We nearly named it MBTI (Mithril Batch Tokens Inference) last minute, so we could make Myers Briggs puns (although maybe that's more of a LinkedIn joke). Wrt "intelligence too cheap to meter", INTJ: I Need Trust in Jevon's*. If you're a pun enjoyer, comment or retweet with

+1! The compound AI systems or "menagerie of models" (Greg Brockman's wording) future is becoming more and more clear. Lots of room for a renaissance of "basic research" and for novel ml+systems co-design.

Cool work Dilara Soylu and Lakshya A Agrawal !

Extremely cool, modifying GRPO to optimize compound systems.

Spencer Schiff yes we may consolidate two models into one but there’s an explosion of model sizes and finetunes and whatnot

Really cool -- looks like system design and prompts can have more impact than foundation model weight training past some point.

If you're running agents in production, consider taking this short survey from my research group! We're collaborating with IBM, Stanford, UIUC, Intesa Sanpaolo and others to better understand the challenges in building agents. It only takes 5 minutes: agents-survey.github.io/AI-Agent-Surve…

Thrilled to announce our $10M seed for Nuance Labs, led by Accel with support from South Park Commons and Lightspeed, to build a multimodal human foundation model that enables emotionally intelligent conversational AI.

I gave a keynote at VLDB 2025 🇬🇧 about why we think it's time to rethink OLTP databases with Lakebase, which combines the cloud-native Postgres design of Neon - Serverless Postgres with Lakehouse. Reasons include the cloud, changing demands on DWs to be more real-time, and AI agents. Slides ⬇️

Congratulations, Zach (Zhihao) Xia and team!