Search: “research”

Meta announced Muse Spark today, their first model release since Llama 4 almost exactly a year ago. It's hosted, not open weights, and the API is currently 'a private API preview to select users', but

Self-improving agents isn’t a single algorithm - it’s a systems engineering problem involving: - eval data curation + maintenance - experiment design to battle overfitting - an update algorithm - huma

Some sober thinking about Mythos (full version with links at my newsletter): 1It’s probably not as bad as they say, as AI and cybersecurity expert @HeidyKhlaaf explains elsewhere (in a thread “As some

This is literally why I wrote Taming Silicon Valley. Director James Cameron on why Big Tech owning AGI is scarier than any science fiction he's ever made: 'AGI will not emerge from a government funded

In February 2019, OpenAI announced GPT-2, a language model trained on text from 8 million webpages to predict the next word in a piece of writing, capable of adapting to the style and content of a...

AI critic Gary Marcus uses the irony of ChatGPT's Voice mode being unable to perform a basic task — starting a timer — as a pointed illustration of the gap between AI industry hype and real-world c...

We're seeing even more autonomous AI coworkers. The new MLE agent on the market is Disarray. In Kaggle competitions, Disarray: - won 28 medals across diverse domains (vision, NLP, tabular data) - plac

Before limited-releasing Claude Mythos Preview, we investigated its internal mechanisms with interpretability techniques. We found it exhibited notably sophisticated (and often unspoken) strategic thi

Julien Chaumond (co-founder of Hugging Face) posted a tweet referencing the infamous 2019 OpenAI decision to initially withhold GPT-2-large from public release due to fears it was 'too dangerous,' ...

Mythos is very powerful, and should feel terrifying. I am proud of our approach to responsibly preview it with cyber defenders, rather than generally releasing it into the wild. Model card here: https

SuperClaude (Mythos) still seems irreducibly Claude-y given the transcripts in the system card. Here two versions of Mythos are forced to talk to each other across multiple rounds. They are less philo

Thank you to @AnthropicAI for sending FFmpeg patches Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software. It’s powered by our newest frontier model, C

PwC rolled out Copilot to employees globally. PwC reported $150 million in time savings from Copilot use in a Microsoft customer story.