Fable 5 is another step up the abstraction ladder on the long journey from “ai is better autocomplete”. We need to yet again throw out old assumptions about how to work with LLMs. Old prompt techniques that worked well before are now too constrictive. It’s a different animal entirely, and using it well means unlearning some habits. Here I’ll describe how I’ve been using it, what works and what still doesn’t. ...
Alex Graveley, the guy behind GitHub Copilot, showed me how I’d been thinking about agentic coding wrong. He used pen and paper. Alex dropped by our Mission Street office on a Tuesday. I told him we had about a week of work left to get self-serve working. “No. That should take 4 hours.” He pulled out a sheet of paper, wrote down ten steps, and handed it to me. That was the plan. ...
We’ve lived with like ChatGPT for over a year now. Last year seemed to be the height of the hype cycle, at least for me. It felt like everyone, from friends to Uber drivers, was excitedly discussing the potential of this groundbreaking technology. However, as often happens with new technological advancements, the initial excitement has dimmed somewhat when faced with the reality of day-to-day use. The trough of disillusionment is real. In this article, I’ll mostly discuss the limitations of LLMs in technical tasks, though I believe these limitations apply to any sufficiently complex domain. ...