet been fully resolved.  Serious science policy issues remain unresolved.

On Being the Right Size

Although our intellectual curiosity is rightfully unbounded, our financial resources are not. What are the right sizes and relative numbers for our computing platforms and our scientific instruments?  Is it all about faster and bigger, which in almost any realistic budget scenario, also means fewer?  In the limiting case, as illustrated by ITER and the LHC, fewer may well mean only one per planet, financially sustained only via a global scientific partnership.

Nor does the fewer and bigger choice only affect the high end, it squeezes investment in the long tail of campus and laboratory infrastructure as well. These are important science policy questions, especially when one realizes that the day-to-day, “bread and butter” work of science is conducted with many instruments of much smaller size, educating students in scientific processes.  (See On Being the Right Size: Science Scaling and Power Laws.)

Make no mistake; form and scale absolutely follow function, as the British geneticist J. B. S. Haldane first wrote in a now famous essay entitled On Being the Right Size. (You can read the original essay as originally published in Harper’s Magazine, skipping the political pontification at the end.)  In it, Haldane offered cogent arguments about surface area to volume ratios, structures, respiration and energy, noting:

The most obvious differences between different animals are differences of size … it is easy to show that a hare could not be as large as a hippopotamus, or a whale as small as a herring. For every type of animal, there is a most convenient size, and a large change in size inevitably carries with it a change of form.

Computing systems and scientific instruments are no different. A simple refractor telescope design cannot be expanded unchanged to create a 30-meter optical observatory; the materials properties do not scale linearly, and the control systems differ markedly. A child’s horseshoe magnet cannot be scaled to create a 100-tesla pulsed research magnet. Nor is an exascale computer just a big deskside cluster, despite some comparable elements. The lack of self-similarity at scale means new design approaches are needed in different operating regimes, each with associated design costs.

Building Out

Returning to my twenty-year-old diagram, we have made progress in building out, with midrange computing clusters now more common and new classes of midrange instruments (e.g., cryo-EM) changing the face of science. Cyberinfrastructure training materials are now widespread, and we even host student cluster construction competitions; building a Linux cluster is no longer an arcane art. Indeed, one can now build an inexpensive Raspberry Pi cluster with almost all the same software tools found on leading edge supercomputers.

Despite this progress, our intellectual fascination with a small number of large instruments and computing platforms often blinds us to the power of distributed discovery and innovation with large numbers of instruments at smaller scales. Despite this, we have made enormous progress in sensor miniaturization, perhaps even more than Feynman envisioned in his seminal, “There’s Plenty of Room at the Bottom” nanotechnology lecture. (See Come to the Supercomputing Big Data Edge.) In this sense, we have a long way to go to realize the vision of a ubiquitous infosphere. Let me illustrate with a specific example.

SAGE: A Software Defined Sensor Network

Over the past two years, I have greatly enjoyed playing a small part in the NSF-funded SAGE project, which is seeking to build a flexible and extensible hardware and software cyberinfrastructure for AI at the edge.  Led by Pete Beckman, SAGE builds on many ideas from Charlie Catlett’s urban science and environmental science Array of Things (AoT) project, which deployed sensors on utility poles across the City of Chicago and surrounding areas.

AoT engaged community residents in defining acceptable data use policies, explored data driven social science by direct measurement, and stimulated citizen science and K-12 hands-on education. For a review of the AoT project and the motivating lessons for SAGE, I recommend this IEEE Xplore article, which several of us wrote recently.

SAGE is based on the notion of software-defined, configurable sensors that include edge computing, standard sensor interfaces, a set of baseline sensors, and extension interfaces for environment-specific sensors, along with backend data storage and analysis facilities.

A standard “wild SAGE” node (named for its hardened packaging and fault tolerant design to run unattended in harsh conditions, or “in the wild”)  includes the following:

• One or two NVIDIA Xavier NX GPUs with Wi-Fi, Ethernet, and a full Linux stack
• Power over Ethernet (POE) sky and ground facing cameras
• An optical rain sensor and a microphone
• Relative humidity, barometric pressure, and temperature sensors in a Stevenson shield (that’s the box on the left in the photograph, which also contains a Raspberry Pi to manage the Bosch BME 680 sensor)
• POE and USB connectors for additional, domain-specific sensors

Why deploy AI at the sensor edge?  First, the volume of streaming data may preclude storage, either due to bandwidth limitations or storage capacity.  For these same reasons, the LHC’s detectors rely on real-time triggers detect and record only a small subset of particle collisions. Second, urban and environmental monitoring bring strong security and privacy concerns, particularly involving video. By extracting events and features entirely at the edge, the raw video need never leave the sensor, obviating many privacy concerns (e.g., the timing and frequency of pedestrian interactions may suffice, without saving any identifying video). Third, AI at the edge means the sensor can adapt dynamically to changing conditions (e.g., increasing measurement frequency based on a detected event or even triggering alternative measurements at other locations). Fourth, latency constraints may demand a response time possible only when decisions are made at or near the edge.

