Director, Tech Center @heritage / infonomena substack com / @entropyecon
Joined February 2009
- Tweets13.1K
- Following3.8K
- Followers4.9K
- Likes18.9K
Bret Swanson retweeted
The WSJ previously reported that you, Jack Keane, and Lindsay Graham were the three primary domestic activists who coordinated to convince Trump that the war was a great idea against the advice of most of his cabinet, agreed. It is true that none of you work on the inside for Trump. And yes, it was his decision, agreed. And we're all now living through the political results of that decision.
WSJ: "To make the case, Graham said he worked closely with two other men—retired Army Gen. Jack Keane, a Fox News contributor, and Marc Thiessen, onetime chief speechwriter for former President George W. Bush. The trio took turns calling the president, he said, and compared notes. Graham said “there were not a lot of other voices” advancing the argument for action."archive.ph/YLlgF
Bret Swanson retweeted
Evidence is growing that we’re turning the corner on productivity.
US productivity grew at a 3.6% annual rate this past quarter.
That’s more than double the 1.5% average from 2007-2019.
The new numbers put me further ahead for my bet with Bob Gordon about productivity growth in the 2020s. 💪
Here’s how far the US is now ahead of the CBO estimate at the start of the decade (before they knew about the pandemic or LLMs) courtesy of @jasonfurman’s
Econvitals.
Bret Swanson retweeted
Practically no one read this when I published it a couple of months ago but I'm going to immodestly propose that everyone who wants to build a data center or whose livelihood depends on them getting built should read it now:
wsj.com/tech/ai/data-center-…
Bret Swanson retweeted
Re: Texas governor election. Let's take a deep breath.
If companies ask for no subsidies, promise a ton of union jobs, and guarantee that water and electricity costs won't rise/plus add supply, data centers can be built under her governorship.
This is directly from Gina's website:
Bret Swanson retweeted
What the 2026 Elections Mean for AI, Robotics & Data Centers: 5 Trends to Watch
adamthierer.substack.com/p/w…
The likely and lamentable political turn against AI data centers in some states (Texas) presents a gigantic, even historic, opportunity for other states (Indiana) who welcome AI investment. We’re talking hundreds of billions of dollars, which could remake a state’s economy.
Capital flows where it’s welcome.
Outer space 🪐 is another winner.
Unreal. @elonmusk is the BEST of the WEST.
Scott Galloway calls Elon Musk a "traitor to the West" and invokes the executions of the Rosenbergs after WW2.
"We lined up the Rosenbergs against a wall. This is a traitor to the West, and he should be held to the same standards as every other traitor in the West."
*Applause*
These people are absolutely unhinged.
“SpaceX has even revised its estimate up to ~13GW in 2027. This would be the largest amount of datacenter capacity ever built by any one company in that timeframe, roughly equivalent to all of Google’s datacenter capacity from 1999-2026 in just 1.5 years.”
Elon Is Doing It Again: SpaceX Will Be The World’s Biggest AI Cloud
Making 13GW in 1.5 Years A Reality
newsletter.semianalysis.com/…
Bret Swanson retweeted
‘His gift to the Nobel museum was the Roman oil lamp that he had acquired to estimate the amount of labor that was necessary to produce an hour of light two thousand years ago.’ Phenomenal @TedNordhaus tribute. RIP to a true giant.
open.substack.com/pub/thebre…
One of my favorites:
Thank you Uncle Bill, for all that you did for the world, and for me. breakthroughjournal.org/p/th…
Bret Swanson retweeted
Elon Musk just gave a massive update on Colossus 2 in Memphis, and the scale of what's being built is absolutely insane
Two buildings already packed with GB300s, two more are coming online within weeks, and it is heading toward roughly 860,000 NVIDIA GB300s
Here's the breakdown:
• Macroharder: 220,000 GB300s, 800-gig NICs, configured for training
• Minihard: 220,000 GB300s, same training configuration
• Gun Range: 220,000 GB300s, tracking to be operational by the END OF NOVEMBER
• Spam: ~200,000 GB300s, configured for inference, targeting DECEMBER
That's roughly 860,000 GB300s across these four buildings
Spam is designed for inference, which needs much simpler networking and less than HALF the fiber of the training buildings. And yes, the building once housed Hormel products
Elon personally walked the entire Colossus 2 site around midnight and again the next morning to check progress
Building by building, SpaceXSI is assembling SI compute at a speed no one else on Earth can match
Macroharder and Minihard both have 220k GB300s with 800 gig nics configured for training. Gun Range will have the same. Spam (used to house Hormel products) will have ~200k GB300s, but configured for inference, so much simpler networking and less than half the fiber.
