Knowledge catalogue
Some fun examples, I just gave basic prompts and the AI executed: Balatro, but for coin flipping (all the design and ideas was Fable): https://play-flipside.netlify.app/ The best self-aware snake game
The fact that Anthropic may take away subscription access to Fable in two weeks is weird & discourages investing in learning about the model. Subscription use is how you figure out what the model is g
The New York Times published a roundtable discussion between @DAcemogluMIT, @deanwball, @clarashih & myself about the future of AI & who wins at work. I think it is a really nice overview of the core
This article by Ethan Mollick describes the user experience and practical workflow of working with Mythos, an AI system. It likely covers the system's capabilities, interface, strengths, limitations,
One year prior to this post, o3 represented the most advanced AI agent available, marking a significant milestone in AI development. The statement reflects how rapidly AI capabilities have evolved, wi
Last time around Apple released a lot of information about how their AI version of Siri worked between local and cloud models, not so much this time It is nice to have a Gemma-like model on device, bu
The Matrix idea of keeping humans as batteries is obviously weird... we would be more useful as dice. LLMs default to very similar kinds of arguments & structure, and even different LLMs seem to colla
It is a really good time to store up a few of your hardest, most valuable, and most unusual ideas - whether for work, hobbies, or a new venture. Thanks to AI, really good & unique ideas are getting ex
As you can see It does seem like meaningfully better AI releases are accelerating, especially from OpenAI & Anthropic. To illustrate, I caused this timeline to be created. It only lists new models tha
One reason you want AIs to be better writers is that there is a lot of writing even in software, and it is incredibly painful to hit a menu which is filled with Claudisms or ChatGPTish phrases. A repo
The Gemini Pro models do not seem to be iterating anywhere near as quickly as Claude or GPT (last release was 3.1 Pro in February). Its causing a growing performance gap between Google and the other t
This chart from Anthropic is useful, since Agent Teams and Workflows are both very new and very powerful (and token hungry). On the other hand, maybe it doesn't matter as a lot of the decisions about
Also, a lot depends on Chinese labs continuing to ship open weights models. If they stop, the frontier falls further and further behind to those who want to use local/fine-tuned models. I think this i
At least until (if?) rapid improvement stops, it seems less likely someone is going to catch the Big Three AI Labs. Microsoft and Meta released their models, which were fine, but not frontier. SpaceX
A real problem with feeling the acceleration viscerally is that current models are really good and it is hard to feel the vibe difference on most individual tasks with new models, even as AIs continue
'As of May 2026, more than 80% of the code we merge into Anthropic’s codebase was authored by Claude.' Matches independent measures. There really is no sign this is slowing down (which doesn't mean th
Based on anecdotal conversations with peers, there is enthusiasm for AI among academics in the humanities (while still being worried - rightly - about the negative consequences as well), but they gene
This article explores the evolution of human-AI collaboration, arguing that the era of 'co-intelligence'—where humans and AI work together as complementary partners—may be transitioning as AI capabili
I actually read this & it is super weird, it appears to be an argument that prior machine learning systems (not generative AI) did not generate savings due to data issues so that will lead to a lack o
I think it is really worth reading this piece on RSI at Anthropic. There is a bit of navel-gazing, some marketing, and a lot of very sincere beliefs about what Anthropic thinks is likely in the near f
Ethan Mollick observes that both OpenAI and Anthropic's official documentation frequently falls behind their actual product releases by several months, resulting in outdated and conflicting informatio
Leaving aside the question of consciousness, the Ted Chiang piece has a reasonable point about moral atrophy if you let AI make choices. But it is also interesting in light of the fact that repeated r
See also: https://x.com/emollick/status/1799046218917822682?s=20 Alignment, of a sort: this paper conducts what they call a “moral Turing Test,” asking people to compare GPT-4o to humans on ethical qu
The capabilities of Claude Code and Codex have expanded a lot in recent months, they added many ways to approach work (subagents, skills, goal, workflows, plugins, etc). Given the AI labs can use thei
There are lots of issues with GenAI implementations but surveys seem to suggest broad positive ROI. Data & stories about AI failures are real and useful but need to be clearer about what they are meas
Had Claude Code build a snake game where the snake becomes aware it is in the game and then... stuff happens. Some impressive creative decisions by the AI (& also some very AI ones), I just gave a fir
Hey, its our paper! One of the most-viewed PNAS articles in the last week is “Persuading large language models to comply with objectionable requests.” Explore the article here: https://ow.ly/wOxl50Z6f
This appears to be Ethan Mollick's personal reflection on an AI-generated or AI-assisted project that includes racing and Stardew Valley game elements, with commentary that the conclusion exhibits cha
This appears to be a GitHub repository shared by Ethan Mollick containing code or a project called 'snake-awakening' that is available for public download and modification. Based on the repository str
In early May, the best superforecasters predicted that, by the end of the year, the longest METR 80% task horizons would reach 3-4 hours. In late May, Claude Mythos achieved that number. We also asked
