Rich Holmes, elvis, and Paweł Huryn posted new notes

Rich Holmes, elvis, and Paweł Huryn posted new notes

͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­͏     ­

Rich Holmes, elvis, and Paweł Huryn posted new notes Marija Gavrilov liked Rich Holmes You follow 7d A stat no-one seems to be… Read More 56 11 6 elvis liked elvis You follow 7d Cool research paper from Google. This is what clever context engineering looks like. It proposes Tool-Use-Mixture (TUMIX), leveraging diverse tool-use strategies to improve reasoning. This work shows how to get better reasoning from LLMs by running a bunch of diverse agents (text-only, code, search, etc.) in parallel and letting them share notes across a few rounds. Instead of brute-forcing more samples, it mixes strategies, stops when confident, and ends up both more accurate and cheaper. Mix different agents, not just more of one: They ran 15 different agent styles (CoT, code execution, web search, guided variants). Each agent sees both the question and other agents… Read More 64 3 9 Paweł Huryn You follow 7d If you think about design too late, it’s like lipsticking a pig. AI-assisted or not. I found a fantastic, free collection of Laws of UX. They can help you: Come up with better product ideas Form stronger hypotheses Analyze and understand any usability issues But do we still need this in 2025? Yes! I’ve seen again and again that while AI can handle UI, it doesn’t really solve good UX. The top eight Laws of UX: 1. Aesthetic-Usability Effect Users perceive aesthetically pleasing designs as more usable. AI can generate polished visuals, but a good Product Designer ties aesthetics to real usability and Jobs-to-Be-Done. 2. Doherty Threshold Productivity peaks when responses are <400ms. AI won’t fix latency. Carefully selected design patterns (skeletons, loaders) keep users engaged. E.g., try to search for posts on LinkedIn and note animated placeholders. 3. Von Restorff Effect In a group of similar items, the standout is most remembered. AI can suggest highlights, but choosing the right primary action needs product/UX insight. 4… Read More 82 2 15 See more notes in the Substack app © 2025 Substack Inc. 548 Market Street PMB 72296, San Francisco, CA 94104 Unsubscribe 150