Research and news on AI alignment, value learning, and scalable oversight.
11 archived stories · Last updated Sep 14, 2026
This page is background reading from an earlier news archive. MisalignAI now focuses on practical AI safety tools that run in your browser.
Advertisement
I don't think this is particularly impressive or interesting for anyone else, but I think it may turn out to be useful in the future to have an easily visible public record of what happened, so here goes: I am an L5 SWE at Google Israel. I have been there since May 2021, was promoted once [1] , and...
Note from Austin: The EA Community Choice was an experimental funding round we ran in Fall 2024. We had 340 users help us distribute $200k to various community projects; $100k via a sliding-scale airdrop, $100k via quadratic funding match. See more in the launch post . Now, as a huge wave of...
I've just started to rewatch the film, "Executive Decision" after a long enough time to have forgotten most of it. There is a short moment early on where a news broadcast is shown for exposition purposes to explain the background of what is going on to set up the premise of the story. What struck...
TL;DR: We ( Sage , a nonprofit) built Talleyrand, a free, open-source AI workspace for thinking through hard questions. Your questions are kept in a tree, not a linear chat. After each answer it suggests the next questions and learns from your reactions. Try it at talleyrand.app . Why we built it I...
This is a research update for an on-going replication of no-CoT [1] evals done as part of the Second Look Fellowship . In following posts, we will run more comprehensive replications of previous work and release open source tooling for no-CoT eval elicitation. Code can be found here . tl;dr We...
(Originally published on No Set Gauge June 24th 2026.) Rembrandt, Saul and David When people talk about AI being “aligned”, I think there’s a lot of conflation between two different bars of success: The AI does what you say and does not go rogue. If you ask it to do a machine learning experiment...
On Equity, we discussed the AI industry's latest debate about whether it poses an existential threat to humanity.
An Anthropic researcher resigned this week, warning in a post on X that the company is “racing straight to self-improving superintelligence and gambling with our lives”. The company’s own alignment lead even co-signed the message rather than walking it...
The CoTControl eval asks reasoning models to follow formatting constraints in their chain-of-thought (e.g. write in all lowercase, avoid a specific word) while solving questions. Models seem to mostly be pretty bad at this: recent models score between 0-30% with the exception of Mythos Preview [1]...
GPT‑6 Astra improves Devin’s ability to test software and show that it works, with the goal of helping engineers review less code and ship more.
Learn how OpenAI evolved Habitat from a Python library into a globally distributed storage platform serving 1 billion ChatGPT users and 22M requests per second.
Advertisement