There is something poetic about the fact that an AI security agent found a critical vulnerability that an AI coding assistant introduced. Wiz Research's autonomous Red Agent identified and fully exploited a script-injection bug in one of Snowflake's GitHub Actions workflows, gaining access to an internal Jira instance. The bug had been sitting there for five days. It was introduced by GitHub Copilot Autofix in a pull request, which removed an existing sanitised input pattern and replaced it...
15 days ago • 5 min read
There is a lot of breathless writing in this industry about how AI is a great leveller, a tool that erases the gap between the expert and the beginner. Sean Goedecke's post LLMs reward expertise — which was Hacker News's top story this week — argues the opposite, and I think he's right. The quality of what you get back from a model scales with the quality of what you bring to the conversation. The gap between an expert and a novice wielding the same model is wider now, not narrower. That's...
29 days ago • 7 min read
A study from Capraro, Marcoccia and Quattrociocchi, written up by The Next Web, gave people AI assistance on a reasoning task and watched their accuracy fall from 27% to 9%. Confidence went the other way, from 30% to 76%, and willingness to say "I don't know" collapsed from 44% to 3%. That gap between being wrong and feeling right is the same failure mode we keep building into our agents, and the mood is showing elsewhere too: Alec Scollon reckons he has LLM burnout, Yael Grauer would like...
about 1 month ago • 5 min read
The AI Engineer World’s Fair was in San Francisco last week. No, I wasn’t there, and yes, I was feeling all of the FOMO. If you’re in the same position, then you can catch up with what went on with Latent Space’s write-ups from the event and subscribe to the YouTube playlist as it starts to to publish the talks. You can also check out the talks from the online track while you wait. In this edition of AI++ we take a look at the skills it takes to build great Skills, get hands-on with some...
about 2 months ago • 3 min read
When we last published AI++ Anthropic had just launched Fable 5 and everyone was very excited about another step-change in the quality of models. That excitement didn’t last long as the US government issued an export directive that effectively meant Anthropic had to withdraw access. If you missed out, Ethan Mollick wrote about what it was like to work with Fable 5. I found two interesting looks at the world of AI this week. First, Anthropic published the results of 81,000 interviews with...
2 months ago • 4 min read
Claude Mythos is here, except it’s called Fable 5 and comes with a few restrictions. It appears to be the largest model released and, according to the benchmarks, the most accomplished. Even more so than Opus 4.8 that was only released 2 weeks ago. It’s also the most expensive model, so you might want to think twice before swapping it into your RAG support chat bot. While it’s impressive to see the frontier march forward, this week in AI++ we’ll take a look at some of the techniques people...
3 months ago • 3 min read
The last couple of weeks has seen students booing commencement speakers at graduation ceremonies in Florida and Arizona when they mentioned AI. This is a visceral reaction to what they see as a threat to their careers. Meanwhile, developers working with AI are burning out and getting “Brain Fry” from doing more work at higher intensities without the same fulfillment. I wrote a bit about this myself, sharing that I found it hard to be proud of a useful little app that I built. AI is changing...
3 months ago • 4 min read
Working with LLMs is weird, but I never thought it would be as weird as OpenAI having to specifically tell their models not to talk about goblins, gremlins, raccoons, trolls, ogres, or pigeons. It raises so many questions. Thankfully after someone spotted the instructions in the Codex base instructions, OpenAI did give an explanation as to where the goblins came from. They never mentioned why raccoons and pigeons got caught up in the fantasy creature fascination though. In this edition of...
4 months ago • 3 min read
Is a token crunch coming? This week GitHub paused sign-us for GitHub Copilot Pro, Pro+ and Student plans, tightened up their usage limits, and removed Opus from their Pro plans. And today, Anthropic seemed to remove Claude Code from new Pro plans, though that has been reversed quickly. In general, while this is only seeming to affect individual plans related to coding agents, it could point to an inflection point where AI companies start considering how their pricing matches up to their...
4 months ago • 3 min read