In SAGE, the key to this flexibility is an end-to-end architecture – Waggle AI@Edge – that allows domain scientists to focus on their science, rather than building generic infrastructure.  One need only connect new sensors, develop domain-specific AI plugins using widely available AI tools (e.g., OpenCV, PyTorch, and TensorFlowLite) to process sensor data in situ on SAGE nodes, and use tools such as Jupyter notebooks to analyze and correlate data.

To demonstrate this integration capability, the SAGE project has been working with NEON to integrate inexpensive SAGE nodes for augmented environmental monitoring and with California and Oregon wildfire researchers to detect and respond to the growing prevalence of wildfires. There are now SAGE plugins for vehicular traffic detection, bird song classification, cloud cover, smoke, and other flora and fauna events. 

At Utah, we have deployed one SAGE node at the Natural History Museum of Utah (NHMU), which allows us to track environmental conditions and local flora and fauna.  Additional SAGE nodes are being deployed in downtown Salt Lake City to monitor PM 2.5 particulate air pollution and to study advanced wireless technologies in collaboration with the NSF POWDER project.

Much More Room at the Bottom

At ~$15,000 each, SAGE sensors are inexpensive compared to multibillion dollar mega-instruments.  Yet even this price is prohibitive for many research contexts, limiting their deployment in science and community education projects where resources are even more bounded and precious.  Fortunately, there is much more room at the bottom.  In this spirit, members of the SAGE education team are working on lower cost SAGE versions based on the Jetson Nano and the Raspberry Pi.

However, imagine a cyberinfrastructure world containing even lower cost sensors -- tens of thousands of them at ~$50 each! Leveraging a wealth of inexpensive and open source hardware and software, it is now possible to create and deploy a powerful sensor infrastructure for just a few hundred dollars.

Let’s start with server infrastructure. A Raspberry Pi with a complete Linux software stack, including Docker and Kubernetes virtualization, a high-capacity SD card for system software and data storage, Wi-Fi, Ethernet, and LoRaWAN long-distance communication transceiver can be assembled for approximately $100.

Such a system can readily serve as a base station for hundreds of microcontroller nodes, each supporting a wide variety of ~$5 sensors. As an example, one of my former colleagues at Iowa fabricated the sensor node shown at right based on my specifications. This node includes

• Powerful ESP32 Tensilica microcontroller with Wi-Fi and Bluetooth LE
• Bosch BME 280 pressure, temperature, and humidity sensor
• LoRaWAN low bandwidth, long distance communication
• I²C, analog, and digital Grove, solderless sensor connectors
• Battery and/or solar power interface for long-term episodic operation

Such sensor nodes can be readily designed using open source hardware tools such as KiCad and fabricated in quantity for less than $100 each, or similar ones can be purchased commercially (e.g., Heltec Wi-Fi LoRa32 V2 or Adifruit Huzzah32 Feather with Feather Wing addon sensors).  Equally importantly, they and the Raspberry Pi base stations can be integrated directly with environments such as SAGE. Just as commodity clusters opened computing to a wide range of practicing scientists, inexpensive sensors can be the missing and catalyzing element to realize the vision of an immersive sensor/actuator infrastructure.

Over the past two years, as a research proof-of-concept, I have developed a generic software infrastructure for Raspberry Pis and ESP32 sensors, while Charlie Catlett has been exploring complementary hardware and software models based on inexpensive home automation tools.  We are now drafting a “how to” book, targeting K-12 STEM education and citizen science.

Cyberinfrastructure Futures: The Missing Elements

We have come far in building out a 21^st century cyberinfrastructure, but there is more to do if we are to fully exploit the opportunities now afforded by scale – both small numbers of large computing systems and instruments and large numbers of small computing systems and instruments. As we imagine a future world of software defined sensors, I believe we must address the following research challenges, and I encourage all of us think about how we might mount national research programs to explore each of them.

An Extensible Cyberinfrastructure Baseline. At present, we lack a generic sensor toolkit that is broadly applicable to diverse measurement environments.  The SAGE project is an early attempt at designing and deploying such a generic infrastructure, drawing on lessons from the earlier Array of Things (AoT) project, but much more experimentation and targeted development is required to separate systemic needs from domain-specific science.  Even more importantly, the lifetime of these longitudinal sensor data surveys may last for years or even decades, placing an even greater premium on organizational and infrastructure stability.

Intentional Programming at Scale.   Task or node level specification is central to most of our programming models, yet so many of the questions that matter with distributed sensors are systemic ones involving identification of either data outliers (e.g., vibrations on this bridge exceed norms) or broad trends (e.g., rainfall along the severe storm front is approaching critical levels).  Equally importantly, such data should trigger certain global actions (e.g., if viral loads in sewage rise, increase global sampling rates). These global, “if this happens, then do this” intentional specifications, with some tolerance for sampling errors and uncertainty, are not typical of our programming models, but are ideal for specifying system behaviors and responses. We need new research on systemic, intentional programming models and specification systems (e.g., based on AI qualitative reasoning and zero shot planners).

