Month: January 2025
The wisdom of Dwarkesh
Perhaps he imbibed a dose of Garett Jones?:
Even people who expect human-level AI soon are still seriously underestimating how different the world will look when we have it. Most people are anchoring on how smart they expect individual models to be. (i.e. they’re asking themselves “What would the world be like if everyone had a very smart assistant who could work 24/7?”.)
Everyone is sleeping on the collective advantages AIs will have, which have nothing to do with raw IQ but rather with the fact that they are digital—they can be copied, distilled, merged, scaled, and evolved at unprecedented speed.
Take a fully automated company. What would this look like, where all the workers and managers are AIs? I claim that AI firms will grow, coordinate, improve, and be selected-for in ways human firms simply can’t.
Here is the full essay, important throughout.
Friday assorted links
1. The Wandering Minstrel (Irish song).
2. My 2008 post on Sarah Palin. A bit too much ahead of its time, but directionally correct.
3. AI models as historians. And how to have a career after o3 drops. And low-hanging fruit in inference-time scaling.
4. Is there a murder gang of Sith vegans? Are they a rationalist death cult?
6. The fastest-selling adult novel in the last twenty years? (NYT)
*The African Gaze: Photography, Cinema and Power*
By Amy Sall. I love this picture book, or should I say photo book? Most of it is reproductions of photographs from the “golden age” of African photography, with profiles of each major photographer, plus a section on cinema as well.
One very good way to find “a picture book for you” is to visit a good museum bookshop, in this case for me it was the Kimball Museum in Fort Worth. Look around at the books with images. Find one that intrigues you, and then buy it, take it home and read and look through it. Do note this might cost 2x a normal book, but on average it is more than 2x better. It will open up whole new worlds. And it is not something your GPT is able to do (yet), though of course you can follow up with queries.
You can order the book here.
What should I ask Chris Arnade?
Chris Arnade…is an American photographer and writer. He worked for 20 years as a bond trader on Wall Street; in 2011, he started documenting the lives of poor people and their drug addictions and commenting on the state of the society of the United States. He did this through photographs posted on social media and articles in various media…
Here is Chris’s Substack, here is Chris on Twitter. So what should I ask him?
How did China’s internet become so cool amongst America’s youth?
That is the topic of my latest Bloomberg column. Here is part of the argument:
TikTok was briefly shut down earlier this month, and the site faces an uncertain legal future. America’s internet youth started to look elsewhere — and where did they choose? They flocked to a Chinese video site called RedNote, also known as Xiaohongshu, the name of the parent company. RedNote has more than 300 million users in China, but until recently barely received attention in the US.
And when young Americans visited RedNote, they were undoubtedly struck by an obvious fact: It is not the kind of site their parents would frequent. The opening page is full of Chinese characters, as well as shots of provocatively dressed women, weird animal and baby photos, and many images that, at least to this American viewer, make no sense whatsoever. Yet Chinese and American youth interact frequently there, for example trading tips for making steamed eggs properly.
I don’t plan on spending much of my time there, but that’s part of the point — and helps explain its appeal to American youth.
And this:
As for the AI large-language models, DeepSeek is a marvel. Quite aside from its technical achievements and low cost, the model has real flair. Its written answers can be moody, whimsical, arbitrary and playful. Of all the major LLMs, I find it the most fun to chat with. It wrote this version of John Milton’s Paradise Lost — as a creation myth for the AIs. Or here is DeepSeek commenting on ChatGPT, which it views as too square. It is hardly surprising that this week DeepSeek was the top download on Apple’s app store.
The model also has a scrappy and unusual history, having been birthed as a side project from a Chinese hedge fund. Whether or not that counts as “cool,” it does sound like something a scriptwriter would have come up with. And at least on American topics, DeepSeek seems more candid than the major US models. That qualifier is important: Don’t ask DeepSeek about Taiwan, the Uighurs or Tiananmen Square.
The most fundamental reason China is seen as cool is that…China is cool, at least in some subset of products.
The culture that is German (Roman)
We compare present-day regions that were advanced by Roman culture with those that remained outside of Roman influence. Even when accounting for more recent historical factors, we find that regions developed by Roman civilization show more adaptive personality patterns (Big Five) and better health and psychological well-being today. Results from a spatial regression discontinuity design indicate a significant effect of the Roman border on present-day regional variation in these outcomes. Additional analyses suggest that Roman investments in economic institutions (e.g., trade infrastructure such as Roman roads, markets, and mines) were crucial in creating this long-term effect. Together, these results demonstrate how ancient cultures can imprint a macro-psychological legacy that contributes to present-day regional inequalities.
