When VCs brag about their portfolio companies going public, they don't mention when they invested — usually it was the last round before the IPO.
When @stephen_wolfram introduced me to @DeepFission all they had was a blackboard. Ringing the bell at NASDAQ just a few years later is the best proof that our Deep Future Invention Capital strategy is working.
Much more to come!
I believe @WalterIsaacson completely failed to understand Steve Jobs. Reading that biography made me realize this was probably true for his other subjects, who I didn’t know much about.
You don’t need his failed attempt to channel Steve Jobs on any modern technology. Try things out for yourself. Learn to use them. Have your own ideas and see if new technologies can help you bring them to life. Form your own informed opinion, and if the tech sucks, go bury it in the dirt and play with some rocks instead.
.@WalterIsaacson says Steve Jobs would be appalled by agentic AI.
"[Steve Jobs] says the goal is to connect human creativity to the processing power of our machines."
"What we're doing now with agentic AI is where I worry."
"When I start using agents that try to do things without me and to be out of my control, Steve Jobs would be appalled by that."
"It doesn't connect human creativity to the product. It doesn't have the music that makes our hearts sing, as Steve said, with the iPod."
"These people aren't stepping back and saying, 'Why are we making this?' The answer should be to help humanity."
"That was Steve's lodestar."
Creating or updating open standards has always been slowed by the cost for all parties to upgrade their software. For example, this is why we never got practical encryption for email.
Now, it is almost trivial to integrate anything for everyone.
We can finally make dramatically more secure and adaptable protocols for everything.
💡Blockchain based provenance for artists, authors, journalists. Imagine being able to see who posted anything first.
💡E2E encryption for email.
💡Web of trust so you know who you are talking to.
💡Decentralized social graph to keep track of who you know and care about.
💡Next Generation RSS to break our feeds out of walled gardens.
We are still running a lot of internet 1.0 shit that could be rearchitected. A multitude of protocols should be created and fight for adoption. Maybe analogous to what has happened with blockchains.
“If you show up acting like every other founder, you've literally just gotten a no.” — Chris Kantrowitz
xitter.cf/ChrisKantrowitz
deepfuture.tech/interview-th…
🟢 Spotify → open.spotify.com/episode/4Yw…
🍎 Apple Podcasts → podcasts.apple.com/us/podcas…
🔴 Pocket Casts → pocketcasts.com/podcast/deep…
Pablos retweeted
geoffrey hinton mass-quit google in 2023 and mass-told everyone he was scared of what he built. nobody asked the right question.
not "is AI dangerous." the right question is "why does backpropagation work at all."
hinton couldn't answer it. neither could lecun. neither could sutskever when he left openai. the three founders of deep learning built a $4.6 trillion industry on an algorithm none of them could mathematically prove should work.
until last month.
a physicist at ETH zurich published a 41-page proof that backpropagation is a special case of the second law of thermodynamics. not an analogy. a derivation.
the math:
thermodynamics: dS/dt ≥ 0. entropy of a closed system never decreases.
backpropagation: dL/dt ≤ 0. loss of a trained network never increases.
one is entropy going up. the other is loss going down. they are the same equation with a sign flip.
the proof shows that gradient descent on a neural network is mathematically identical to a physical system falling toward thermal equilibrium. every weight update is a molecule finding its lowest energy state.
this is why transformers scale. this is why chinchilla scaling laws look like thermodynamic equations. this is why nobody could prove convergence guarantees - they were using the wrong field. the answer was never in computer science. it was in statistical mechanics.
now watch what this means for the frontier labs [Fig. 1] -
openai trained GPT-4 at an estimated mass-energy cost of 50,000 MWh. anthropic trained claude 3.5 at approximately 31,000 MWh. google trained gemini ultra at over 60,000 MWh.
the ETH proof predicts a theoretical minimum energy for each model's capability level. the equation:
E_min = kT × N × ln(1/ε)
k = boltzmann constant. T = operating temperature. N = parameters. ε = target loss.
GPT-4's actual energy usage: 17x above the theoretical minimum.
claude: 12x above minimum.
gemini: 23x above minimum.
anthropic is closest to the thermodynamic floor. not by accident. dario amodei has 3 physicists on the training infrastructure team. openai has zero.
the companies that understand this equation will compress 3 years of scaling into 6 months. the companies that don't will spend $10 billion on training runs that a physicist could tell them are 20x less efficient than possible.
hinton understood the gradient. he never understood the entropy.
the network doesn't learn. it cools.
save this
Readers added context they thought people might want to know
The specific 41-page proof by an ETH Zurich physicist published last month does not exist. A 2025 paper explores how backpropagation emerges in physical models of neural networks: arxiv.org/abs/2503.09980 Training energy claims and the exact equivalence described are not supported by it.
