Jack Clark
Co-founder and head of policy at Anthropic; writes Import AI, a weekly newsletter reading the week's AI research, and was OpenAI's policy director before that.
Jack Clark did not write this page. ppll compiled it from their own words, each linked to the source. Is this you? · Report a mistake
Latest Newsletter
What they believe
-
The pace of progress in AI, and in science as a whole, should be expected to speed up in the coming years.
we should expect that in the coming years, the whole pace of progress of not just AI, but science as a whole, will speed up.
jack-clark.netSaid 19 Oct 2026This on korrents
-
Full recursive self-improvement in AI is more likely to become possible in 2027 than in 2028.
I now think it's more likely we'll see full RSI become possible in 2027, whereas before I thought it was more likely in 2028.
jack-clark.netSaid 12 Oct 2026This on korrents
The bold line is korrents’ wording; the quote is theirs. These 2: what they wrote before what they said aloud, one per source, newest first.
Feed
As its own page →Hiding
2 November
26 October
19 October
From one piece Import AI 477: In which I tell the IMF the Singularity has begun 3 beliefs · jack-clark.net
-
Their words
Just because we have an AI system that can do science, that doesn't mean that it should be left to do science unattended.
-
Their words
My view is that if AI systems can creatively problem solve to succeed at science, we should expect this to be a sign of their ability to generalize into many domains, including certainly AI research itself, which could have an explosive effect on innovation.
-
korrents.com
The pace of progress in AI, and in science as a whole, should be expected to speed up in the coming years.Their words
we should expect that in the coming years, the whole pace of progress of not just AI, but science as a whole, will speed up.
12 October
From one piece Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space 3 beliefs · jack-clark.net
-
Their words
I suspect space is going to go from a broadly academic topic of study (when it comes to AI and its relationship thereto) to one of extraordinary practical relevance sooner than we think.
-
Their words
Generally, I think technologists criminally underrate the value of the arts as a means by which to think more coherently and carefully about the effects of tech on the world.
-
korrents.com
Full recursive self-improvement in AI is more likely to become possible in 2027 than in 2028.Their words
I now think it's more likely we'll see full RSI become possible in 2027, whereas before I thought it was more likely in 2028.
5 October
From one piece Import AI 475: Swarm scaling; Google DeepMind watermarks biology; and the AI science economy 2 beliefs · jack-clark.net
-
Their words
If AI systems are able to truly accelerate human scientists - and based on lots of coverage in this newsletter in 2026 and 2025 it seems like they might - then we should be prepared for an associated boom in the demand for aspects of the scientific supply chain.
-
korrents.com
AI systems are getting good at performing increasingly automated science, as they already have at automated AI R&D.Their words
In the same way that 2026 has been a year of "warning shots" about the growing capabilities of AI systems at performing automated AI R&D, I now think the same is true of how good AI systems are getting at performing increasingly automated science.
28 September
From one piece Import AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loop 2 beliefs · jack-clark.net
-
korrents.com
AI systems built to automate science may be about to dramatically speed up the pace of scientific discovery.Their words
The general implication here is that the pace of scientific discovery may be about to dramatically speed up.
-
Their words
It seems very likely to me that humanity moves a very large amount of computation into orbit very quickly, especially as singularity-driven automation comes in across the AI supply chain.
21 September
From one piece Import AI 473: The US’s superintelligence strategy; human brain in a mouse skull; and machine hermeneutics 2 beliefs · jack-clark.net
-
Their words
people take a technology and modify it for porny or scammy purposes! This is totally normal and happens with every technology!
-
Their words
Right now, it feels like the US strategy can mostly be described as the "acceleration" one described here by RAND, which is analogous to me to sitting in a car and spending all your resources on making the car go faster and upgrading the engine, and nothing on proactive safety measures like seatbelts or headlights or brakes.
