Let’s play all playoff games at Disney world from now on #repbx

in 2027 engineers are going to be reviewing prs through HTML files with pictures+diagrams instead of diffs. much better experience

My dot is great. Really like how it can spinout codex sessions and the cloud/computer integration they have implemented. If only it could control Claude sessions. Codex is like the postal service. It takes 1-3 business days to complete anything

Jorge posada would never do that #repBx

AI engineer for the Mets making 200k sounds like a dream job. Until you realize you have to work for the Mets

dot is great because instead of grok bot where it launches a different application to code, dot just spins out a codex session in the same control pane. great feature @thsottiaux

Jev is a huge power-up for research agents! @edwardirby built a @youdotcom agent that monitors business risks, and compared Jev to an ordinary LLM judge. • Same report quality • But the LLM flip-flopped: 0.35 -> 0.68 -> 0.50 on the same threat • the LLM missed 5 of 11 investigations. Jev missed none. • Jev was 250x cheaper, 3-6x faster Write-up 👇

super cool from edward check it out!

How to build a reliable risk agent without a frontier model ($0.02/sweep): • @youdotcom search • @typesafeai Jev for typed judgments • @QwenDevs for proposals & synthesis • MCP for integration Full architecture & cost breakdown below 👇

ABS challenges in postseason baseball are the most electric part of the game

everyone should be spending at least 50$ personally on AI tooling. helps a lot with coming up with new ideas

i may not need ultrafast. but it looks incredibly addictive

Unreal 2 weeks for jev

In a world of metrics and distractions there’s incredible alpha in first principles thinking. What do you need to do and how are you gonna do it

engineering+design should use opus and everyone else should use sonnet. maybe give people opus for their power points and spreadsheets

Every ai application layer startup is just a differentiated way to sell you tokens

3x fewer tokens. 84% accuracy on the Vertical RTK benchmark. That’s what we got using Jev through @OpenRouter as a reranking layer for Knowledge, which we recently launched in the Web Search API. Knowledge returns structured, real-time data from authoritative sources across domains like stocks, sports, and more. For many queries, that data answers the question directly, so much of the retrieved web context is unnecessary for the synthesis model. Jev sits between retrieval and synthesis, selecting only the most relevant results to pass downstream.

Thrilled to launch Knowledge: extending our capabilities to data beyond the web. Stock prices, crypto, sports, weather, prediction markets, and more. All available now through our web search API at ZERO additional cost, and just +140ms higher latency. Try it out and let us know what you think!

I've been working on this for a while and I'm really proud of what we shipped! The web is a crucial source of context for agents, but it's limited by what can be crawled. So we added structured knowledge to the web search API to push the frontier of what agents can do. Excited to see what people build!

the grok bot + cursor + opus-5.5 stack is OP in the office right now