Pablos retweeted
Replying to @BachFrancis
(1/7) I'd like to provide some additional context for the remarks that can be found in this interview.
(2/7) My discomfort with the attention being paid to existential risk---and with the ongoing dominance of the public discussion of such risks by a few so-called thought leaders---is that there's an elephant in the room that is being missed. That elephant is the effect of all of this discussion on young people, and indeed the effect---beginning with social media a decade ago---of information technology on young lives broadly speaking.
(3/7) I recently read an interview by Jill Lepore in which she reports on discussions that she had recently with a group of college students: "They had just three things to say. One, social media completely destroyed their childhoods. Two, AI is destroying their college education and their likely prospects for employment. And three, what the hell are we all doing?" To add a fourth bullet that I've heard in similar discussions: "I'm not going to have children, because the existential risk is just too high."
(4/7) Please contemplate such a statement, those of you who think that it's OK that the public, widely diffused, discussion coming from seemingly authoritative figures is so adamant about existential risk.
(5/7) Let me add that I have little interest in a debate about what "superintelligence" might mean, or whether we're currently seeing the emergence of superintelligence. For what it's worth, I would certainly agree that the solution of extremely challenging math problems (for example) exhibits a very impressive form of intelligence, although I would want to argue that such intelligence is impossible to dissociate from the collective of humans that provided data to the machine learning system (which itself has become a kind of collective).
(6/7) Such philosophical debates are entertaining and even important, but when they then head in the direction of "when such superintelligence emerges, there's a good chance that it will want to kill all humans", I say "science fiction". Again, I don't think that it's bad to consider such thoughts, but I strongly believe that it's bad for society, and especially young people, for such thoughts, unaccompanied by any serious real-world analysis, to become a dominant form of discourse of the era.
(7/7) We are damaging human lives, and the consequences of that damage will no doubt be a major theme of the next several decades.
Pablos retweeted
The one and only Michael Jordan was interviewed last week by the French newspaper Libération. Here is an English translation. Enjoy!
di.ens.fr/~fbach/MJordan_ITW…
I got an Apple ][ in 1979.
Alarmists in 1979:
“the microprocessor is being transformed into the means for destroying millions of people’s jobs and making the rest even more tedious and dehumanised than before.”
The microprocessor is probably the reason I haven’t had to spend a single dehumanizing day of my life tediously mining shit out of the ground that can burn.
marxists.org/archive/harman/…
In In 1979 Clive Jenkins and Barrie Sherman wrote The Collapse of Work, arguing that microelectronics was advancing so fast that a society with many millions of people permanently unemployed was both inevitable and imminent. Their follow-up, The Leisure Shock, warned of Britain facing five million unemployed by about 1990.
Surely they were just ahead of their time and these predictions will finally come to pass, now that we have more microelectronics.
I’m gonna need to see some verification that you are a real New Yorker.
nypost.com/2026/10/12/us-new…
Pablos retweeted
We all finish the rounds and put our hands up and the world just sees successes.
But make no mistake, raising for any deep tech company is really freakin hard and most don’t get through it without resorting to, let’s just say, a wider investor audience…
Replying to @beffjezos
Deep Tech founders shouldn't have to take the money of last resort like this.
Kudos to Isaiah for pushing through.
Lean is verifying a lot of hallucinated proofs, but WTF is it actually verifying? We are probably going to need an army of mathematicians to figure that out.
Math is universal in the universe. The problem is that WE don’t understand math. The goal of mathematics is to fix that.
If your AI generates a 600 page proof that no human can parse, but Lean verifies it, has mathematics really advanced?
It ain’t over until it’s over folks.
Don’t blame @stephen_wolfram for my musings, go read his essay.
writings.stephenwolfram.com/…
“We've accidentally cracked the code on regulatory.” — Chris Kantrowitz
xitter.cf/ChrisKantrowitz
deepfuture.tech/our-kpi-is-d…
🟢 Spotify → open.spotify.com/episode/2yW…
🍎 Apple Podcasts → podcasts.apple.com/us/podcas…
🔴 Pocket Casts → pocketcasts.com/podcast/deep…













