Omnivorous Explanations

Can we even try to explain what’s happening right now?

I was recently asked to provide a CliffsNotes summary of the macro chaos enveloping our world, and my response was a meme of Homer Simpson carrying a sign that displayed “The End is Near.” Flippant, if accurate.

The original question though is worth pondering: How exactly do you summarize the present state of the world? And with each tightening of brevity, what should an analyst or storyteller leave out?

I joked back in “Dissonant Loops” that “The Economist could move from a weekly to an hourly publication and still fill its dense pages.” It’s as true today as it was two months ago, if not more so as we head into the fall U.S. election cycle, Europe begins to stare into a winter abyss of energetic discontent, and China's 20th National Congress convenes.

The world is not a well-manicured system anymore, with abstractions that allow individuals to understand their part of the whole without glimpsing the immense edifice of it all. Not unlike art, computer science prioritizes abstractions above almost all other principles (not surprising then that one of the most famous books in the field is The Art of Computer Programming by Donald Knuth). Well-abstracted code allows the software engineer to ignore whole sections of the codebase, saving mental capacity and intensifying focus. Maintain one part, and everyone else will maintain theirs.

Functioning abstractions was the theme of the global international system the past two decades. Chinese factories financed by European capital produced goods that were placed on South Korean-built shipping vessels crewed by international sailors that then arrived on American shores and ultimately delivered goods onto American shelves. As we discussed on the podcast in “The geopolitics and digital future of agricultural commodities,” most people don't want to think about how food gets to their table. In fact, the miracle of modernity is that food arrives on shelves with almost no thought from consumers whatsoever.

And that doesn’t just apply to food commodities. Anyone who has tried to peek through Daniel Yergin’s duet of The Prize (929 pages) and The Quest (834 pages) knows that oil is exceedingly hard to capture as a market. Ditto for metals, energy, and finished goods, and that’s not even getting to intellectual products. The beauty though is that you didn’t need to read Yergin until recently to fill up your car — you popped the fuel cap, placed the nozzle in, and the entire world system delivered those precious and energy-dense black gold drops into the tank. And at an incredible price to boot.

Now we need to think about the ramifications of Russia’s war on Ukraine on oil price inflation. And the droughts in China and its effect on rice and pork prices. There are thousands of individual phenomena woven together, which itself creates a whole slew of second-order effects that have huge impacts on our daily life — let alone investments, company building or the future of the nation.

One bandaid that’s emerged has been the soothsaying explanatory oracles who now litter the bestseller shelves at bookstores. I’m thinking of Adam Tooze, whose Shutdown: How Covid Shook the World’s Economy rocketed up bestseller lists upon debut. Or Niall Ferguson’s recent migration from the history of money to the simply-named Doom: The Politics of Catastrophe, or even backing up a bit in time to Yuval Noah Harari’s Sapiens. In all cases, authors (and really, men) who try to condense the immense complexity of the global macro environment into the confines of an acceptable consumer-length paper volume.

How does Tooze do it? In Molly Fischer’s profile of him for New York magazine, the answer appears to be to just engorge on more and more information, and then spit it back out as fast as possible:

The Tooze Bros look upon his productivity in wonderment. “It is mind-boggling. Beyond mind-boggling. I mean, I’ll see him tweeting at 4 a.m. sometimes,” [Columbia student Henry Williams] said. “I do that sometimes, but that’s really a mistake. That’s a problem for me.” … Since the beginning of 2022, Tooze has published eight articles in The New Statesman, Foreign Policy, and The Guardian; 40 editions of [his newsletter] Chartbook (not including link roundups); and 1,500 tweets. “The man’s output is insane,” said [Ethan Winter, an analyst at the think tank Data for Progress]. “He just crushes content. I have no idea how it’s possible.”

Fischer summarizes the situation:

What Tooze gives a reader like Williams is not a piercing, singular insight but a sense of rigorous mastery. […] Tooze’s great intellectual power is a gift for synthesis. “He just digests staggering amounts of information,” said Ted Fertik, one of his former Ph.D. students and now a policy strategist. Tooze roves across vast fields of data — historical data, technical data, data about Russian currency reserves, data about the Nazi steel-tube industry — and returns with a reasonably accessible brief in hand.

When confronted with a massive heap of software code, Tooze — like his other oracular peers — simply memorizes the entire codebase. Remove the abstractions and just read the whole thing. Don’t focus on one well-designed 10,000 line module, but just download the entire system and let the brain have at it. If everything in the system is connected, well, better get reading. It’s almost reminiscent of the Whiz Kids of the Pentagon in the Robert McNamara era — just collect more data and jam it into more and more theories. It’s all dominoes.

The bounded rationality here should be obvious though. Without abstractions, the cognitive burden of “the world” is more than any mind can process. No one can understand all the shapes, let alone all the lines or all the dots of the world as it is.

James M. Banner Jr. recently wrote an essay for Humanities, the magazine of the National Endowment for the Humanities titled “All History Is Revisionist History.” He lays out a distinction about the past that I think is equally helpful to understanding our current present:

A fundamental feature of historical thought is the distinction between “the past” and “history.” What we call “the past” is just that: It’s what happened at some point before now. Once it occurs, “the past” is gone forever—beyond repeating, beyond reliving, beyond replicating. It’s recoverable only by the evidence, almost never complete, that it leaves behind; and that evidence must be interpreted by individual humans—historians principally, but archaeologists, anthropologists, and others, all of whom differ in all sorts of ways. Distinct from “the past” are the narratives and analyses that historians offer about earlier times. That’s what we call “history.” History is what people make of the forever-gone past out of surviving documents and artifacts, human recall, and such items as photographs, films, and sound recordings. Indeed, history is created by the application of human thought and imagination to what’s left behind.