I walked the whole Colossus 2 site around midnight and again this morning.
Good progress by the SpaceX team and contractors. Gun Range is tracking to be operational by end of November and Spam in December.
Bret Swanson retweeted
Two key facts that are often missed in reporting on ACA premium increases:
1. The taxpayer pays nearly the entire premium increase for subsidized enrollees.
2. Nearly 90 percent of enrollees are subsidized.
Affordable Care Act plan premiums will rise by 13% in 2027, the second straight year of double-digit increases in a strained market that is already facing shrinking enrollment. on.wsj.com/4jQmzqC
Bret Swanson retweeted
Excerpts:
1. On the limits of alignment research:
"you will not find a technical solution to politics and power"
2. On OpenAI's Navier-Stokes result:
"a famous unsolved problem solved with 130B tokens is the equivalent of a single human thinking for a couple thousand years."
3. On who knows what first:
"we're already in the world of 'a 25 year old on a beanbag in SF knows about crucial encryption developments before the head of CIA'."
4. On Europe:
"Europe will have neither AI nor the power infrastructure. But it will have a lot of columns about AI being fake..."
5. On AI-safety people who won't take political risks:
"Dude, YOU are the one talking about the 'entire future lightcone' and you don't have the balls to annoy some academics on a committee!?!"
"Courage and sacrifice win followers in ways intelligence does not."
Amazing 70k word substack from Dom: AI and Power
dominiccummings.substack.com…
... it’s useful to take a wide look at the context, to see how many questions connect, before ploughing into lots of detail. Much of the news is extremely hard to make sense of even for tech Insiders — so there’s these people call ‘Sam’ and ‘Dario’ who think they’re close to building super-intelligence then putting it in robots, and they apparently think there’s a 20% chance it kills everyone, and they’re accelerating their plans because they don’t trust each other, and some of their opponents claim this is a marketing campaign, and the governments have mostly ignored it all and have no Actual Plan?!
... I’m one of a tiny number of people who’ve had senior roles in politics AND has followed the AI debate since ~2000, listened to three physicists about AI before 2016 and took it seriously, hired AI experts to do AI on two historic elections (2016 and 2019) when this seemed kooky in politics and in 2019-20 helped create an AI office in No10, as part of a general push to make AI and robotics a central strategic priority for the British state and No10, when this idea seemed ludicrous to ~100% of the ‘sensible mainstream’.
... It’s normal to listen to some bigshot talking about AI and someone from a lab will say, sometimes with a pitying smile, ‘they just don’t feel the AGI’: i.e they don’t really grasp how fast change is happening and will accelerate because it’s exponential.
My friend physicist Steve Hsu has made a documentary, Machine God, coming out soon (check out the trailer, and his blog for details). Steve also was boss of the PRC side in the AI wargame organised by Marc Warner (film here). Steve wanted to capture the feel of the Bay Area in this extraordinary time. Here, the people who feel the AGI talk about ‘the future lightcone’.
This is one of the best substantive policies … and most amazing feel good results of American free enterprise … and redemptive reassurance that our leadership class is not totally lost. @MichaelDell @altcap
Robert and I are thrilled that our gift of over 2 million shares of SpaceX stock will be distributed to children’s Trump accounts in December!
Our initial gift (shares transferring to escrow as we speak) will go to 11 to 17-year-olds who live in zip codes with median family income (MFI) of $57,000 or less, and to those who live in zip codes with too few residents for an MFI to be reported in the census. These families should receive a notice by Monday.
