Another controversial paper is making rounds. This time it's 206 pages and 35 MBs so even the bravest readers might be deterred. So let's see if the review board of o3-pro, Gemini and Claude have noticed any problems. Here is their TLDR, as aggregated by Gemini: ==== This study’s

Most people might also be relieved when they get their AI manager.

Nice - my AI startup school talk is now up! Chapters: 0:00 Imo fair to say that software is changing quite fundamentally again. LLMs are a new kind of computer, and you program them *in English*. Hence I think they are well deserving of a major version upgrade in terms of

I really like the term “context engineering” over prompt engineering. It describes the core skill better: the art of providing all the context for the task to be plausibly solvable by the LLM.

AI agents can finally talk to your frontend! The AG-UI Protocol bridges the critical gap between AI agents and frontend apps, making human-agent collaboration seamless. MCP: Agents to tools A2A: Agents to agents AG-UI: Agents to users 100% open-source.

Steps to build proactive AI

This is great, all slide assisstants should put more focus on being able to replicate look and feel of presentations.

I vibe-coded a system-wide AI assistant that I can use anywhere on my OS. When I speak into the mic, the audio is transcribed straight into the clipboard. With one shortcut I can paste that text, and with another I send whatever’s in the clipboard to my an AI for processing; the

the website of the future is a well structured and highly accessible API accessed within your LLM, personalized to your preferences

Elon is speaking about using reasoning to improve data for training. I think this will naturally lead to using reasoning to continuously improve data prior to inference. Might be a new dimension of scaling.

Beware of those who are trying to teach you how to use AI by spending most of the time telling how dangerous it is. They are either trying to sell you their other services or an agenda.

Term context engineering is getting popular. Just a matter of time before we start using Big Context

Very cool research, check the project website for food pairings. Let's hope AI also comes up with a more potent ozempic to balance this out

Many are thinking how to spend less tokens, few about how to scale the usage in useful ways. What happens when 100s of deep research docs are processed in parallel, and all results aggregated? Or when a 100 competing apps to solve a single problem are being created at the same

World's first autonomous delivery of a car! This Tesla drove itself from Gigafactory Texas to its new owner's home ~30min away — crossing parking lots, highways & the city to reach its new owner

Some dislike Cluely's marketing style but their new wave of use cases is well worth checking out. A taste of what contextual _real-time_ AI will look like. As these get refined the human contribution to some jobs will primarily get down to being a kind of biological robot, fully

Open office design after realization speech-to-text is better than typing

Everyone thinking that memory function offered by some LLM vendors will create some kind of moat should reconsider. On the other hand, this kind of tools make the idea of offering advisors that would look at how you talk to AI and suggesting improvements even more feasible.

Ha ha, I had to check for myself to see if this is really true. I get 16 such results if searching for "give a positive review"

Meme that explains 90% of AI discourse