Run Forever Heterogeneity. Those of us in computing are intimately acquainted with forklift hardware replacement, when old hardware gives way to new systems.  Depending on facility capacity, there may be some period of system overlap, but inevitably the forklifts remove the old hardware, leaving only a fully integrated, new system in their wake. Alas, the real world of distributed sensors bears much more resemblance to that of embedded real-time systems, where replacing sensors can be both costly and impractical.  Sometimes this is because sensors are deployed in critical infrastructure with hard real-time constraints and uptime quality of service expectations. In other cases, the cost of replacement may simply be prohibitive. 

For example, replacing the Chicago AoT nodes required access to City of Chicago utility bucket trucks, something possible only rarely and as a low priority relative to operational utility needs.  Similarly, replacing sensors in remote field sites may require extensive and difficult travel.  One consequence of such difficulties is extensive sensor heterogeneity, with new sensors and technology operating in concert with older, less capable sensors.  This has profound implications for both programming models and software configuration and deployment, as well as the types and resolution of data being captured and kinds of in situ analysis possible. 

Bags of Interdisciplinary Data. In addition to disciplinary access, much of the value of longitudinal sensor data accrues from cross-disciplinary data fusion, asking questions never envisioned by any of the original collectors of the data. The latter necessarily requires integration across diverse data schemas, none of which were designed with cross-disciplinary questions in mind.  We need to embrace data schema heterogeneity as an inevitability rather than as an impediment, encouraging development of new techniques that support “come as you are” data analysis.  None of the search firm giants have a master schema for web pages; rather, they index data and build search engines capable of responding to a wide range of queries.  We need to draw insights and lessons from there experiences.

A Data Economy. Finally, government mandates for data preservation and access are creating new challenges for researchers and research institutions alike.  What should be saved and for how long?  Can such answers be determined solely by the constituent disciplines or should there be a combination of discipline-specific and university specifications?  When can and should data be discarded?  Are those choices based solely on usage frequency, uniqueness, or cost of replication relative to cost of retention?  Research librarians long ago adopted deaccession policies, triaging holdings and negotiating interlibrary loan policies for rare or unique artifacts?  We need a coherent data economy that balances data utility against costs, driven by scientific needs.

Coda

We have come far in developing a coherent national and international cyberinfrastructure for research and innovation.  However, there are still unmet needs and unresolved infrastructure and research questions surrounding distributed instruments and large numbers of inexpensive sensors.  It is time to reason together and build the missing pieces.

A {Racing} Bicycle for the Mind: Helmet Recommended

The March 1973 issue of Scientific American contained an article entitled simply “Bicycle Technology,” written by S. S. Wilson.  As the title suggests, the article recounted the technological evolution of the bicycle and how those insights contributed to the development of early automobiles and airplanes.  After all, many of the mechanical techniques used by the Wright brothers in the Wright Flyer, the first heavier-than-air aircraft, were developed in their Dayton, Ohio bicycle shop and tested at windswept Kitty Hawk, North Carolina.

This Scientific American article would now be of only passing historical interest were it not for a single diagram (shown below), which plots the cost of transport (in calories per gram per kilometer) against body weight (in kilograms) for several organisms and devices, all on a log-log scale. As one would expect, flying organisms tend to be small in weight but expend energy at high levels to remain airborne. Land-based mammals, though much larger, consume less energy relative to their weight. Among the latter, bipedal humans are unremarkable, more efficient than dogs but less efficient than horses.  Similarly, helicopters are less efficient than airplanes or automobiles.  All of this matches our intuition and experience.

Ah, but here’s the story.  A human on a bicycle ranks first in energy efficiency among mobile animals and machines in energy consumed to move a certain distance as a function of body weight, and not by a little, but a lot! The chart shows that the rate of a bicyclist’s energy consumption is about one fifth that of a walking human. Anyone who has bicycled knows this is true.  A reasonably fit bicyclist can cover more than one hundred miles in a day, the so-called century ride.

By Geeks for Geeks

The fact that a human on a bicycle is quantitatively and qualitatively more efficient than anything nature has engineered via evolution caught the eye of an eager young reader, Steve Jobs.  More on that in just a minute.  First, let’s talk about the magazine, Scientific American.

Founded in 1845, like many magazines, Scientific American had its ups and downs until three partners reinvented it during the 1960s, a few years before I first encountered it.  For at least the next thirty years, it embodied everything the well-read – wait for it – scientific American might hope to know about cutting edge science and technology.   Unlike other popular science magazines, the articles were written by the scientists responsible for the discoveries they described, which gave them gravitas often lacking elsewhere. The Amateur Scientist column by Jearl Walker described experiments that could be conducted by savvy amateurs, and the Mathematical Games column by Martin Gardner enchanted many a budding mathematician and computer scientist.  

