I loved watching this. My takeaway: don’t just say what it does, explain how it does what it does.
I recently spoke with the CTO of a popular AI note-taking app who told me something surprising: they spend twice as much on vector search as they do on OpenAI API calls. Think about that for a second. Running the retrieval layer costs them more than paying for the LLM itself.
— James Luan, Engineering architect of Milvus
Tags: vector-search, embeddings
This is the trailer for a documentary celebrating the life and work of actor & comedian John Candy.
I loved John Candy; how could you not? Uncle Buck was my favorite of his movies. I can’t believe he died more than 20 years ago already. (via craig mod)
Tags: John Candy · John Candy: I Like Me · movies · trailers · video
Yesterday, Jackson Goldstone won the 2025 Downhill Mountain ike World Championship. Here’s the POV of his winning run down the *very* steep course in Champéry, Switzerland. (This is *bananas*!)
Forthcoming book by Bernie Sanders: Fight Oligarchy. “Sanders explains how the United States today is an oligarchic society in which a small handful of multibillionaires exercise enormous economic and political power.”
I agree with the intellectual substance of virtually every common critique of AI. And it's very clear that turning those critiques into a competition about who can frame them in the most scathing way online has done zero to slow down adoption, even if much of that is due to default bundling.
At what point are folks going to try literally any other tactic than condescending rants? Does it matter that LLM apps are at the top of virtually every app store nearly every day because individual people are choosing to download them, and the criticism hasn't been effective in slowing that?
Tags: ai-ethics, anil-dash, ai, generative-ai
I am once again shocked at how much better image retrieval performance you can get if you embed highly opinionated summaries of an image, a summary that came out of a visual language model, than using CLIP embeddings themselves. If you tell the LLM that the summary is going to be embedded and used to do search downstream. I had one system go from 28% recall at 5 using CLIP to 75% recall at 5 using an LLM summary.
Tags: vision-llms, generative-ai, ai, embeddings, llms, jason-liu
A new blood test could dramatically improve survival rates for women with ovarian cancer by detecting it much sooner. The researchers tell us it's a game changer that could go on to save tens of thousands of lives around the world every year. Also: we meet the female Muslim jockey breaking new ground and inspiring the next generation. Khadijah Mellah has just made her amateur debut. Plus the woman who underwent a rare triple organ transplant talks about the joy of getting back to normal life; how a hand-cranked washing machine is saving thousands of people from hours of back breaking work; football's Homeless World Cup; and why a member of indie band Kasabian is opening a new music venue in a tiny Welsh town. Our weekly collection of inspiring, uplifting and happy news from around the world.
