Oh man... it's going to be bad, Doug. No one can live code an HNSW data structure in one hour with 600 watching.

This should be a really fun talk, Shawn Simister and I have been recalling to ourselves how it all worked. What we realized is that we were doing an early form of LLM-as-judge back in early 2023. maven.com/p/da8264/how-e…

Part two of live coding a vector database. Thanks to Eric Pugh of OpenSource Connections for emceeing so it feels participatory :) maven.com/p/eca048/build…

Digital art from Doug Turnbull in celebration of the new American pope

Writing in NYT, former Google CEO Eric Schmidt warns that “China Tech Is Starting to Pull Ahead”: “China is at parity or pulling ahead of the United States in a variety of technologies, notably at the A.I. frontier. And it has developed a real edge in how it disseminates,

Jim Fan says NVIDIA trained humanoid robots to walk and move like humans -- zero-shot transfer from simulation to the real world. 10 years of learning, compressed into 2 hours of simulation. Turns out you don’t need a giant model to master motion. “1.5 million parameters, not

In my LLM app dev course, I show how that Agency, RAG, workflows are all built on really simple ideas. From simple ideas you can build differentiated user experiences. In this post I demonstrate how to create Artifacts which help users get more work done maven.com/p/e4d1ea/cut-t…

A huge thing that most AI chat applications are missing (and especially code gen) is the ability to ask sensible clarifying questions. It's a tough thing to build because you _also_ don't want it to ask _all possible_ clarifying questions. There is an art (that humans are good

My technical reviewer's copy has finally arrived! Last year, I had the pleasure of reviewing “Prompt Engineering for LLMs” by Albert Ziegler and John Berryman. Both authors were early researcher developers on GitHub Copilot. And in this book, they share the knowledge on prompt

Interested in learning about Evals? Hamel Husain and Shreya Shankar are dropping gold in this short talk.

Semantic search has become the defacto standard for RAG. But sometimes semantic is a bad fit. Lexical search offers an attractive alternative. This talk will open your eyes to new possibilities with lexical search. maven.com/p/a1e564/lexic…

Here is the recording of the deep dive into GitHub Copilot evals with John Berryman and Shawn Simister - probably my fav talk of the series on evals. Chapter summaries below Links to more resources in the reply. 06:01 - Overview: how evaluations were central to Copilot's development

I had to pop out a little early, but this was a great discussion between 3 very different but very experienced consultants/contractors. jason liu how do I find links to all your AI Consulting content?