In a few weeks, an additional gift will be finalized for eligible 11 to 17-year-old children in military families serving overseas, children in foster care, and more children whose MFI is $90,000 or less living in counties near our home in the great state of Texas.
Thank you to the incredible team @SpaceX, @Treasury, @IRS, @DellFdn, @GreaterHorizons, @InvestAmerica24, and the legal firms of Morgan Lewis and Gibson Dunn for working over the past few months to make this happen.
We are deeply grateful for the groundbreaking work that Brad Gerstner @altcap and @MichaelDell have done for the long-term financial security for millions of kids.
God Bless America!
💯👏
The Knowledge Problem in Public Health
In 1945, economist Friedrich Hayek identified a fundamental weakness in centralized economic planning: the knowledge necessary to make sound decisions is dispersed throughout society. No central authority, however intelligent or well-intentioned, can fully possess the information held by millions of individuals responding to their particular circumstances.
The same problem arises in public health. Knowledge about disease, clinical outcomes, patient behavior, environmental conditions, and local vulnerabilities is distributed among physicians, patients, scientists, laboratories, hospitals, and communities. National public health institutions contribute essential scientific expertise and coordination, but they cannot possess all the knowledge required to determine the optimal intervention for every population under every circumstance.
Centralization creates an additional danger. When a mistaken scientific assumption is implemented independently in one community, its consequences may be limited, and the experience of other communities may expose the error. When that same assumption is embedded in a uniform national policy, the resulting error can become systemic. The institutional mechanisms that ordinarily permit discovery through variation, comparison, and correction may be weakened precisely when they are most needed.
This suggests a fundamental principle of public health governance: "the greater the uncertainty surrounding an intervention, the greater the value of preserving independent investigation, local adaptation, and opportunities for comparative learning."
This principle does not eliminate the need for national coordination. Infectious diseases cross jurisdictional boundaries, and reliable surveillance, scientific standards, and emergency resource mobilization require collective action. Rather, it argues for distinguishing coordination from uniformity. A national public health system should coordinate the acquisition and exchange of knowledge without unnecessarily requiring identical decisions wherever legitimate uncertainty exists.
The objective should not be fifty independent public health systems operating in isolation, nor one centralized system making every consequential decision. It should be a scientifically integrated but operationally adaptive system that preserves institutional diversity, measures outcomes, identifies errors early, and rapidly incorporates new knowledge.
The greatest danger is not simply that a public health authority may be wrong. All scientific institutions are sometimes wrong. The greater danger is constructing a system in which one institution's error becomes everyone's error, while the mechanisms needed to discover that error have been suppressed.
Bret Swanson retweeted
I joined @GoogleDeepMind to develop a research program on the economics of AGI. Today, together with Andrew Koh (@andrewjkoh) and Shane Legg (@ShaneLegg), we are incredibly excited to release The Economics of Intelligence, the first in a series of projects from our group on how AGI will change the economy.
How do capability improvements transform the economy, and when? In this project, we define thresholds of economic capabilities, e.g., minimal AGI, based on how they transform the economy. We then track how the economy–market structure, employment, and wages–changes during the transition across capability regimes.
We have seen exponential increases in what AI agents can do but the broader economy seems almost remarkably unchanged. Part of the reason is that the solo AI agent is the wrong economic object; economic transformation is an equilibrium outcome where organizations compete. Looking at how intelligence functions within organizations, transformation will depend on when two key capability thresholds are crossed: competence (gains on individual skills, e.g., coding, math) and coordination (gains on ability of agents to cooperate and work together).
We show that what economic transformation looks like depends on whether AI becomes supercompetent, supercooperative, or both—the regime of artificial superintelligence (ASI). Tracking this transition shows why the real visible transformation may come from improvements in coordination—a capability most benchmarks don’t yet track. Past a certain threshold, AI capabilities can break the Coasian ceiling on firm size and lead to an economy that looks drastically different than the one we see today.
Two main takeaways:
1. To understand the path of transformation, we should be tracking agentic coordination more closely.
2. The impact of AI will likely show up in wages before market structure or employment.
We are really excited about this paper. Excited to hear people’s thoughts.
Link: institute.deepmind.com/essay…
