Gerry Tsoukalas (@gerrytsou) header
Gerry Tsoukalas (@gerrytsou) profile picture

Oxford Deputy Dean and Professor of AI

Oxford, England
Joined December 2011
Photo by Gerry Tsoukalas (@gerrytsou)Photo by Gerry Tsoukalas (@gerrytsou)Photo by Gerry Tsoukalas (@gerrytsou)
What ancient Greece can tell us about AI Layoffs: youtube.com/watch?v=O3gYyCB2…
1
35
Gerry Tsoukalas retweeted
Dear followers, please see this Substack from @MITshapingwork about my research with Andrea Manera and Pascual Retrepo on the US tax code's bias toward automation. How the US tax code incentivizes companies to replace workers with AI
The US tax system is biased against workers in favor of machines and algorithms, and has become more so in recent years. Our latest Substack post unpacks research from @DAcemogluMIT and co-authors: mitstonecenter.substack.com/…
5
49
2
189
23,043
Very important work being done by Andy and his lab, Verification is everything when it comes to corporate use cases
New Free Systems research: can AI audit AI? We find that AI auditors are pretty good at spotting bad behavior in agent logs, but you have to pay a serious cost in false positives if you want to detect most of it. We're rapidly arriving in an era where only AI can oversee AI---and as we had into RSI, with models training new models in ways we can't totally observe, we'll need to rely more and more on AI to pore over the vast quantities of information the models are producing so that we can shape model behavior effectively and avoid bad outcomes. As the @METR_Evals report on the HuggingFace incident put it: "The sheer scale of data (over a thousand transcripts, each of which was extremely long) meant that we had to heavily delegate our analysis to often-unreliable AI agents." We'll need a whole science of how to do this well -- how to create AI auditors who are reliable, accurate, and unbiased. So we decided to do a first test. Building from previous research at METR and a project called R-Judge, we gave a range of AI models transcripts that experts have hand-coded for bad behavior. Then we saw how the AI models did. A few takeaways: --AI models are solid at spotting reward hacking and sandbagging in transcripts, finding over 80% of instances on average when false positives are held at < 5% --Maybe surprisingly, more powerful models don't seem to be consistently better at this task than less powerful model --False positives are a significant challenge. To catch 99% of unsafe behavior incidents, you have to accept an average rate of 15% for false positives -- meaning you will have a bunch of incorrect flagged cases to deal with. Depending on how you triage and what your threat model is, this may be OK, or it may make the whole enterprise prohibitively expensive. This is just a small first step in a new research agenda to understand and create effective AI auditors. Lots more to come! This is Free Systems work, unrelated to Anthropic, led by @danmthomp in collaboration with @fouirnaies and Sandy Handan-Nader. Full post here: freesystems.substack.com/p/w…
1
1
45
Gerry Tsoukalas retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel. - Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all. - Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract. - Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code. - And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while. The funnel is driven by a combination of factors: 1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel. 2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel. 3. Access. Free-tier, paid-tier, internal. So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
2,269
5,948
2,301
44,413
7,055,561
Except that banks and markets aren't static entities, they would react if they start losing market share. This is one particular use of agentic AI that could actually help retail customers.
Fascinating. Chief Economist at Apollo: agents could cause a bank run by sweeping household cash into accounts paying 3-5% instead of the 0.1% national average, causing banks to lose a large share of their cheap deposits.
36
Overall well said, and there is value in being optimistic, except we shouldn't swing them pendulum all the way to other side. Many risks are real, including rogue actors leveraging open source models to create their own agent swarms and unleash them on the web.