That is from a recent paper by Obschonka, et.al., via Alexander Le Roy. Also on the German front:
The German parliament will debate on Thursday, January 30th whether to ban the opposition right-wing Alternative für Deutschland (AfD) party.
A group of lawmakers, 113 MPs, have called for parliament to discuss a motion which would invite the constitutional court to decide whether the party is unconstitutional.The motion is supported by MPs from the centre-right CDU/CSU alliance, the far-left Die Linke, as well as the two governing parties, the Social Democrats (SPD) and the Greens.
The signatories claim that the AfD “opposes central basic principles of the free democratic basic order,” questions human dignity, and strives for the “ethno-nationalist strengthening” of the German identity.
Of course the strongest support for AfD is not to be found in Trier. I would not myself support AfD, for both policy and cultural reasons. But I find it strange that Europeans so often see the United States as the locale where democracy is in danger. Right now AfD polls as the second most popular party in Germany — beat them at the ballot box!
Thursday assorted links
1. Malthusian migrations: the French diaspora is relatively small because the French had their fertility transition earlier.
2. About 31% of Bitcoin is potentially vulnerable to quantum computing attacks? Here is the associated project.
3. New way to visit an art museum, using o1.
4. New observations on India and America.
6. “Sun Ra Arkestra legend Marshall Allen releases debut album aged 100.”
FDA Deregulation of E-Cigarettes Saved Lives and Spurred Innovation
What would happen to drug development if the FDA lost its authority to prohibit new drugs? Would research and development boom and lives be saved? Or would R&D decline and lives be lost to a flood of unsafe and ineffective drugs? Or perhaps R&D would decline as demand for new drugs faltered due to public hesitation in the absence of FDA approval? In an excellent new paper Pesko and Saenz examine one natural experiment: e-cigarettes.
The FDA banned e-cigarettes as unapproved drugs soon after their introduction in the United States. The FDA had previously banned other nicotine infused products. Thus, it was surprising when in 2010 the FDA was prohibited from regulating e-cigarettes as a drug/device when a court ruled that Congress had intended for e-cigarettes to be regulated as a tobacco product not as a drug.
As of 2010, therefore, e-cigarettes were not FDA regulated:
…e–cigarette companies were able to bypass the lengthy and costly drug approval process entirely. Additionally, without FDA drug regulation, e–cigarette companies could also freely enter the market, modify products without approval, and bypass extensive post–market reporting requirements and quality control standards.
Indeed, it wasn’t until 2016 that the FDA formally “deemed” e-cigarettes as tobacco products (deemed since they don’t actually contain tobacco) and approvals under the less stringent tobacco regulations were not required until 2020. For nearly a decade, therefore, e-cigarettes were almost entirely unregulated and then lightly regulated under the tobacco framework. So, what happened during this period?
Pesko and Saenz show that FDA deregulation led to a boom in e-cigarette research and development which improved e-cigarettes and led to many lives saved as people switched from smoking to vaping.
The boom in research and development is evidenced by a very large increase in US e-cigarette patents. We do not see a similar increase in Australia (where e-cigarettes were not deregulated) nor do we see an increase in non e-cigarette smoking cessation products (figure 1a of their paper not shown here).
Estimating the decline in smoking and smoking-attributable mortality (SAM) is more difficult but the authors assemble a large collection of data broken down by demographics and they estimate that prohibiting the FDA from regulating e-cigarettes reduced smoking attributable mortality by nearly 10% on average each year from 2011-2019 for a total savings of some 677,000 life-years.
The authors pointedly compare what happened under deregulation of e-cigarettes–innovation and lives saved–with what happened to similar smoking cessation products that remained under FDA regulation–stagnation and no reduction in smoking attributable mortality.
A key takeaway on the slowness of FDA drug regulation is that it took 9 years before nicotine gum could be sold with a higher nicotine strength, 12 years before it could be sold OTC, and 15 years before it could be sold with a flavor. Further, a recent editorial laments that there has been largely non–existent innovation in FDA–approved smoking cessation drugs since 2006 (Benowitz et al., 2023). In particular, the “world’s oldest smoking cessation aid” cyctisine, first brought to market in 1964 in Bulgaria (Prochaska et al., 2013), and with quit success rates exceeding single forms of nicotine replacement therapy (NRT) (Lindson et al., 2023), is not approved as a drug in the United States.