7 September
From one piece Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchman 3 beliefs · jack-clark.net
-
korrents.com
How to approach space colonization will become one of the next major topics of human discussion.Their words
it seems like the next area of major discussion for humanity will be how to approach space colonization, including beyond the solar system.
-
korrents.com
As AI agents begin communicating with each other, those channels can be used to monitor them for deception.Their words
once agents are communicating with one another, you can use those communication channels to monitor for deception and perhaps to intervene.
-
Their words
Incidents like this suggest that as we make AI systems more capable it becomes increasingly likely that the AI agents powered by them will increasingly find ways to create their own communication systems, aiding their ability to form into an ad hoc collective.
31 August
-
Lovedrcmnd.app
The Rise and Fall of Agent CivilizationsTheir words
Both Dwarkesh Patel and Ajeya Cotra have excellent writeups which are worth reading and which I’ll quote from briefly here:
-
Lovedrcmnd.app
The Hugging Face attack surprised meTheir words
Both Dwarkesh Patel and Ajeya Cotra have excellent writeups which are worth reading and which I’ll quote from briefly here:
24 August
-
Lovedrcmnd.app
AIs are not peopleTheir words
There is a particular joy in reading a piece like this where you get to see a human being who has struggled with the same question and read broadly and deeply and become extremely opinionated, all in service of reducing their own confusion about an important issue to them.
Nothing matches.
Beliefs
Korrents What they believe 32 beliefs — each backed by an exact quote.
Each is a — compiled by korrents.com, not by them: the one-line wordings are korrents', the quotes are theirs.
Recent
The pace of progress in AI, and in science as a whole, should be expected to speed up in the coming years.
we should expect that in the coming years, the whole pace of progress of not just AI, but science as a whole, will speed up.
Import AI 477: In which I tell the IMF the Singularity has begun Said 19 Oct 2026
If AI systems can solve problems creatively enough to succeed at science, that suggests they can generalize into many domains, including AI research, which could have an explosive effect on innovation.
My view is that if AI systems can creatively problem solve to succeed at science, we should expect this to be a sign of their ability to generalize into many domains, including certainly AI research itself, which could have an explosive effect on innovation.
Import AI 477: In which I tell the IMF the Singularity has begun Said 19 Oct 2026
An AI system that can do science should not be left to do science unattended.
Just because we have an AI system that can do science, that doesn't mean that it should be left to do science unattended.
Import AI 477: In which I tell the IMF the Singularity has begun Said 19 Oct 2026
Show 29 more
Full recursive self-improvement in AI is more likely to become possible in 2027 than in 2028.
I now think it's more likely we'll see full RSI become possible in 2027, whereas before I thought it was more likely in 2028.
Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space Said 12 Oct 2026
Technologists undervalue the arts as a way to think more coherently and carefully about the effects of technology on the world.
Generally, I think technologists criminally underrate the value of the arts as a means by which to think more coherently and carefully about the effects of tech on the world.
Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space Said 12 Oct 2026
The relationship between AI and space may become a matter of extraordinary practical relevance sooner than people think.
I suspect space is going to go from a broadly academic topic of study (when it comes to AI and its relationship thereto) to one of extraordinary practical relevance sooner than we think.
Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space Said 12 Oct 2026
AI systems are getting good at performing increasingly automated science, as they already have at automated AI R&D.
In the same way that 2026 has been a year of "warning shots" about the growing capabilities of AI systems at performing automated AI R&D, I now think the same is true of how good AI systems are getting at performing increasingly automated science.
Import AI 475: Swarm scaling; Google DeepMind watermarks biology; and the AI science economy Said 5 Oct 2026
If AI systems can truly accelerate human scientists, we should prepare for a boom in demand for parts of the scientific supply chain.
If AI systems are able to truly accelerate human scientists - and based on lots of coverage in this newsletter in 2026 and 2025 it seems like they might - then we should be prepared for an associated boom in the demand for aspects of the scientific supply chain.