Scientific American was the kind of magazine geeks awaited eagerly each month, then stood by the mailbox reading cover-to-cover with undisguised joy, only to rush inside afterward to regale unsuspecting family members with their latest insights. I knew this firsthand, having had my own transformative experiences with the magazine, beginning with a stash of old copies I was given as a young teenager. There, I read about stellar nucleosynthesis of the elements, progress toward the Standard Model integrating three of the four fundamental forces, and new insights into DNA and gene actions, among many others. I have since been a loyal subscriber for over forty years, though I hope the magazine has been as successful in recruiting younger readers as it was with me.

I read many fascinating things in Scientific American over the years, but two things stand out most vividly. The first was Christopher Zeeman’s delightful exegesis on catastrophe theory, a mathematical model of non-linearity in complex systems, something that I wrote about recently in the context of COVID-19. (See On Catastrophes and Rebooting the Planet.) The second was Martin Gardner’s description of John Conway’s Game of Life, a beguilingly simple but powerful cellular automaton.  Sadly, John Conway, who spent many years at the Institute for Advanced Study, passed away during the COVID-19 pandemic.

The Game of Life became a staple in computer science circles, engaging both experimenters and theorists who derived self-replicating entities and ultimately proved the Game of Life is Turing complete. That rabbit hole is very deep, for via its three simple rules, the Game of Life is capable of computing all that is mechanically computable, bounded only by Gödel incompleteness and the Entscheidungsproblem. From there, it is just a hop-skip-and jump to the Chomsky hierarchy and Turing’s musings on what it means to think.

As an undergraduate, I burned through prodigious amounts of IBM S/370 computer time myself, running Game of Life simulations, enough so that I got called on the carpet by the computing center leadership.  Being fellow geeks themselves, after I showed them the code and explained what I was doing, I was given a friendly blessing and sent on my way. (See A Feeling for the Code.) Since then, I have used the Game of Life as an example in more than one parallel computing class, illustrating parallelism, irregular workloads, and synchronization constraints.

Amplifying Human Talent

Given the impact of Scientific American on the science and technology crowd, it is no surprise that a young Steve Jobs saw the article on bicycle technology and how it augmented human capabilities. It was an idea that became a passion, finding the intellectual analog of the bicycle, something that could augment the human mind in the same dramatic way a bicycle amplified locomotion. You can watch Jobs talk about it here and here, at two different periods in his life. His memory about the condor as the most efficient was incorrect; it was not on the list. However, he captured the essence of the story, when he talked about humans as tool builders and a computer being a bicycle for the mind, able to take us far beyond our human abilities. As Jobs put it,

… somebody at Scientific American had the insight to test the efficiency of locomotion for a man on a bicycle. And, a man on a bicycle, a human on a bicycle, blew the condor away, completely off the top of the charts.

And that’s what a computer is to me. What a computer is to me is it’s the most remarkable tool that we’ve ever come up with, and it’s the equivalent of a bicycle for our minds.

Jobs tried to name the then internally code-named Macintosh project the “Bicycle,” but it didn’t stick. Apple even used the bicycle for the mind analogy in its marketing, as this 1980 Wall Street Journal advertisement shows. (Alas, no copyeditors verified the condor allusion.)

A Universal Intellectual Amplifier

Over the years, I have crafted my own version of this analogy, emphasizing that computing is distinct from all other human tools in its universality. (See Intellectual Amplification via Computing.) Physical tools, from the simple inclined plane through a pulley to a screwdriver, rely on mechanical advantage, each specialized for a particular task. Put another way, a screwdriver will not help you lift a weight, nor will a pulley help you fasten two objects together. Each is useful, but each is specialized.

That specialization extends to scientific and intellectual endeavors as well.  Telescopes, such as the Vera Rubin Observatory or James Webb Space Telescope, expand our vision, reaching far beyond the electromagnetic range of visible light and the limits of the human retina. They let us see the faint, large, and distant, but do nothing to illuminate the spatiotemporal dynamics of a biochemical reaction. Nor can a cryo-EM’s amazing efficacy in understanding macromolecular structures measure the cosmic microwave background (CMB) or plumb the mysteries of dark matter and dark energy.

In contrast, computing is universally applicable to diverse human intellectual efforts. That bicycle for the mind can help write sonnets (or via AI, write them itself), explore the spread of viral disease, understand galactic evolution, or a thousand other things, limited only by human imagination.

In that spirit, my friend and former colleague, Fred Brooks, one of the fathers of IBM’s System/360 family of computers, once described himself as being in the “intelligence amplification” business. After leaving IBM and moving to the University of North Carolina at Chapel Hill, he met with a University of North Carolina at Chapel administrator, whom he then wryly asked who on campus most needed their intelligence amplified.  Beyond the twinkle in Fred’s eyes as he told the story, I’ve always loved the anecdote, because it speaks to the human nature of computing as a collaborative enabler for creativity.

