Commerce’s CHIPS Contortions

How not to reanimate the undead industrial base of America

America disemboweled its own industrial base, a dishonorable yet intentional seppuku of the economic dreams of the middle class and the long-term strategic competitiveness of the nation. Now after decades of watching jobs swim far away from American shores and factories rise overseas, there’s momentum to at least act like that matters.

After an era of denigrating industrial policy and letting markets be markets, Congress finally conceded the reality that well-executed, government-initiated economic coordination is foundational to industries that require gargantuan capital investment alongside complex synchronization of academic scientific research, workforce talent development and robust supply chains. The CHIPS and Science Act passed the Senate nearly 2:1, and among its many provisions, the law offers tens of billions of dollars to the semiconductor ecosystem with the goal of returning America to the competitive frontiers of fabrication.

Now for the painfully tough part: to win the victory.

This was a huge week for the Act, with the CHIPS Program Office at the Department of Commerce making its first major announcements on the application process and decision criteria it will use to select the winners for the first tranche of funding from the law. Unfortunately, those announcements showed the two-faced nature of America’s approach to its economy: leadership and a drive to win, but with a deep desire to lose in the details. The doctor gets the prognosis right, but dithers on the available treatments.

The announcements started positively with Commerce Secretary Gina Raimondo’s trenchant remarks at Georgetown last week outlining her vision for the future of America’s prosperity. A former venture capitalist as founder of Point Judith Capital and governor of Rhode Island, she began with a look at the industry back in the halcyon Silicon Valley years of the 1950s and 1960s (highlight in bold):

And although the chip companies were fiercely competitive, there was an industry-wide effort to advance the technology. The government drove these advances through purchases and tech transfers. Tens of thousands of engineers in these companies would make daily incremental innovations in manufacturing techniques, resulting in improved scaling and yield, through expertise that is only possible by producing millions and millions of wafers. This relentless pace of lab-to-fab and fab-to-lab innovation became synonymous with America’s tech leadership, doubling our computing capacity every two years. This ecosystem enabled every smartphone, cloud computing service, new car, medical device, and weapons system we use today. But what was once a self-propelling engine of innovation and production fell out of balance. We sacrificed our manufacturing capacity and workforce in the mistaken belief that we could somehow maintain our technological leadership without them. In 1990, the U.S. accounted for 37% of global chip manufacturing capacity. Today, that number is only 12%. We once manufactured nearly all of the world’s most advanced semiconductors. Today, we manufacture none.

She then crescendoed to an audacious vision for America’s leadership in chips:

I want the United States to be the only country in the world where every company capable of producing leading-edge chips will have a significant R&D and high-volume manufacturing presence. We will be the premier destination in the world where new leading-edge chip architectures can be invented in our research labs, designed for every end-use application, manufactured at scale and packaged with the most advanced technologies. This combination of technological leadership, supplier diversity, and resiliency does not exist anywhere else in the world today.

It wasn’t the most soaring of rhetoric, but the historical perspective was accurate and the strategy for America’s future was thoughtfully constructed. The United States lost the battle for semiconductor supremacy to far more competitive ecosystems in Korea, Taiwan, and Japan, but it hasn’t lost the war in totality. The latent talent, research labs and capital exist to reconstitute the industry, and careful synchronization and capital investment could reembowel the corpse of America’s semiconductor manufacturing base.

What I (and presumably the industry) was looking for was an all-hands-on-deck competitive fusillade that would simultaneously ameliorate America’s competitive disadvantages: high labor costs, soaring construction expenses, low infrastructure investment, meager talent pipeline, and limited university-industry coordination.

Apparently, word never got through to the CHIPS program staff that the cadaver of the American semi industry is in the emergency department and not a California holistic wellness center. For what we got instead of an aggressive economic resuscitation plan was a wish list of other priorities outside the singular strategic goal: investing toward the zenith of competitiveness in an industry where America is now years behind the fab frontier.