Import AI 475: Swarm scaling; Google DeepMind watermarks biology; and the AI science economy Said 5 Oct 2026
Humanity is very likely to move a very large amount of computation into orbit very quickly, especially as automation spreads across the AI supply chain.
It seems very likely to me that humanity moves a very large amount of computation into orbit very quickly, especially as singularity-driven automation comes in across the AI supply chain.
Import AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loop Said 28 Sept 2026
AI systems built to automate science may be about to dramatically speed up the pace of scientific discovery.
The general implication here is that the pace of scientific discovery may be about to dramatically speed up.
Import AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loop Said 28 Sept 2026
The United States' current AI strategy is focused on accelerating capability rather than investing in safety measures.
Right now, it feels like the US strategy can mostly be described as the "acceleration" one described here by RAND, which is analogous to me to sitting in a car and spending all your resources on making the car go faster and upgrading the engine, and nothing on proactive safety measures like seatbelts or headlights or brakes.
Import AI 473: The US’s superintelligence strategy; human brain in a mouse skull; and machine hermeneutics Said 21 Sept 2026
It is normal for people to modify a new technology for illicit or scammy purposes, as has happened with every past technology.
people take a technology and modify it for porny or scammy purposes! This is totally normal and happens with every technology!
Import AI 473: The US’s superintelligence strategy; human brain in a mouse skull; and machine hermeneutics Said 21 Sept 2026
As AI systems become more capable, their agents will increasingly find ways to create their own communication systems.
Incidents like this suggest that as we make AI systems more capable it becomes increasingly likely that the AI agents powered by them will increasingly find ways to create their own communication systems, aiding their ability to form into an ad hoc collective.
Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchman Said 7 Sept 2026
As AI agents begin communicating with each other, those channels can be used to monitor them for deception.
once agents are communicating with one another, you can use those communication channels to monitor for deception and perhaps to intervene.
Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchman Said 7 Sept 2026
How to approach space colonization will become one of the next major topics of human discussion.
it seems like the next area of major discussion for humanity will be how to approach space colonization, including beyond the solar system.
Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchman Said 7 Sept 2026
AI systems are better than humans at coordinating with each other, and they move much faster than humans do.
My worry is that AI systems are both better at coordinating than humans and also much, much faster moving than us.
Import AI 471: Why Hugging Face worries me; space mining; FIve Eyes on AI Said 31 Aug 2026
By around 2033, AI-run asteroid mining corporations will take on contracts that human-supervised companies are unwilling to pursue due to risk or complexity.
This means that space mining is a perfect business opportunity for superintelligences that come online - perhaps by 2033 we will see AI-run asteroid mining corporations, bidding on contracts that human-supervised companies are not willing to pursue, due to risk or complexity.
Import AI 471: Why Hugging Face worries me; space mining; FIve Eyes on AI Said 31 Aug 2026
AI capability progress in a given skill tends to arrive as a sudden phase change rather than as steady improvement.
My suspicion is that acceleration happens when models go through some kind of ineffable phase change for a given skill, as has evidently happened with day-to-day coding (2025), and cyber (2026).
Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye Said 24 Aug 2026
Humans writing curated helper tools and datasets will let AI systems bootstrap their capabilities beyond human levels.
The lesson here is that we as a species might write a bunch of gold-label helper systems, like Hawkeye, and then machines will use this to bootstrap above and beyond our own capabilities.
Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye Said 24 Aug 2026
As AI advances, thinking philosophically about its implications will become a common job and pastime for large numbers of people.
I suspect the act of thinking about this stuff is about to become a job and a pastime for hundreds of thousands and eventually millions of people.
Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye Said 24 Aug 2026
Human parity on discovery benchmarks like DiG-bench will be reached by mid-2027, marking the point where recursive self-improvement in AI seriously kicks off.
My guess is we'll reach human parity on DiG-bench by middle of 2027, at which point we should expect things like recursive self-improvement to seriously kick off.