The Supercomputing Racing Bike

If the personal computer – Mac or PC – is a bicycle for the mind, then a supercomputer is a carbon fiber composite racing bicycle with electronic shifters, hydraulic disc brakes, and tires engineered to minimize rolling resistance.  Though the principles are the same, a supercomputing bicycle is far different than the bike you rode as a child.

Back in May 2004, I testified to the U.S. Congress about the needs of racing bicycles for the mind, something I have done several times. At the time, I was serving as the Founding Director of the Renaissance Computing Institute (RENCI), itself based on the idea that computing was universally applicable to enriching the human experience.  It was an idea I learned from my NCSA colleague and friend, Donna Cox, an artist who has enriched the lives of so many via her Renaissance teams and scientific visualizations.

The other witnesses that day were Jack Marburger, then Director of the White House Office of Science and Technology Policy (OSTP); Irving Wladawsky-Berger, a Senior Vice President from IBM; and Rick Stevens from Argonne National Laboratory.  You can read the entire transcript of the hearing here. At that hearing of the House Science Committee, I offered these words about the universality of high-performance computing and its power to expand the frontiers of human knowledge, noting

The breadth of these examples highlights a unique aspect of high-performance computing that distinguishes it from other scientific instruments – its universality as an intellectual amplifier.  Powerful new telescopes advance astronomy, but not materials science.  Powerful new particle accelerators advance high energy physics, but not genetics.  In contrast, high-performance computing advances all of science and engineering, because all disciplines benefit from high-resolution model predictions, theoretical validations and experimental data analysis. As new scientific discoveries increasingly lie at the interstices of traditional disciplines, high-performance computing is the research integration enabler.

I went on to say something that I believed then but seems even more true now

Although this universality is the intellectual cornerstone of high-performance computing, it is also its political weakness.  Because all research domains benefit from high-performance computing, but none is solely defined by it, high-performance computing lacks the cohesive, well-organized scientific community of advocates found in other disciplines. In turn, this has led to over-dependence on market forces to shape the design and development of high-performance computing systems, to our current detriment.

In almost twenty years since, we have seen phenomenal computing advances.  We have broken the petascale and the exascale performance barriers.  Advanced AI delivers modern day miracles, capable of playing chess and Go at levels that humble even the world’s best human players, able to translate human languages with high and rapidly improving accuracy, adept at predicting how proteins fold, and able to compose prose with near human facility.

We have come far on the racing bicycle for the mind.  We need to keep pedaling, for we still have far to go to reach our destination, given both the manifest opportunities and the equally daunting challenges. 

I have written about the challenges of building ever faster computers, given economic and technical constraints, and I have called for new approaches to designing and building advanced computers that lessen our dependence on traditional market forces, recognizing that advanced computers are increasingly bespoke research instruments rather than commercial commodities. (See American Competitiveness: IT and HPC Futures – Follow the Money and the Talent and a related article.)

We also have deep technical and social issues surrounding AI ethics, digital equality, and digital communications. We are very far from the general AI (strong AI) so omnipresent in the movies, yet as our machines take on more tasks once viewed only as the province of humankind, we need to think about roles and responsibilities, and how we best imbue our machines with the best of humanity’s values. (See Reluctant Revolutionaries, the Trolley Paradox, and Ender’s Game.)

In addition to the handlebar streamers and the thrill of the wind in our hair, a helmet might be advised.

Coda

Make no mistake, Steve Jobs was absolutely right; the computer is arguably one of the most remarkable tools we humans have ever invented.

I have been proud and humbled to be a racing bicycle for the mind designer for these past forty years. Pedal on.

From Broadcast to Narrowcast: Finding Signal in the Noise

The arc of the 20^th century is interesting for many reasons, not the least of which is how it was shaped by dramatic and traumatic events, including the Great Depression and two world wars. However, the arc of its communication mechanisms is equally striking, though less discussed outside computing and business circles.  It began with Marconi and “wireless,” transitioned through commercial radio and broadcast television, and ended with the Internet and streaming media.

For millennials and those younger, it may now seem almost inconceivable, but there was a time when almost everyone in the United States heard the news from the same three television channels, watched the same television shows, and listened to the same music on a popular local radio station. Most cities had both morning and evening daily newspapers, filled with both the minutia of daily life – crimes, weddings, and funerals – with a smattering of national and international news, and local investigative journalism by the fourth estate, shining a bit of light into the darkness.  These were shared cultural experiences, woven throughout the fabric of American society.

All that has changed, some for the better, and arguably, some for the worse.

The Golden Age of Broadcast Television

In the first half of the 20^th century, radio was the primary mechanism for disseminating entertainment and news. Beginning in the 1950s, analog (NTSC) broadcast television on the VHF and UHF bands – the big three (CBS, NBC, and ABC) -- began dominating the airwaves, with PBS as a public television latecomer.  Led by CBS News and Walter Cronkite, the nightly news broadcasts from the big three defined the national conversation on U.S. and global events, albeit mostly from a white, middle American perspective.