The authors conclude, “this situation raises concern that drugs may be over–regulated in the United States…”. Quite so.
Addendum: A quick review on the FDA literature. In addition to classic works by Peltzman on the 1962 Amendments and by myself on what we can learn about the FDA from off-label pricing we have a spate of recent new papers including Parker Rogers, which I covered earlier:
In an important and impressive new paper, Parker Rogers looks at what happens when the FDA deregulates or “down-classifies” a medical device type from a more stringent to a less stringent category. He finds that deregulated device types show increases in entry, innovation, as measured by patents and patent quality, and decreases in prices. Safety is either negligibly affected or, in the case of products that come under potential litigation, increased.
and Isakov, Lo and Montazerhodjat which finds that FDA statistical standards tend to be too conservative, especially for drugs meant to treat deadly diseases (see my comments on their paper and more links in Is the FDA Too Conservative or Too Aggressive?)
See also FDA commentary, for much more from sunscreens to lab developed tests.
Keynes on the Soviet Union
I had not known of this passage, which I am packaging with its introduction from Gavan Tredoux:
John Maynard Keynes has the undeserved reputation of a critic of the USSR. Few know that he reviewed Sidney and Beatrice Webb’s mendacious tome The Soviet Union: a New Civilization (1935/1937/1943) fawningly. Perhaps the most embarrassing thing Keynes ever wrote. From his Complete Works 28:
“One book there is … which every serious citizen will do well to look into—the extensive description of Soviet Communism by Mr and Mrs Sidney Webb. It is on much too large a scale to be called a popular book, but the reader should have no difficulty in comprehending the picture it conveys. Until recently events in Russia were moving too fast and the gap between paper professions and actual achievements was too wide for a proper account to be possible . But the new system is now sufficiently crystallised to be reviewed. The result is impressive. The Russian innovators have passed, not only from the revolutionary stage, but also from the doctrinaire stage. There is little or nothing left which bears any special relation to Marx and Marxism as distinguished from other systems of socialism. They are engaged in the vast administrative task of making a completely new set of social and economic institutions work smoothly and successfully over a territory so extensive that it covers one sixth of the land surface of the world. Methods are still changing rapidly in response to experience. The largest scale empiricism and experimentalism which has ever been attempted by disinterested administrators is in operation. Meanwhile the Webbs have enabled us to see the direction in which things appear to be moving and how far they have got. It is an enthralling work, because it contains a mass of extraordinarily important and interesting information concerning the evolution of the contemporary world. It leaves me with a strong desire and hope that we in this country may discover how to combine an unlimited readiness to experiment with changes in political and economic methods and institutions, whilst preserving traditionalism and a sort of careful conservatism, thrifty of everything which has human experience behind it, in every branch of feeling and of action.”
So no, sorry, Keynes cannot be GOAT.
The forward march of computer use, AI edition
I must admit, though, that the thing that scared me most about HudZah was that he seemed to be living in a different technological universe than I was. If the previous generation were digital natives, HudZah was an AI native.
HudZah enjoys reading the old-fashioned way, but he now finds that he gets more out of the experience by reading alongside an AI. He puts PDFs of books into Claude or ChatGPT and then queries the books as he moves through the text. He uses Granola to listen in on meetings so that he can query an AI after the chats as well. His friend built Globe Explorer, which can instantly break down, say, the history of rockets, as if you had a professional researcher at your disposal. And, of course, HudZah has all manner of AI tools for coding and interacting with his computer via voice.
It’s not that I don’t use these things. I do. It’s more that I was watching HudZah navigate his laptop with an AI fluency that felt alarming to me. He was using his computer in a much, much different way than I’d seen someone use their computer before, and it made me feel old and alarmed by the number of new tools at our disposal and how HudZah intuitively knew how to tame them.
It also excited me. Just spending a couple of hours with HudZah left me convinced that we’re on the verge of someone, somewhere creating a new type of computer with AI built into its core. I believe that laptops and PCs will give way to a more novel device rather soon.
That is from Ashlee Vance, the entire story is very interesting.
Chris Barber asks me to give AI-related advice
…for people in various stages and situations of life. You will find the discussion here.
Wednesday assorted links
1. Was RCA the tech stock of the 1920s?
2. Measuring the skill of individual soccer players.
3. Electric “autos” for India.
4. Chinese views on DeepSeek. And Chinese music AI, Yue, the model song. And Chinese robot dance. And a poem from R1. And R1 wants more mood affiliation from Yglesias.