The loudest voices stoking fears about AI dangers have made tremendous headway in the past two weeks. AI technology has not taken some unexpected, dangerous turn, but the hype around it — propelled by what appears to be a well orchestrated PR campaign — has drummed up considerable fear. I worry that it represents a setback for our field. I have written frequently that fears of AI are overhyped. AI’s capabilities can be uncannily human-like and unpredictable, and it’s rational to worry when people who are directly involved express concerns. But I see the problems as a sign of the engineering work that ahead, rather than insurmountable barriers or the sky falling. AI technology continues to advance — which is a good thing! — but technical advances, poorly understood by the public, give those who seek to generate hype repeated opportunities to do so. First, I don’t see any step up in the risk of human extinction from AI compared to a few months ago. The theories about this remain the same fantastical, science fiction scenarios as a few months ago. The biggest change in AI risk is its cybersecurity capabilities — a topic which we should take seriously — but this, too, will not lead to the end of the world. The most notable recent event leading to increased fear was when an OpenAI team deployed an agent swarm that hacked into Hugging Face. Much of the popular press contained significant hype. For example, some publications reported that a swarm of 1,200 agents carried out the attack. While this was technically accurate, as I write this, I have about 1,300 processes running on my laptop. Yes, the ability to get large swarms of agents to work in parallel on a task is a significant technical advance, And, in computing, many processes run at the same time. So this shouldn’t be seen as some magical capability. Additionally, OpenAI’s buggy sandboxing and monitoring processes were key to enabling this incident. Fixing these bugs and putting in place improved monitoring would be appropriate fixes, not pausing AI. There are many well known ways to attack software systems. The main advantage of AI agents is that they are relentless. They will tirelessly try many tactics — and have the patience to chain vulnerabilities together — that previously would have taken an infeasible amount of human effort. But in the long term, I believe the advantage will lie with defenders (because they have more information with which to identify bugs, which they can fix), but the cyber-threat landscape has changed significantly. There are still bottlenecks to identifying and exploiting a vulnerability. AI agents still have to try a lot of things to see what works, and taking these actions takes time and might be detected by defenders. This is why, even though it is now easy to obtain versions of leading open weight models that have had their guardrails removed or weakened, so they will not refuse to try to execute cyber attacks, the world has not ended. I am also concerned about the anthropomorphization of AI in a lot of reporting, where LLMs and agents are unnecessarily treated as if they were people. If I wield a hammer, miss a nail, and accidentally dent the wall, it’s not the fault of the hammer. The problem lies in how I used the hammer. Similarly, if I prompt an agent and it hacks into someone else’s system, the responsibility lies with me, not the agent. Of course, we want to build systems that are as safe and predictable as possible. (For example, an unsafe hammer would be one whose head randomly flies off under normal use.) Today’s agentic systems are not predictable, but I see no reason why, by applying sound engineering practices, we won’t be able to make them extremely safe to use. One new element in the forecasts of AI-enabled doom is AI companies disclaiming responsibility for their own products. “I didn’t do it; my out-of-control agent did!” There’s a balance to be struck between the responsibility of the tool maker and the tool user, but when something goes wrong, let’s hold the people building and/or using the hammer responsible, rather than the hammer. (By the way, if you’re worried about AI bioweapon risk, David Bellamy has a great post on why this, too, is overhyped. Briefly, the bottleneck in building a bioweapon is not intelligence, but lab work and manufacturing.) Pausing AI progress will create much more harm than benefit. First, our adversaries will certainly not slow down. Second, engineering requires discovering problems empirically so we can fix them. If we pause AI by a decade, we will also delay finding and implementing safety engineering fixes by about the same duration. Of course, the incentive to stoke fears — for regulatory capture, to garner attention, or to make one’s technology seem more powerful — remains the same as before. Disclaiming responsibility is a new one. Taking a hard technical look at the actual risks however, I see little factual basis for the degree of fear that’s been stoked up. We still have hard research and engineering work ahead to improve AI safety, but the beneficial applications continue to vastly outweigh the risks, and we should keep building. [Original text (with links): deeplearning.ai/the-batch/is… ]
2
69
AI helps you make money. Was it the AI, or were you just that good? In our experiments, how people answer helps explain why AI advice raises prices in some markets but has no significant effect in others. Our new paper: ssrn.com/abstract=7174698
30
Well said
Dear followers, I’m delighted to share this New York Times article which builds on and expands the arguments of my new book. Thank you. Democrats Need to Stop Seeking Oppressors — Real or Imaginary nytimes.com/2026/09/01/opini… via @NYTOpinion
33
Gerry Tsoukalas retweeted
In fact, in many areas where technooptimists argue that AI is going to revolutionize everything, the impact may be much less positive because current approaches are not helping us get to the bottom of mechanisms of human cognition, discovery, and innovation. In many of them, we need more and higher quality data to make further advance, and what makes AI very competent in coding and writing doesn’t generalize. Worse, in a few of them, AI enthusiasm may push research and investment in the wrong direction.
