Applications close in one week! If you (or someone you know) has what it takes to run an ambitious research program, we want to hear from you.

Feet! #inktober

Applications for the 2026 cohort of the Brains Research Accelerator are now open! If you have a deep technical background and an ambitious research idea that's beyond the scope of a startup or academic lab, you should apply! spec.tech/brains

We realized that (to our knowledge) nobody had unpacked where the money for “misfit research” — work that is a poor fit for academia, startups, or large companies — was coming from. So I wrote a short but hopefully comprehensive overview. blog.spec.tech/p/who-funds-...

It's funny how at meetings in DC I'm the whippersnapper and in SF I'm basically a walking corpse.

I do think AI could increase scientific productivity. But so much of science isn't easily mechanized tasks like pipetting -- it's adjusting optics, troubleshooting custom equipment, and keeping critters from dying. Here's a plan to capture and use this "tacit knowledge." ifp.org/teaching-ai-...

We finished building the Millennium Falcon! #caturday

Sometimes I wonder if the thing that makes startups possibly the most dynamic modern institution is just because it's the one place in today's world where it's normalized for young/undercredentialed people to be in charge of organizations and significant resources.

(Re)published a piece on the Spectech Blog about how efficiency warps the types of ideas that get support, which ones get killed too early, and which ideas people choose to pursue in the first place. blog.spec.tech/p/efficiency...

For the historians out there -- There are some hardcore, dig-into-the-undigitized-archives-and-try-to-piece-things-together-from-account-books history research projects about the financial history of technology that nobody is tackling: blog.spec.tech/p/some-open-...

For some reason I get a lot of joy out of the flight path from Helsinki to Tokyo

A lot of new thoughts about: - State capacity - Scale dependence for governments - The difference between governments and people of a state - American LARPing That are probably too politically-charged and out of my lane for the internet.

Ok! Time to stop procrastinating. You’ve done harder things than applying to AI Brains. Knock it out this weekend. Applications close Monday June 16th!

😭

Don't forget! AI Brains Q&A session today at 1p ET. You bring questions, we'll bring answers.

I’m noticing that there’s a funny motte-and-bailey attribute of most disciplines that I think are BS: If done really really well, they’re super insightful and valuable; it’s just that the vast majority of ideas are garbage and practitioners are charlatans.

This is today at 8p ET! Q+A session about the impending AI cohort of the Brains Research Accelerator. We'll talk about anything: how it works, who should apply, our theory of change, etc.

This is the kind of not-incremental metascientific experiment I want to see more of!

Testing gentle ways of reminding people in group chats to clean tracking information off of links

I've always been a bit annoyed by the advice that goes "What would you do if you had a year to live? Why aren't you doing that now?" Obv there are some things you shouldn't delay, but there are some things that take more than a year no matter how much effort you slam into them.

If you have questions, we have answers! We are doing two Q+A sessions about AI Brains in the coming weeks: 1. June 3 at 8p ET: lu.ma/eli2c08p 2. June 9 at 1p ET: lu.ma/vxx85ug5

Some proposals to replace metascience: Epistotechnics Technogenesisics Epistodynamics

I finally had a chance to search for "wirecutter wirecutter"

It's funny that there are two complete opposite arguments for where to locate an ambitious thing: "It should be where the good people want to be, otherwise you won't get good people" "It should be in the middle of nowhere because then only the really committed people will join"

Although curmudgeonly, the actual scientists who do takedowns of hype-y things have a great record I think the Heinlein quote holds — scientists saying “that is impossible” is often wrong. But “this actually existing work is not as exciting as you think it is” is often right.

Sometimes I'm tempted to just send this gif in follow-up emails instead of the nice polite "Hey just wanted to make sure this didn't get lost in your inbox..."

Has anybody ever used a CRM that actually sparks Joy?

Applications are now open for a special cohort of the Brains Accelerator targeting ambitious AI research programs, with a special focus on security and governance capabilities! spec.tech/ai-brains 1/

My first tariff-related sticker shock: $5 wing nuts 😬

Honest question about a fusion drive like this: where does the activation energy come from? If the majority of the energy from the fusion energy is blasting out the back, how do you sustain the reaction? newatlas.com/space/pulsar...

Everything about this is a pile of horrific unforced errors. I’d really advocate for step one being less foot-shooting and more cultures of competence regardless of your political bent. nymag.com/intelligence...

"The problem with most people's writing is located between their ears" Lots of people think they need to get better at writing, but nobody thinks they need to get better at thinking, and this is why they don't get better at writing. Crucial observation from @mastroianni.bsky.social

Good historical fiction feels a bit like fan fiction, but for history. Lots of "oh yeah! I got that obscure reference!"

I worry that most of the focus on "AI for science" is getting it backwards. Try to do great science and build awesome technology. Build AI tools when it would be useful. Setting out to do "AI for science" will lead to all sorts of lamplight effects and square peg round holes.

This attitude of "doing it right no matter what it takes" is what separates good work from great work. (From @antonhowes.bsky.social's recent excellent piece)

I find trees splitting rocks inspirational. Long game. (From a hike this past weekend)

You don't want to know my algorithm for whether to bias towards taking someone seriously or not.

Make starting a factory or a research lab like opening a restaurant.

We (mostly @adammarblestone.bsky.social and Mary with the help of the amazing scientists in our network) made a map of “fundamental development” gaps in science and technology. Go explore gap-map.org.

Honest question: are there examples of companies whose products or services got better when they were acquired by private equity?

The explosion of acronyms (three letters or otherwise) is another relatively-recent-but-feels-like-it-has-existed-forever phenomenon that we can chalk up to WWII.

Spectech is hiring a program associate for the Brains Research Accelerator! Are you/do you know someone relatively early in their career who has: - Technical training - Intense attention to detail - An interest in ambitious research programs? If so, please get in touch. spec.tech/jobs

An LLM agent that tracks what I type and creates a little popup whenever I write something factually false.

Colossal / Boring Company partnership to genetically engineer earthworms into tunnel boring sandworms

Functional humanoid robots will be awesome, but not in manufacturing. I finally did an analysis of - What it takes for a humanoid to match the cost of human labor - What that will do to the price of goods - How it compares to other, specialized forms of automation blog.spec.tech/p/humanoid-r...

April 1st on the internet is sadly no longer fun :-/ There are already so many absurd claims and over-the-top rhetoric that it just makes the adversarial information environment even more adversarial.

April 1st on the internet is sadly no longer fun :-/ There are already so many absurd claims and over-the-top rhetoric that it just makes the adversarial information environment even more adversarial.

Looking at a DoD slide deck from 1996. The quality of graphics hasn't improved in 30 years. If anything, it's gotten worse.