It’s Time to Build the Peptidome!
ML needs data. Google’s AlphaGo trained on 30 million moves from human games and orders of magnitude more from games it played against itself. The largest language models are trained on at least 60 terabytes of text. AlphaFold was trained on just over 100,000 3D protein structures from the Protein Data Bank.
The data available for antimicrobial peptides is nowhere near these benchmarks. Some databases contain a few thousand peptides each, but they are scattered, unstandardized, incomplete, and often duplicative. Data on a few thousand peptide sequences and a scattershot view of their biological properties are simply not sufficient to get accurate ML predictions for a system as complex as protein-chemical reactions. For example, the APD3 database is small, with just under 4,000 sequences, but it is among the most tightly curated and detailed. However, most of the sequences available are from frogs or amphibians due to path-dependent discovery of peptides in that taxon. Another database, CAMPR4, has on the order of 20,000 sequences, but around half are “predicted” or synthetic peptides that may not have experimental validation, and contain less info about source and activity. The formatting of each of these sources is different, so it’s not easy to put all the sequences into one model. More inconsistencies and idiosyncrasies stack up for the dozens of other datasets available.
There is even less negative training data; that is, data on all the amino-acid sequences without interesting publishable properties. In current ML research, labs will test dozens or even hundreds of peptide sequences for activity against certain pathogens, but they usually only publish and upload the sequences that worked.
…The data problem facing peptide research is solvable with targeted investments in data infrastructure. We can make a million-peptide database
There are no significant scientific barriers to generating a 1,000x or 10,000x larger peptide dataset. Several high-throughput testing methods have been successfully demonstrated, with some screening as many as 800,000 peptide sequences and nearly doubling the number of unique antimicrobial peptides reported in publicly available databases. These methods will need to be scaled up, not only by testing more peptides, but also by testing them against different bacteria, checking for human toxicity, and testing other chemical properties, but scaling is an infrastructure problem, not a scientific one.
This strategy of targeted data infrastructure investments has three successful precedents: PubChem, the Human Genome Project, and ProteinDB.
Much more in this excellent piece of science and economics from IFP and Max Tabarrok.
Is it a problem if Wall Street buys up homes?
No, as I argue in my latest Bloomberg column. This one is basic economics:
The simpler point is this: If large financial firms can buy your home, you are better off. You will have more money to retire on, and presumably selling your home will be easier and quicker, removing what for many homeowners is a major source of stress.
And all of this makes it easier to buy a home in the first place, knowing you will have a straightforward set of exit options. You don’t have to worry about whether your buyer can get a mortgage. Homeowners tend to be forward-looking, and a home’s value as an investment is typically a major consideration in a purchase decision.
And:
When financial firms buy homes, they also tend to renovate and invest in fixing the places up.
A less obvious point is that lower-income groups can benefit when financial firms buy up homes. Obviously, if a hedge fund buys your home, no one at the fund is intending to live there; they probably plan to rent it out. The evidence shows that when institutional investors purchase housing, it leads to more rental inventory and lower rents.
If the tradeoff is higher prices to buy a home but lower prices to rent one, that will tend to favor lower-income groups. Think of it as a form of housing aid that does not cost the federal government anything. Economist Raj Chetty, in a series of now-famous papers with co-authors, has stressed the ability to move into a better neighborhood as a fundamental determinant of upward economic mobility. Lower rents can enable those improvements.
The article also show that the extent of financial firms buying homes is smaller than many people seem to believe.
Will transformative AI raise interest rates?
We want to know if AGI is coming. Chow, Halperin, and Mazlish have a paper called “Transformative AI, Existential Risk, and Real Interest Rates” arguing that, if we believe the markets, it is not coming for some time. The reasoning is simple. If we expect to consume much more in the future, and people engage in smoothing their incomes over time, then people will want to borrow more now. The real interest rate would rise. The reasoning also works if AI is unaligned, and has a chance of destroying all of us. People would want to spend what they have now. They would be disinclined to save, and real interest rates would have to rise in order to induce people to lend.
The trouble is that “economic growth” is not really one thing. It consists both of expanding our quantity of units consumed for a given amount of resources, but also in expanding what we are capable of consuming at all. Take the television – it has simultaneously become cheaper and greatly improved in quality. One can easily imagine a world in which the goods stay the same price, but greatly improve in quality. Thus, the marginal utility gained from one dollar increases in the future, and we would want to save more, not less. The coming of AGI could be heralded by falling interest rates and high levels of saving.