Across six fact sheets (that’s a lot of facts!), Commerce set out its criteria for judging applications for funding. The first fact sheet outlined the application process and was complemented by five other fact sheets that highlighted the department’s criteria across a series of themes. Miraculously, the department managed to insert a poison pill provision in each theme to undermine the overall competitiveness of the program. Walking through them one by one:

“Catalyzing Private Investment” — Smart encouragement of private capital and creative financial structures poisoned by a cash flow requirement that demands that companies receiving CHIPS funding pay back unanticipated upside resulting from CHIPS funds.
“Protecting U.S. Taxpayers” — Rigorous transparency requirements marred by a focus on “[evaluating] applications based on the extent of the applicant’s commitments to refrain from stock buybacks. Applicants are also prohibited by law from using CHIPS funds for dividends or stock buybacks.”
“Building a Skilled and Diverse Workforce” — Reasonable nostrums of encouraging sectoral partnerships and coordinating with workforce partners decimated by requirements to pay prevailing wages for construction (i.e. the highest union wage in a region) and requiring that on-site child care be provided to all construction and on-going semiconductor fab workers.
“Engaging with U.S. Partners and Allies” — Excellent balance of raising up America's manufacturing in concert with allies tainted by a 10-year ban on semiconductor investments in “any foreign country of concern” (i.e. China).
“Spurring Regional Economic Development and Inclusive Economic Growth” — Thoughtful goals to improve regional semiconductor hubs tarnished by a demand that all fabs operate with “100% renewable energy” and “[c]ommit to using iron, steel, and construction materials produced in the United States as part of their projects.”

Here we see the crux of Commerce’s CHIPS contortions: the department simultaneously wants to win the fight for semiconductor supremacy (an almost impossible goal albeit one that is still attainable) while also limiting corporate profitability and shareholder value, expanding child care and labor unions, banning global expansion, and hitting Paris climate goals.

For me and you as readers, some of Commerce’s goals might be ones we share and want to prioritize. I for one would love to see more universal child care for workers as a way of equalizing employment opportunities across all demographic groups and particularly women. But as always with American industrial policy, we can’t seem to keep our eyes exclusively on the (industry) prize; we have to multitask, with the concomitant drop in focus and ultimate performance that our approach portends.

I wrote back in “Consensus Functions” last year (with a highlight in bold):

Society, meanwhile, doesn’t have all those layers of consensus to build upon for new decisions. There’s no algorithmic blockchain ensuring that the basic facts of reality are cross-validated, or the scientific method to ensure that evidence is considered with appropriate context. Consensus is recursive, and without better consensus functions around values and tradeoffs, it’s impossible for a nation to make decisions. America and much of the West escaped that lack of deep consensus simply through abundance. Multiple values and tradeoffs could be supported by building institutions that allowed them all to coexist. Venture funds can’t make every investment and scientists can’t accept every contradictory result, but wealthy nations can support a wide range of consensuses, each underwritten in parallel.

America is a wealthy nation, but competition is catching up very, very quickly. China, only recently a rural and immiserated economic backwater, fell just shy of the World Bank’s high-income threshold for 2022, hitting a per capita income of $12,608 against a benchmark of $13,205. Korea’s domestic Hyundai auto workers now hold salaries equivalent to American Midwest factory labor.

America isn’t a leader protecting its edge in chips. America is a follower racing to catch up with smarter and faster rivals, or as Raimondo remarked, "We once manufactured nearly all of the world’s most advanced semiconductors. Today, we manufacture none.” High labor costs and low productivity drove many factories overseas, and our response is to add wage increases and demand more benefits for workers? American fabs are the most expensive in the world to construct and operate, but they’re also supposed to solve climate change and union membership growth as well? Intel, as I wrote in “Intel’s Malaise”, hit a profitability nadir last year simultaneous with the passage of the CHIPS and Science Act. Will banning excess profits, dividends, stock buybacks and overseas expansion in China turn around its bleak corporate narrative?

One can hope that this week’s fact sheets are merely desires expressed by the department to feign fealty to President Joe Biden and liberal economic priorities, and the ultimate applications selected will be far more internationally competitive. I doubt that, for the same reason I doubt that America can get its industrial base back: it never wants to truly mobilize all facets of government and industry to “do what it takes” to win.

The patient’s organs are splayed out on the operating table, and we’re making sure the artificial heart is zero carbon and domestically made (“It uses recycled aluminum, right?”). The CHIPS rules are a perfect vivisection of America’s industrial policy malaise, and the disembowelment of U.S. factory jobs continues.

“Securities” Podcast: Observations on Generative AI

OpenAI’s ChatGPT and generative AI more broadly have been slathering the news the past few months, and that excitement has translated into a white-hot venture market for the next generation of AI startups. We previously discussed some of the limits and ethical issues around this crop of large language models with NYU professor and AI critic Gary Marcus on the podcast in “That’s 100% what keeps me up at night”, and I also talked last week in the newsletter about the coming creative class warfare in “Garrulous Guerrilla.”