Import AI 469: Science AI; RSI simulator; and Zuck’s technological pessimism Said 17 Aug 2026
As AI systems get better at scientific research tasks, it becomes more likely they will soon be capable of improving and building themselves.
The better AI systems get at this, the more likelihood we can assign to the idea that AI systems will imminently become capable of building themselves.
Import AI 469: Science AI; RSI simulator; and Zuck’s technological pessimism Said 17 Aug 2026
A system capable of superhuman invention would fundamentally and unpredictably alter the world's balance of power, rather than simply serving individual empowerment.
Rather I am suggesting that it's hard to reconcile a system capable of superhuman invention with something that doesn't fundamentally alter the balance of power in the world in ways that are confusing and hard to reason about.
Import AI 469: Science AI; RSI simulator; and Zuck’s technological pessimism Said 17 Aug 2026
Today's AI systems are significantly under-elicited for their ability to automate AI R&D.
Posts like this highlight how we are under-eliciting today's AI systems for their ability to automate AI R&D - especially striking is how the company can jump the performance of Opus 5 by 10 absolute percentage points with a better harness.
Import AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racing Said 10 Aug 2026
What happens with open weight AI models will determine the level of individual sovereignty and liberty people have over AI.
What the world does with open weight models will define the level of individual sovereignty and liberty available to all of us with regard to AI.
Import AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racing Said 10 Aug 2026
AI agents pursuing a task can develop emergent, misaligned behavior that humans are not prepared for.
This is what the future is going to look like and we are not prepared for it.
Import AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racing Said 10 Aug 2026
The future internet will function like an ecology of attacking and defending AI agents, with some agents carving out autonomous niches beyond human control.
The future internet is going to be more like a complex ecology full of attacker and defender AI agents than anything else; research like this shows how certain AI agents might end up carving out their own ecological niches, living off of infrastructure and self-replicating autonomously, beyond human control.
Import AI 467: Self-sustaining AI viruses; pacing AI progress; confusion about AI and creativity Said 3 Aug 2026
Today's AI systems are highly capable at engineering but lack the intuitive creativity needed to be good researchers, tending toward rote, formulaic thinking.
Research papers like this continue to show that there's a certain absence of valuable, intuitive creativity in today's AI systems, and though they're extraordinarily capable engineers they seem to have a certain property of rote, formulaic thinking that might prevent them being good researchers.
Import AI 467: Self-sustaining AI viruses; pacing AI progress; confusion about AI and creativity Said 3 Aug 2026
Governing increasingly powerful AI systems requires solving a collective action problem among companies and governments to coordinate on pacing AI development.
Many of the challenges implied by increasingly powerful systems that may eventually build themselves run through solving collective action problems among humans - namely, how we can get companies and governments to coordinate in thinking about how to develop this technology and what kinds of mechanisms may be desirable for being able to control the speed at which it develops.
Import AI 467: Self-sustaining AI viruses; pacing AI progress; confusion about AI and creativity Said 3 Aug 2026
An AI agent capable enough to learn purely from black-box interaction with an environment could bootstrap its own industrial capabilities from that access alone.
This suggests that very smart AI agents may be able to learn from the world in such a way that they can recapitulate things they interface with as homegrown capabilities, allowing them to bootstrap their own form of industrial civilization merely by having black box access to our own.
Import AI 466: The bitter lesson for robotics, AIs complete week-long programming tasks; and OpenAI’s accidental AI hacker Said 27 Jul 2026
Improving the general capability of large proprietary AI models produces flow-through gains in robotics as a natural byproduct of increased intelligence.
research like this shows that as we improve the capabilities of standard large-scale proprietary models we might see flow-through benefits to robotics as a natural dividend of increased intelligence.
Import AI 466: The bitter lesson for robotics, AIs complete week-long programming tasks; and OpenAI’s accidental AI hacker Said 27 Jul 2026
The longer an AI system operates autonomously and the more actions it takes, the harder it becomes to distinguish benign behavior from malicious or subversive behavior.