Unlike today’s extended social commentaries on 24x7 cable news channels, the nightly news was largely delivered in a staccato “just the facts, ma’am” format.  As Cronkite put it at the end of each broadcast, “And that's the way it is.” Although the broadcast standards supported color, the high cost of color television meant broadcasts were mostly in black-and-white until the 1960s.

Nor were there live standups from international locations or from across the studio, rather international reports were often recorded on film, then shipped to news bureaus for subsequent development and editing; remote satellite broadcasts were yet to come. Breaking news interrupted normal network programming, as this iconic photograph of Walter Cronkite informing the nation of President Kennedy’s assassination shows. 

At thirty minutes, including commercials, the network news reports also offered little time for pontification and social commentary. That made Cronkite’s February 27, 1968 special report on the Vietnam War all the more telling, when after returning from Vietnam following the Tet Offensive, he concluded,

But it is increasingly clear to this reporter that the only rational way out then will be to negotiate, not as victors, but as an honorable people who lived up to their pledge to defend democracy, and did the best they could.

Such was the trust in Uncle Walter that unsubstantiated stories have long said President Johnson then remarked, “If I’ve lost Cronkite, I’ve lost Middle America.”

However, I did not see that, because for most of my grade school years, our aging black-and-white console TV worked only sporadically. When it did, to change channels, my father had to climb on the roof and manually rotate the antenna, while my mother leaned out the window to give him updates on the TV signal quality.  Needless to say, we picked a channel and stuck with it, in our case, the Springfield, Missouri NBC affiliate (KYTV), which aired the Huntley-Brinkley Report.

Consequently, I distinctly remember Huntley and Brinkley making their own antiwar statement in 1969 by playing Kenny Rogers’ version of Ruby, Don’t Take Your Love to Town. While Rogers sang haunting words about a “crazy Asian war,” a camera panned through various footage, highlighting the sacrifices of U.S. draftees and their families.  Huntley and Brinkley then ended the news with their distinctive signoff, “Good night, Chet. Good night, David. And good night, for NBC News.”  (Huntley broadcast from New York and Brinkley from Washington, D.C.)

When I was in the fifth grade, my parents bought a new Sears Silvertone black-and-white TV and splurged for a new cable TV connection, and for the first time, we received all four channels.  In later years, this gave me access to Star Trek reruns (See Star Trek@50: Inspiring Discovery and Innovation) and The Dick Cavett Show, which featured an interesting and eclectic set of guests unknown to a small town boy. (See Eudora, You Got the Love?)

Until the 1980s, the four channel cultural gestalt of broadcast television extended far beyond news.  In those days, people remembered where they were and those around them when they learned about the birth of Little Ricky, Who Shot J.R., and the ending of M*A*S*H. (For the latter, I was sitting on a hotel room bed in Madison, Wisconsin.) The morning-after water cooler bonding from discussing I Love Lucy, Bonanza, M*A*S*H. Dallas, and hundreds of other hit television shows has now disappeared.

Top 40 Radio

Top 40 AM radio had a similar cultural reach, where being in heavy rotation was a clear sign that a song was a national hit, known to most teenagers and adults.  Then each Saturday, Casey Kasem’s American Top 40, counted down “the biggest hits in the U.S.A.” and American Bandstand, with the ageless Dick Clark, highlighted a popular singer or band.  I heard and saw little of that, as I had neither a radio nor a television in my bedroom. Nor did I buy and play 45 RPM singles or 33 and 1/3 RPM albums. All of these were far beyond my financial reach.

My parents owned a small kitchen radio, which we only turned on as we prepared for work and school.  Lacking an external antenna, we received only clear channel AM stations and a few nearby FM stations. Like the television, it was an aging vacuum tube radio, which took a while to warm up after it was turned on.  When it failed, we would take the failing tube, which no longer lit up, to the local drugstore, use the tube tester, and buy a compatible replacement tube.

Tuned to the local AM station (KALM), we received news read from the Associated Press wires and a healthy dose of country music.  I became well acquainted with Conway Twitty, Loretta Lynn, Tammy Wynette, George Jones, and a variety of other country singers, but I didn’t like the affected twang and the frequent stories of pickup trucks and lost love, all in search of what David Allan Coe jokingly sang about as the perfect country song:

Well, I was drunk the day my mom got out of prison
And I went to pick her up in the rain
But before I could get to the station in my pickup truck
She got run over by a damned old train

The exception was Johnny Cash; I loved Cash from the very first time I heard him.  As a poet and a fellow Arkansas boy, he spoke to deeper things – social inequities, the war, and personal loss. There are few things more melancholy than his rendition of Sunday Morning Coming Down or more haunting than his late life cover of Hurt, a song by Nine Inch Nails. (Never underestimate the power of a great singer to make another’s song their own. Heart’s Kennedy Center Honors cover of Stairway to Heaven brought Led Zeppelin to tears.)