This week, I wanted to get more of a ground floor perspective on what’s happening on the frontlines of the AI wars, so I had Lux’s own partner Grace Isford join me to talk about what she’s seeing in the market right now. We talk about the first-ever AI Film Festival, how a new generation of tools is empowering developers to incorporate generative AI into their applications, and how the big trio of Microsoft, Google and Amazon are approaching the market.

🔊 Take a listen to “May the AI be ever in your favor”

Lux Recommends

“Securities” producer Chris Gates chimes in with his first-ever Lux Recommends with an episode of The Ezra Klein Show interviewing author Adrian Tchaikovsky of the Children of Time trilogy. Chris is deep into the third and final book, Children of Memory, and heartily recommends what’s described as a space opera on “humanity’s battle for survival on a terraformed planet.”
In terms of breakthrough papers, Shaq Vayda recommends “De novo design of luciferases using deep learning” just published in Nature. "The creation of highly active and specific biocatalysts from scratch with broad applications in biomedicine is a key milestone for computational enzyme design, and our approach should enable generation of a wide range of luciferases and other enzymes.”
Our scientist-in-residence Sam Arbesman recommends Marcin Wichary’s post chronicling the work of Dana Sibera, who devises ingenuous (and often useless) tech product designs that illuminate and reconfigure our notions of how we interact with our devices. Dozens of brilliant images included: “These ‘mock-ups that mock’ all feel great to me, a perfect antidote to the pretentious Jonny Ive white-room videos of the previous decade.”
I recommend “Israel risks turning into a shut-down nation” in the Financial Times by John Thornhill describing how the Israeli tech community is mobilizing itself against the recently reinstated premiership of Benjamin Netanyahu. “’Restraint has been thrown out the window. Tech people are the muscle behind the protest movement,’ another entrepreneur told me. ‘It’s great that we live in a real democracy as we are seeing today. But the situation is scary as shit.’”
Zaid Fattah recommends Florian Kronawitter’s Substack “The Next Economy”, which covers macro and micro economic trends.
Finally, Sam recommends this retrospective from the Computer History Museum's David C. Brock in IEEE Spectrum, “50 Years Later, We’re Still Living in the Xerox Alto’s World. "They conceived and developed the Alto in a remarkable burst of creativity, used it to develop diverse and pathbreaking software, and then moved out of Xerox, taking their achievements, design knowledge, and experiences into the wider world, where they and others built on the foundation they had established.”

That’s it, folks. Have questions, comments, or ideas? This newsletter is sent from my email, so you can just click reply.

Forcing China’s AI researchers to strive for chip efficiency will ultimately shave America’s lead

In incididunt ad qui nostrud sint ullamco. Irure sint deserunt Lorem id officia dolore non. Anim dolor minim sit dolor et sint aliquip qui est. Ex in tempor laborum laboris dolor laboris ullamco quis. Enim est cupidatat consequat est culpa consequat. Fugiat officia in ea ea laborum sunt Lorem. Anim laborum labore duis ipsum mollit nisi do exercitation. Magna in pariatur anim aute.

In incididunt ad qui nostrud sint ullamco. Irure sint deserunt Lorem id officia dolore non. Anim dolor minim sit dolor et sint aliquip qui est. Ex in tempor laborum laboris dolor laboris ullamco quis. Enim est cupidatat consequat est culpa consequat. Fugiat officia in ea ea laborum sunt Lorem. Anim laborum labore duis ipsum mollit nisi do exercitation. Magna in pariatur anim aute.

‍

Right now, pathbreaking AI foundation models follow an inverse Moore’s law (sometimes quipped “Eroom’s Law”). Each new generation is becoming more and more expensive to train as researchers exponentially increase the number of parameters used and overall model complexity. Sam Altman of OpenAI said that the cost of training GPT-4 was over $100 million, and some AI computational specialists believe that the first $1 billion model is currently or will shortly be developed.

