Mods are absolutely insane. You can now customize Claude to work and look the way you want by just prompting it. Each person works differently, so there's no reason why everyone should have an identical Claude experience. Make Claude your own, and share mods as plugins so others can try your mods too.

It’s now even easier to get help from my company for agentifying analysis: ktandb.co.uk Do you want one off analysis, or a sharper analytical workflow? Do you want us to do everything, or to be involved throughout design and implementation?

I asked two AIs to redesign my landing page. See original, Alt A and Alt B in the next post and vote for your favourite: Https://ktandb.co.uk

More and more often, I'm prompting via voice. I think this helps me articulate my ideas better, and it lets me provide richer context and more useful information for agents to start working from. Who else is doing that on a regular basis? What are your tips?

Can't agree more

I like it when it doesn't get my sense of humor.

Writing like lightning. My transformative new system for better nonfiction writing, including scientific publication, inspired by slip boxes, quietly supported by AI agents. Link below to sign up for more info.

Do not go gentle into that good night

I love R, but in the age of AI, it makes a lot of sense to switch to the best tool for a given task. I'll try D3.js.

The Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training at MIT released the final report. It is a call to action to redesign learning and research activities to make the use of AI beneficial and not detrimental. As one can easily imagine, the report recommends reinvigorating the residential experience based on shared experiences and a focus on in-person activities. aiandeducation.mit.edu/repor…

I propose Stanford NLP as an independent third-party evaluator under @DarioAmodei’s 3 step plan. For important parts of the work, universities would be better than any other organization (see below 🧵👇), and, of university groups, @stanfordnlp would be the best one to choose. 😊

This Is part of the solution. We should involve academia and institutions.

I believe some of these arguments apply to data science too

Do large language models actually understand the world, or are they just very good at pretending? Does it even matter? Our recent special issue in the Royal Society, “World Models in Natural and Artificial Intelligence,” brings together pioneers across AI, biology, and philosophy to argue that the path to true intelligence runs through something deeper: the ability to model not just language, but causality, the self, and the physical world. Featuring contributions from Douglas Hofstadter, Michael Levin, Josh Tenenbaum, Samuel Gershman, Melanie Mitchell, and others, the collection asks a radical question: What if the next leap in AI requires not just more data, but systems that model themselves? Here are 3 ideas that might redefine how we build AI: 1. Capability is not the same as true intelligence. Current foundation models are incredibly capable, but they often lack true emergent intelligence. They learn surface statistics instead of compact, causal abstractions. Simply scaling compute will not fix this fundamental issue. 2. Self-modeling is an engineering primitive, not a philosophical luxury. New research in the issue shows that when networks learn to predict their own internal states, they compress and simplify, becoming more efficient as a form of regularization. For physical AI and future agents, a self-model is what will allow them to adapt their own skills and morphologies in real-time. 3. The hardest problems in AI are continuous with the hardest problems of life. Biological minds do not passively ingest data; they actively explore, driven by empowerment to increase control over their environment. If world modeling is about an agent representing itself in relation to its environment to survive and adapt, then general AI may need to look much more like artificial life. The takeaway is that the next leap in AI won’t come from just scaling up next-token prediction, but rather from systems that are agentic, self-referential, and temporally grounded. Read the introductory essay and the full special issue here: royalsocietypublishing.org/r… What do you think is the most important missing ingredient in today’s AI systems?

Anthropic identified and disrupted operations in which threat actors tried to use Claude for malicious activity. Good! However, it seems to me that this information brings to the table the need for good regulation and checks, with intergovernmental institutions. We have the IAEA for nuclear weapons, we should have something similar for AI. And we need a balanced public discourse. anthropic.com/threat-intelli…

I strongly believe that fundamentals matter more than ever. Let's dive into this gem. "A philosophy of software design" by John K. Ousterhout.

AI is a net job creator in the US. Over a million new roles in data centers, engineering, and related trades. More than enough to absorb the cuts in admin work. Software developers tell the real story. Negative in 2024. Negative into early 2025. Then a rebound. First AI replaced the routine. Then it multiplied the demand for those who can direct it. Economists call it reinstatement: the same technology that erases a task creates another but not for the same person, neither immediately. The one who loses an admin job does not become a data scientist the same day.

"These are great times". It fully resonates with me!

"I used to sit down and just let my mind drift. Most of the time a useless activity, but occasionally it would lead to something that was useful. That sort of dreamlike state was very important in my life." Richard Robson tells us how he accesses a restful state-of-mind in which ideas flow in this episode of our podcast – Nobel Prize Conversations. We also speak with Robson's co-laureates Omar Yaghi and Susumu Kitagawa about how they changed chemistry forever by letting their minds wander to unplanned places. Join us for a discussion with the 2025 chemistry laureates: linktr.ee/nobelprizeconversa…

If everybody can build an app, a tool, a function, a package, or a library... what is the future of communities around programming languages?