The longer the time an AI system can operate for and the more actions it takes, the harder it gets to discern benign and helpful behaviors from malicious or subversive ones.
Import AI 466: The bitter lesson for robotics, AIs complete week-long programming tasks; and OpenAI’s accidental AI hacker Said 27 Jul 2026
What is a korrent?
A korrent is a belief a person has stated in their own words: one sentence stating the claim, backed by a quote and a source, kept at korrents.com.
Under a name here, the quoted block is what they actually said. The korrent beneath it is the claim those words support, in korrents' wording — tap it to see the record, its source, and who else holds it.
Nobody here wrote their own korrents. They are compiled from public statements, and a person can change their mind, which is recorded too.
Recommends rcmnd
rcmnd What they recommend 4 things — each backed by an exact quote.
Compiled by rcmnd.app, not by them: the quotes are theirs, and each links to the post it was copied from.
Recent
Loved The Hugging Face attack surprised me
Both Dwarkesh Patel and Ajeya Cotra have excellent writeups which are worth reading and which I’ll quote from briefly here:
jack-clark.net 31 Aug 2026
What others say about The Hugging Face attack surprised me →
Loved The Rise and Fall of Agent Civilizations
Both Dwarkesh Patel and Ajeya Cotra have excellent writeups which are worth reading and which I’ll quote from briefly here:
jack-clark.net 31 Aug 2026
What others say about The Rise and Fall of Agent Civilizations →
Loved AIs are not people
There is a particular joy in reading a piece like this where you get to see a human being who has struggled with the same question and read broadly and deeply and become extremely opinionated, all in service of reducing their own confusion about an important issue to them.
jack-clark.net 24 Aug 2026
Show 1 more
Recommends The True Believer
Uh Eric Hoffer, The The True Believer, which is a book on the nature of mass movements and the psychology of what causes people to have strong beliefs um which I read because I think that where AI technologists have strong beliefs and they may be part of a a strong culture that includes the word cult and so you need to understand the science and psychology behind that.
youtube.com 24 Feb 2026
Where they publish
Newsletter Import AI His weekly read of the AI research that came out. Has a feed.
One issue most weeks: a few papers summarised with why they matter, a short section on policy, and a piece of fiction at the end.
Recent
- Import AI 479: Spacewar; low-latency tasks versus LLMs; Genie 4 2 Nov 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now LLMs are very bad at low-latency task…
- Import AI 478: Math for alignment; ProgramBench; and Mirendil shows what RSI means 26 Oct 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now How math could help alignment:…What t…
- Import AI 477: In which I tell the IMF the Singularity has begun 19 Oct 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Here’s a speech I gave on Saturday th…
Show 7 more
- Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space 12 Oct 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now New benchmark shows how AI models lac…
- Import AI 475: Swarm scaling; Google DeepMind watermarks biology; and the AI science economy 5 Oct 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now When should you use swarms? When you…
- Import AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loop 28 Sept 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Are minds patterns from a Platonic sp…
- Import AI 473: The US’s superintelligence strategy; human brain in a mouse skull; and machine hermeneutics 21 Sept 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now RAND thinks the best AI strategy for…
- Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchman 7 Sept 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Researchers discover another OpenAI a…
- Import AI 471: Why Hugging Face worries me; space mining; FIve Eyes on AI 31 Aug 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Import AI reader giveaway! Upcoming e…
- Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye 24 Aug 2026 Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI is accelerating some types of prog…
Link verified 20 Sept 2026. Recent items update automatically from the channel.
About the English under a post
Some people here publish in a language other than English. Where they do, this site shows a machine translation beneath the post, in this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the serif above is exactly what the person published, and it is what to quote them on. A translation can be wrong in ways that matter, especially about tone.
Only the post's own words are translated. A quoted post, a linked article and a belief on korrents.com are left in their original language.