As semiconductor chips rise in complexity, costs come down because transistors are packed more densely on silicon, cutting the cost per transistor during fabrication as well as lowering operational costs for energy and heat dissipation. That miracle of performance is the inverse with AI today. To increase the complexity (and therefore hopefully quality) of an AI model, researchers have attempted to pack in more and more parameters, each one of which demands more computation both for training and for usage. A 1 million parameter model can be trained for a few bucks and run on a $15 Raspberry Pi Zero 2 W, but Google’s PaLM with 540 billion parameters requires full-scale data centers to operate and is estimated to have cost millions of dollars to train.

Admittedly, simply having more parameters isn’t a magic recipe for better AI end performance. One recalls Steve Jobs’s marketing of the so-called “Megahertz Myth” to attempt to persuade the public that headline megahertz numbers weren't the right way to judge the performance of a personal computer. Performance in most fields is a complicated problem to judge, and just adding more inputs doesn't necessarily translate into a better output.

And indeed, there is an efficiency curve underway in AI outside of the leading-edge foundation models from OpenAI and Google. Researchers over the past two years have discovered better training techniques (as well as recipes to bundle these techniques together), developed best practices for spending on reinforcement learning from human feedback (RLHF), and curated better training data to improve model quality even while shaving parameter counts. Far from surpassing $1 billion, training new models that are equally performant might well cost only tens or hundreds of thousands of dollars.

This AI performance envelope between dollars invested and quality of model trained is a huge area of debate for the trajectory of the field (and was the most important theme to emanate from our AI Summit). And it’s absolutely vital to understand, since where the efficiency story ends up will determine the sustained market structure of the AI industry.

If foundation models cost billions of dollars to train, all the value and leverage of AI will accrue and centralize to the big tech companies like Microsoft (through OpenAI), Google and others who have the means and teams to lavish. But if the performance envelope reaches a significantly better dollar-to-quality ratio in the future, that means the whole field opens up to startups and novel experiments, while the leverage of the big tech companies would be much reduced.

The U.S. right now is parallelizing both approaches toward AI. Big tech is hurling billions of dollars on the field, while startups are exploring and developing more efficient models given their relatively meagre resources and limited access to Nvidia’s flagship chip, the H100. Talent — on balance — is heading as it typically does to big tech. Why work on efficiency when a big tech behemoth has money to burn on theoretical ideas emanating from university AI labs?

Turn to China now, and the story is in a very different place. The Biden administration has placed a series of export controls on advanced computation and semiconductor fabrication to China, including blocking the export of Nvidia’s H100 and A100 chips. Nvidia responded in March by offering tweaked versions that comply with U.S. export controls (with what it dubs the H800 chip), but the general policy remains in force: prevent China from acquiring the best chips that Earth produces.

Without access to the highest-performance chips, China is limited in the work it can do on the cutting-edge frontiers of AI development. Without more chips (and in the future, the next generations of GPUs), it won’t have the competitive compute power to push the AI field to its limits like American companies. That leaves China with the only other path available, which is to follow the parallel course for improving AI through efficiency.

For those looking to prevent the decline of American economic power, this is an alarming development. Model efficiency is what will ultimately allow foundation models to be preloaded onto our devices and open up the consumer market to cheap and rapid AI interactions. Whoever builds an advantage in model efficiency will open up a range of applications that remain impractical or too expensive for the most complex AI models.

Given U.S. export controls, China is now (by assumption, and yes, it’s a big assumption) putting its entire weight behind building the AI models it can, which are focused on efficiency. Which means that its resources are arrayed for building the platforms to capture end-user applications — the exact opposite goal of American policymakers. It’s a classic result: restricting access to technology forces engineers to be more creative in building their products, the exact intensified creativity that typically leads to the next great startup or scientific breakthrough.

If America was serious about slowing the growth of China’s still-nascent semiconductor market, it really should have taken a page from the Chinese industrial policy handbook and just dumped chips on the market, just as China has done for years from solar panel manufacturing to electronics. Cheaper chips, faster chips, chips so competitive that no domestic manufacturer — even under Beijing direction — could have effectively competed. Instead we are attempting to decouple from the second largest chips market in the world, turning a competitive field where America is the clear leader into a bountiful green field of opportunity for domestic national champions to usurp market share and profits.

There were of course other goals outside of economic growth for restricting China’s access to chips. America is deeply concerned about the country’s AI integration into its military, and it wants to slow the evolution of its autonomous weaponry and intelligence gathering. Export controls do that, but they are likely to come at an extremely exorbitant long-term cost: the loss of leadership in the most important technological development so far this decade. It’s not a trade off I would have built trade policy on.