Our present is created by the application of human thought and imagination to what’s before us. What Tooze, Ferguson, Harari and others are attempting to do is narrate what are otherwise unfathomable changes. It’s CliffNotes, if better written.

Yet, when I think of what’s needed to actually solve the on-going crises of the world, it’s not more narratives and more “understanding.” It’s really about agility, both individual and collective. To use the military mental model, it’s OODA — observe, orient, decide, and act. Narratives get at the first “O”, but do nothing to orient and drive decisions and actions together.

In a world of such complexity and emergence, it’s better not to build extensive intelligence systems but to just be fast. Learn quickly, adapt quickly. The Fed raised interest rates 75bps this week, while also publishing a new projection of future interest rates. Per the Financial Times:

[Fed chair Jay Powell’s] remarks followed the publication of a new “dot plot” of Fed officials’ interest rate projections that reinforced the central bank’s commitment to a “higher for longer” approach. The forecast showed the benchmark rate rising to 4.4 per cent by the end of this year before peaking at 4.6 per cent next year. The dot plot was much more hawkish than in June, when it was last updated. At the time, officials predicted the fed funds rate would reach just 3.4 per cent by the end of the year and 3.8 per cent in 2023, before declining in 2024.

If the Fed — with all of its access, resources, talent, and capital — can’t project interest rates out even a few months, how can anyone project the future with any level of accuracy? The answer is that they can’t.

Rather than trying to understand a system that is impervious to understanding, design organizations to respond no matter what state the system finds itself in. Analogizing to computer science, it’s about handling errors and exceptions gracefully, even though such states were never supposed to be reached in the first place.

Will Putin use nuclear weapons on Ukraine? Who knows (even Putin likely doesn’t!), but be prepared for all eventualities. Will the Fed raise rates by 0bps or 400bps? Contingency plan and react when it finally makes a decision.

In computer science, chaos engineering is testing systems by systematically breaking things and seeing what happens, with the goal of identifying faults in systems and fixing them before a real problem arises. We all need to chaos engineer ourselves, our own organizations and our societies, breaking things and building them up stronger for whatever future might arrive.

“Securities” podcast: How to recapitalize America’s bloated defense industrial base

America spends more money on defense than ever, with a budget that more than doubled in the wake of the September 11th terrorist attacks. Despite those dizzying appropriations though, America is now down to just five big prime defense contractors, consolidated from dozens during the Cold War. It’s a pattern of consolidation, higher costs, and diminished resiliency and flexibility seen in sectors throughout the U.S. economy. As the stakes have grown for American defense, however, there are increasingly acute concerns about what our diminished capacity for defense means for the country’s long-term security.

I hosted James “Hondo” Geurts to talk more about the crisis in the defense industrial base. Geurts most recently performed the duties of the Under Secretary of the Navy and formerly was Assistant Secretary of the Navy for Research, Development & Acquisition, in charge of the Navy’s more than $100 billion annual budget.

Lux Capital is also deeply interested in recapitalizing the future of the defense industrial base and bringing more attention to emerging threats across all theaters of defense, including cyber, space, networking, and more in both state and civilian contexts. To that end, Hondo and Lux teamed up to create the Lux Security + Tech Index built on our portfolio company Thematic to offer investors a more methodical way to invest in the future of security. Steve Carpenter, CEO and founder of Thematic, joined as well.

In this episode, we talked about the importance of the new “industrial network” era of defense, the consequences of the 1990s post-Cold War peace dividend, how large projects like the F-35 and the Gerald R. Ford aircraft carrier drive consolidation, the value of simplicity in defense acquisition, the failure of the Defense Department’s high research expenditures, the need to shift from “program of record” to “capability of need”, and finally, details about the new Lux Security + Tech Index and its construction.

🔊 Listen to the episode here

As a reminder, all investments are risky, and nothing in this episode should be construed as an endorsement of any specific investment product for any individual listener. Always do your own research.

Lux Recommends

Our scientist-in-residence Sam Arbesman recommends Erik Hoel’s essay on “How AI’s critics end up serving the AIs”. "In response to these new AI models there is a rising class of AI critics, who downplay the results by focusing on terminology, fuzzy benchmarks, and minimization—and if the tide takes down your sandcastle, just build it further up the beach.”
In the same vein, Sam also recommends swyx's essay “Multiverse, not Metaverse,” which connects to Sam's piece on “Alternate Histories and GPT-3”.
Peter Hébert strongly recommends the new Ken Burns documentary The U.S. and the Holocaust and particularly its first episode, The Golden Door, on how America closed its borders in the 1930s in spite of the looming Nazi threat.
If you didn’t catch it from the column above, I recommend James M. Banner Jr.’s “All History Is Revisionist History” for an interesting take on a complex intellectual subject that threads the needle on one of the history profession's major debates right now.
Finally, Sam recommends “‘Gifs are cringe’: how Giphy’s multimillion-dollar business fell out of fashion” from The Guardian as well as a fun piece by Leslie Jamison in The New Yorker on The Enduring Allure of Choose Your Own Adventure Books.

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.

‍

September 17, 2022

Warming Hyperthreat

<- Previous

continue

reading

October 1, 2022

Sisyphean Maintenance

Next ->