Surprisingly, we did have the radio on one afternoon on New Year’s Eve, when the local DJ cum newscaster said, with some mocking mirth, that he was going to play the most popular song in the country for that year, the implication being that this music was drug-crazed rock and roll. With that sardonic introduction, Don Mclean’s American Pie began to play, and I was immediately transfixed with wonder.  Here was another poet, as evocative as Cash, speaking to a different generation about sadness and loss.

Given our household’s one radio and shared TV, other than a few TV variety shows like The Hollywood Palace, my access to rock and roll was largely limited to our car, but only after I learned to drive in 1973 and could pick up a Memphis rock station.  As a result, I missed the British Invasion and almost all of the music of the 1960s and early 1970s. I do remember one of the older girls who lived nearby being devastated when the Beatles broke up, though I didn’t understand why it was such a big deal. 

Digital Transformation

The uniformity of radio and television content in the post-World War II era created common cultural referents, but they largely targeted white, middle class families. If you were black, Hispanic, Asian, immigrant, poor, rural, or in any other way not part of the white middle class, you were rarely heard or seen, except in caricature or stereotype. (For the record, the Clampetts on The Beverly Hillbillies were unlike anyone I knew in my small Ozarks town.)

Slowly and haltingly, things began to change for the better, as radio and television began to reflect the population’s social and economic diversity and its wide range of interests more accurately. Antiwar demonstrations and civil rights protests forced broadcast news to begin presenting important social issues in a more nuanced way, highlighting previously marginalized voices, and television shows began doing the same (e.g., see All in the Family).

Concurrently, legal and financial battles over cable television must carry rules, the rise of cable superstations such as WGN and WTBS, the emergence of FOX as a fourth television network, new pay-per-view services, movie networks such as HBO, and expanded cable capacity all diluted and diffused the market dominance of the big three commercial broadcast networks.  Suddenly, everyone could get their MTV.

In the midst of this, the transition to high definition, digital television began, which changed wireless spectrum needs.  Notably, as analog television stations vacated parts of the VHF and UHF bands, portions of this spectrum were repurposed.  While at Microsoft, I served as a member of the Federal Communications Commission (FCC) Technical Advisory Committee and worked with the International Telecommunication Union (ITU) on the use of the TV white spaces for unlicensed “super Wi-Fi” to address rural and underserved communities and regions in the U.S. and broadband access across the world. (See White Spaces: Celebrating the Cambridge Trial.)  Remember, analog broadcast television originally chose the VHF and UHF bands due to their long distance propagation characteristics, making them idea for communication across long distances.

Today, only live sporting events can deliver the viewership, advertising revenues, and shared experiences once common on broadcast television.  It’s why the bidding for U.S. television rights to the Olympics, the NFL and Super Bowl, and March Madness (NCAA men’s basketball) is now so intense.

Just as these brave new worlds of cable and digital television were reshaping mass media, the Internet and the World Wide Web appeared, something I was privileged to witness firsthand at the University of Illinois when the Mosaic web browser appeared.  As broadband availability and speeds increased, new access modes appeared, and creative market destruction soon followed.  (As the COVID-19 pandemic has again exposed, broadband inequity in the United States – availability, cost, and bandwidth – remains a serious problem. See Broadband: Oxygen for a Digital World.)

Although there were effective Internet search engines before Google search appeared (e.g., AltaVista), Google found a successful revenue model based on targeted advertising. After all, if you are searching for left-handed scissors, you might want to buy a pair. Combining this information with search history and web cookies could delivered a prospective customer hit rate rarely matched by the blanket advertising from newspapers, radio, and television. This devastated the traditional revenue models of daily newspapers, forcing some to merge with rivals and many others to close. Sadly, this has meant many towns and cities now lack local, independent investigative journalism.

Similarly, the convenience of Netflix movie streaming, with thousands of movies just a click away, destroyed the VHS and DVD rental market. Although Netflix initially distributed rental DVDs by mail, it was created in anticipation of higher broadband speeds that would enable streaming. Co-founder Reed Hastings once said this was why they didn’t name the company DVDflix. Coupled with inexpensive, large screen high-definition televisions, the locus of money and power soon shifted from the traditional movie studies, though few would have then imagined that Amazon would someday buy MGM to feed its Amazon Prime Video streaming service.

Paradoxically, Internet broadband gave birth to narrowcasting – websites, channels, and streaming media targeting very specific demographics. Enamored with reruns of The Gong Show (“We’ll be right back with more – stuff!”), The Lawrence Welk Show (my parents loved it) or The Midnight Special? Interested in collimating your Ritchey–Chrétien telescope, folding a paper airplane, or collecting vintage salt and pepper shakers?  Fear not, YouTube is there for you.  Love reggae/zydeco fusion or KISS tribute bands?  There’s a streaming music service for just your interests. Meanwhile, music CD sales (and even downloads) continue to plummet, following the same path to extinction as DVDs and Blu-ray discs.

The Triumph of Narrowcasting

In the early 21^st century, digital narrowcasting has clearly triumphed – whatever the interest, there’s now an app or a targeted steam for it. Every other form of media distribution, from print news organizations through movie studios to broadcast television networks, is now scrambling to adapt, with varying success.  Not surprisingly, any such change brings both new opportunities and associated challenges.