The life and death of air conditioning

Across six years of working at TechCrunch, no article triggered an avalanche of readership or inbox vitriol quite like Air conditioning is one of the greatest inventions of the 20th Century. It’s also killing the 21st. It was an interview with Eric Dean Wilson, the author of After Cooling, about the complex feedback loops between global climate disruption and the increasing need for air conditioning to sustain life on Earth. The article was read by millions and millions of people, and hundreds of people wrote in with hot air about the importance of their cold air.

A few weeks ago, Bloomberg had a great piece about the rising toll of air conditioning under the similarly scary headline, “A Billion New Air Conditioners Will Save Lives But Cook the Planet.” The numbers, as always, are staggering:

Demand for air conditioners is surging in markets where both incomes and temperatures are rising, populous places like India, China, Indonesia and the Philippines. By one estimate, the world will add 1 billion ACs before the end of the decade. The market is projected to
before 2040. That’s good for measures of public health and economic productivity; it’s unquestionably bad for the climate, and a global agreement to phase out the most harmful coolants could keep the appliances out of reach of many of the people who need them most.

This is a classic feedback loop, where the increasing temperatures of the planet, particularly in South Asia, lead to increased demand for climate resilience tools like air conditioning and climate-adapted housing, leading to further climate change ad infinitum.

And there doesn’t appear to be a way out. Scientists have determined that there is a zone of comfort for humans where we can function at peak performance. Much of the industrialized world happens to sit in these temperate zones, which is one argument for their long-term economic success compared to the tropics. But as climate change intensifies, the world’s temperate zones are expected to narrow dramatically, and some countries like India may have next to no temperate zones left to protect its people. Air conditioning is a necessity (with some even calling it a human right), but greener coolant alternatives are still in their infancy. In other words, cooling innovation needs to heat up in a big way.

Lux on the Farm

Josh Wolfe speaking at Nvidia Auditorium at Stanford. Photo by Chris Gates.

Josh Wolfe gave a talk at Stanford this week as part of the school’s long-running Entrepreneurial Thought Leaders series, talking all things Lux, defense tech and scientific innovation. The
.

Lux Recommends

As Henry Kissinger turns 100, Grace Isford recommends “Henry Kissinger explains how to avoid world war three.” “In his view, the fate of humanity depends on whether America and China can get along. He believes the rapid progress of AI, in particular, leaves them only five-to-ten years to find a way.”
Our scientist-in-residence Sam Arbesman recommends Blindsight by Peter Watts, a first contact, hard science fiction novel that made quite a splash when it was published back in 2006.
Mohammed bin Rashid Al Maktoum, and just how far he has been willing to go to keep his daughter tranquilized and imprisoned. “When the yacht was located, off the Goa coast, Sheikh Mohammed spoke with the Indian Prime Minister, Narendra Modi, and agreed to extradite a Dubai-based arms dealer in exchange for his daughter’s capture. The Indian government deployed boats, helicopters, and a team of armed commandos to storm Nostromo and carry Latifa away.”
In the happier news department of science, Ina Deljkic recommends “’Quite science fiction’: Brain-spine interface system helps man with paralysis walk again.”
Sam recommends Ada Palmer’s article for Microsoft’s AI Anthology, “We are an information revolution species.” “If we pour a precious new elixir into a leaky cup and it leaks, we need to fix the cup, not fear the elixir.”
I love complex international security stories, and few areas are as complex or wild as the international trade in exotic animals. Tad Friend, who generally covers Silicon Valley for The New Yorker, has a great story about an NGO focused on infiltrating and exposing the networks that allow the trade to continue in “Earth League International Hunts the Hunters.” "At times, rhino horn has been worth more than gold—so South African rhinos are often killed with Czech-made rifles sold by Portuguese arms dealers to poachers from Mozambique, who send the horns by courier to Qatar or Vietnam, or have them bundled with elephant ivory in Maputo or Mombasa or Lagos or Luanda and delivered to China via Malaysia or Hong Kong.”
Finally, Sam recommends Brick Technology’s video of “Making A Billion-Year Lego Clock.”

That’s it, folks. Have questions, comments, or ideas? This newsletter is sent from my email, so you can just click reply.

‍

February 25, 2023

ChatGPT as Garrulous Guerrilla

<- Previous

continue

reading

March 18, 2023

Disaster Performance

Next ->