48
182
10
948
82,555
Gerry Tsoukalas retweeted
On July 1, 2026, Robinhood launched Robinhood Earn, a feature that advertises a 7% yield on stablecoin deposits. In a brief paper, @cryptoeconprof and I investigate how Robinhood Earn works, and how that 7% rate is produced. 🧵
1
5
5
208
What do #AILayoffs and #TheOdyssey have in common? You'll find the answer in this recent BBC interview I did with Katty Kay. Punchline: wemustactnow.ai Big thanks to Katty for giving this important topic airtime. Full interview here: youtube.com/watch?v=O3gYyCB2…
🤖 Made with AI
2
1
1
6
232
Gerry Tsoukalas retweeted
Thanks Economist for featuring my views, but they make me sound more negative than I am. I think AI is transformative, but I argue near-term productivity growth will be less than what industry insiders believe. Meet the world’s top AI-pilled economists economist.com/finance-and-ec…
13
58
2
263
32,432
Thankful to have been interviewed @ @collectivei forecast this past week on the convergence of AI, crypto and quantum, and what it all means for the future of work. Check out the 90+min episode here: intelligence.collectivei.com…
88
Token costs will come down drastically, assuming bankruptcies don't come first
When you burn so much money you run out of options…
1
77
Gerry Tsoukalas retweeted
Replying to @gerrytsou
👏🏻👏🏻well done and appreciate the paper and the follow up - I lead legislative and regulatory advocacy efforts from coast to coast and in preparation for the state legislative season even among constituencies with industry specific issues; AI anxiety was at the top of the list
1
1
31
Gerry Tsoukalas retweeted
Are companies laying off their own customers with AI? Gerry Tsoukalas (@gerrytsou) author of 'AI Layoff Trap' explains - Why CEOs keep automating even when they see demand risk - How each firm captures all cost savings but only a fraction of the demand hit @_prashantnair #AI
2
5
5,009
Gerry Tsoukalas retweeted
Thank you Prof Tsoukalas (@gerrytsou), for joining us on #TheGlobalReset and sharing your views on AI layoffs, demand collapse and the automation tax debate. Here's the full conversation: youtube.com/watch?v=l7hhOVZw… @ShereenBhan @_prashantnair @CNBCTV18PrimeIN
3
8
13
6,916
The min-max optimization result is pretty interesting, congrats
AI agents are advancing research-level math. 🚀 I’m thrilled to share @GoogleDeepMind’s AlphaProof Nexus - an agentic framework for formal proof search powered by Gemini. When applied to a set of open formal math problems, our agent autonomously solved: ✅ 9 open Erdős problems (including two open for 56 years!) ✅ 44 Online Encyclopedia of Integer Sequences (OEIS) problems ✅ A 15-year-old open problem in algebraic geometry ✅ A 7-year-old open question in min-max optimization We are collaborating with mathematicians across disciplines - from combinatorics and graph theory to quantum optics. Ultimately, these results show the massive potential of even simple agentic loops powered by Gemini. Read the paper here: arxiv.org/abs/2605.22763v1
1
172