Most people, including really accomplished people, don't have an accurate mental model of how LLMs operate (and why would they?) You see this in wide beliefs that AI is just copying from known sources
My timeline seems to have people surprised that U Chicago is getting Claude, but tons of schools (including U Penn where I teach) have school-wide AI There are lots of things that need to be figured o
HIPAA and FERPA compliant AI systems reduce institutional and legal risks when used by students and researchers by ensuring sensitive health and educational data are properly protected. Compliance wit
Ethan Mollick expresses skepticism about online posts that explicitly use words like 'genuine' or 'honest,' suggesting such language may indicate the opposite of what's being claimed. This observation
There are good things and bad things about AI in education & research just like everywhere else, but it’s also true that it’s not going away and that AI increasingly is becoming part of the tool set a
This story was so implausible that the only way it even (kind of) made sense if it is some sort of internal accounting placeholder at a cloud provider using their own compute. And even then it seems u
Ethan Mollick shared that a free game he deployed on Netlify consumed his monthly bandwidth credits in just one hour, forcing him to redeploy an optimized, more bandwidth-efficient version while prese
Another thing about AI writing is that while a single instance of AI writing on a topic may be fine, any situation where lots of people use AI to respond to a particular prompt (comments sections, hom
Big paper on AI coding agents using Github & other data The auto-complete tools (Copilot) led to 2.2x more code, local agents like original Claude Code led to 7.4x, & current remote coding agents 17.3
Here's the paper, which confirms that big changes are happening very quickly. Some really interesting approaches to teasing out what is happening in the software development cycle as well: https://pap
I wish the logos and textbox-at-the-bottom interfaces for Discord and Codex did not look so alike at a glance. I have confused the two a couple of times, leading to a confused GPT-5.5 and a confused g
It is difficult to know how good MAI-Thinking-1 is from the scores alone (like weirdly low GPQA & Terminal Bench 2.0) But Microsoft makes it really hard to try its models upon release (a general issue
Law professors wrote questions they were asked during office hours. Gemini 2.5 & humans answered them then other law professors blindly judged the results: -Gemini had a 75% win rate vs. professors -G
Since a lot of comments ask about it, they cannot really say anything about code quality using their methods. In part, the reason humans are the bottleneck is because humans are doing some sort of qua
The everything apps still look a lot like hybrids between chatbots and IDEs, rather than something built for general knowledge work. Too much assuming linearity & that final outputs are the only goal,
For those who wants stats, rather than anecdotes https://x.com/emollick/status/1983217724714066310?s=20 👀New data on the corporate ROI from generative AI from a large-scale tracking survey by my colle
I find debates over whether companies find AI useful to be odd at this point I talk to leadership teams at lots of big firms, and it is pretty universal that they are getting obvious and real value. T
'Just use AI' can be useful (I call it The Crowd in my informal framework I gave in my post below), but you also need a Lab where people can go with their ideas and Leadership to make decisions about
Lots of companies are in the 'encourage AI adoption' phase, whether teaching them ChatGPT/Claude or (sigh) tokenmaxxing. That dodges the harder problems of firm leadership: What do you want people to
The gains are most obvious in coding and operational areas. Even conservative organizations are adopting coding agents fast. That doesn't mean AI adoption doesn't come with huge challenges and cost co
/goal and other fully automated AI agents are cool, but not a great model for the future of work with people. Instead you want your AI to know when to ask you GOOD questions, maybe because it is stuck
I think Epoch does a great job benchmarking, but I continue to believe that open weights models are much more fragile, especially out-of-distribution, than their benchmarks indicate. Vibe-wise, I don’
It does seem like meaningfully better AI releases are accelerating, especially from OpenAI & Anthropic. To illustrate, I caused this timeline to be created. It only lists new models that scored 3 poin
On acknowledging AI (don’t take it too seriously) My modest proposal for how to acknowledge AI work is to use the Latin phase 'Fieri Iussit,' which means 'commanded to be made,' a common phrase on Rom
Claude really can roleplay an economist. I love this little comment Claude made after some robustness checks on the paper it wrote: 'On a 1–10 identification scale, I'd now put the paper at about 4.5
I cannot verify this claim as it references a future date (the URL timestamp appears invalid) and describes a model version (GPT-5.5 Pro) that doesn't exist in current public releases. The post appear
How lucky are you to have been born when and where you are? Had Opus 4.8 in Claude Code whip up a new visualization of all humans who ever lived. In addition to being neat, it is an interesting test o
The Veil of History is an open-source project shared by Ethan Mollick on GitHub that appears to be a tool or resource related to historical analysis or visualization, made available for public improve
Interesting that the GPT-5 Pro series models have consistently been the best models for single-shot attempts at the hardest problems since last summer. There has been no real competition in all that t
Prompts (verbatim): assume a universal veil of ignorance and you could be born as any human who has ever lived in history, what are the most likely socioeconomic conditions and locations that you woul