How can we ensure critical safety properties of AI agents? Like, no API keys leaked, no prompt injection, no data loss? It's a hard question which will require fundamentally rethinking the way we build modular software. We have an answer in our paper "Tracking Capabilities for Safer Agents" arxiv.org/abs/2603.00991, which I will present at the CAIS conference caisconf.org/ next week.

New type of insult: You write like an LLM😓

Sadly, I must say that Opus 4.7, as observed by many here, has become too annoying to work with. GPT 5.5 has become noticeably better, but it still not satisfying for hacking the compiler.

Thought @TheOfficialACM had uniform publishing standards across venues. Nope. ACM CAIS camera-ready: rip out biblatex (which acmart officially supports??) other packages which entails significant rework for no reason. SIGPLAN doesn't do this. All for an HTML view nobody wants.

Just gave a my talk: "How Can We Trust Our Agents?" at Scalar conference. Slides: slideshare.net/slideshow/how…

Good news: we'll wipe contributors.scala-lang.org and users.scala-lang.org clean and make it an LLM-only forum. We've already set up two agents that will argue for resp. against braceless syntax. Scala marches on!

Claude Pro vs Codex Pro on Scala compiler: Claude is extremely good at Scala 3 and synthesizes idiomatic code, understands capture checking very, very well! Codex: less elegant code, requires more handholding, but still fairly competent. It flips on pencil-and-paper math.

We're running the 2026 #Scala Survey! Let us know how you are using Scala, #AI with Scala, and give hints to the Scala teams (both at @VirtusLab and @EPFL) on where the focus in driving Scala forward could be. Thank you - your answers count :) virtuslab.typeform.com/Scala…

After a long freeze, the Scala 3 standard library is again open to contributions. The main place for contributing is now the Scala 3 repository. details and link to process document: scala-lang.org/blog/2026/02/…

Coffee’s good, but living rent-free in someone’s head? Delicious.

Congratulations @odersky for receiving the SIGPLAN Programming Languages Achievement Award! Your work is a great inspiration for me :) Well deserved!

We are thrilled to release Velvet: a foundational multi-modal verifier for imperative programs in Lean. It unifies execution, testing, automated and interactive proofs; and is itself proven sound. 💻 github.com/verse-lab/loom 📄 verse-lab.github.io/papers/l… 🧵 Learn more below ↓

I used to be skeptical of whether Martin's Caprese project is going to be able to bring much value to Scala, given how polished and well-designed the language already is. After Scala Days talks from his team I am now quite convinced the direction taken makes a lot of sense. I think that the standard library, for example, will make a lot more sense with capture checker guarding the less safe parts. For example, it is now my hope that separation checking will be able to make iterators safe from invalid use, even when they are returned from collection methods silently (a link to blogpost with the description of this problem in the reply). What is quite surprising to me is to see that some people still think that Caprese is about building some kind of alternative to existing purely functional libs. It seems more and more obvious to me that the objectives of this research are about building type-level tools to build new, safer and performant solutions that are not forced to rely on immutability and monadic suspension to keep us from making mistakes in our implementations. The one bit that is the most promising to me is the separation checking functionality that will yield linear typing and thus, along with capture checking, allow true and actual resource and mutability type safety that cannot be provided by any existing solution. I'm not gonna lie, Scala Days talks made me a believer, even if only a little. It seems to me that we can have nice things. I also agree now with Nicholas Rinaudo, who wrote that it seems the LAMP team will be able to deliver this improvement in a way that won't break everything for everyone. Kudos to Martin and his team, I'm now officially hyped.

Shoutout to @KagiHQ for their paid search! I've never had to look beyond the first half of the first page to find what I wanted.

Update! The Scala Workshop 2025 submission deadline is now July 18, AoE (11:59 PM UTC-12). More time to submit your talk proposals! Original: x.com/scala_workshop/status/…