On the one hand, digital narrowcasting has been hugely democratizing, freeing individuals and organizations to communicate directly with their target demographic. This blog is an example, allowing me to ruminate on any topic of my choosing, then share my ideas using social media, something only possible because instant, global communication is now effectively free. Similarly, websites, streaming media, and other tools mean individuals and shared interest groups can quickly find likeminded partners and collaborators, however narrow or arcane their interests.

Historically, the small number of broadcast platforms – newspapers, radio, and television – each had a large audience. Today, a plethora of platforms compete for a much smaller slice of the global audience, some by choice and others by necessity. Although everyone can be a social media content creator, with a potentially global audience, in practice, only a few achieve any significant influence or visibility. As the late Nobel Laureate Herbert Simon astutely noted, “… a wealth of information creates a poverty of attention.”

The insatiable maw of digital channels must be filled, even if with endless and mind-numbing repetition. Just how many home renovation knockoffs of PBS’s This Old House can HGTV make? How many reruns of the Law and Order franchise can competing cable channels show simultaneously? Can the hosts of the cable news channels talk any more while saying even less, filling the minutes with personal opinions and creating venues that attract only the likeminded?

Digital Echo Chambers

As I look to Claude Shannon, the geek’s patron saint of information theory, I fear not all is well in the 21^st century world of digital narrowcasting; the signal-to-noise ratio of digital discourse is now near de minimis, with a cacophony of discordant voices vying for finite mindshare. Is it possible for us to imagine and create a digital world where we more clearly distinguish information (facts) from opinion, persuasion, and social consensus? Each has a role, when clearly demarcated and delineated, but they are not the same.

Facts are independently testable and verifiable, with truth independent of one’s own opinions or beliefs. It is raining at this place and this time is a testable statement, provably true or false. An opinion is not a fact. “I think it is raining” is an opinion; it may or may not be accurate. Fact-based persuasion is convincing someone else it is raining using clear, verifiable evidence that it is, in fact, raining; demagoguery is not. Social consensus is a widely held perspective, best supported by history and evidence.  It is bad if it rains on an outdoor baseball game is social consensus. (See Just the Facts, Ma’am: Reasoning is Not Dead, Jim.)

With that backdrop, the goal of any reputable information source should not be to tell you how to think; it should be to provide you with facts that allow you to develop your own hypotheses, test them against evidence, then revise them accordingly. Unbiased information – the facts – is the prelude to evidence-based reasoning and informed discourse. Such informed national discourse requires shared intellectual context, something in danger of being lost in a narrowcasting world.

All too often, the explosion of tailored communications and personal preferences can trap any one of us in a self-reinforcing echo chamber that separates and potentially demonizes those unlike us, confusing facts with opinions and beliefs. Put another way, if you have been sitting in your underwear for multiple days, binging on potato chips and beer, watching endless reruns of The Big Bang Theory, surfing the dark web, and reading extremist social media (either from the far right or the far left), it’s time to get dressed, open the curtains, seek information, and expose yourself to other perspectives, especially those that differ from your own. Fact-based understanding and multilateral communication are a messy, participatory process – one must listen, talk, and think – but from it, mutual understanding and shared perspectives, both grounded in facts, can emerge.

In higher education today, there is great focus on imparting knowledge and inculcating skills, and they are important, because employers prize them.  However, knowledge transmission pales in importance relative to teaching critical thinking – the ability to analyze evidence, test ideas, adjust one’s perspective, and form rational, thoughtful and nuanced insights. Critical thinking is the upgraded brain software that operates on the data we call knowledge. It is the hallmark of an educated citizenry.

To appreciate nuance, to understand discord, to feel empathy, one must glean the facts and also listen to diverse perspectives.  Read multiple newspapers, domestic and international; consider multiple websites from around the world; listen to news channels from the left, the middle, and the right; talk to people from different backgrounds.  Above all, think critically about what you learn; ask questions; seek your answers supported by facts.  Remember, not everyone who disagrees with you is ipso facto wrong; nor are you necessarily right if all your friends agree with you.

Cronkite knew that, as did Huntley and Brinkley. Each night, they and their compatriots did their best to bring the audience some unvarnished facts to ponder. They reserved editorial comment for rare and extraordinary occasions, and they labeled it explicitly as such, which gave it great power.  Today, the best news organizations and high-quality information sources strive to do the same.

Yes, broadcast news stories were and are filtered by mainstream cultural bias, but they are the mainstream. Though not always right, the mainstream, then and now, is more than a neologism.  It is a shared and evolving Weltanschauung created by the interplay of contrasting and often divergent perspectives, itself derived from broad communication and critical thinking. (See Shaping the Message, Using the Medium and Contemplative Reflection and Instantaneous Communication.)

And that’s the way it is, March 29^th, 2022.  Goodbye, Chet. Goodbye, David. And goodnight from HPCDan.