> A lot of the claims of usefulness evaporate when tested In your personal experience? Because that's been my personal experience too, in lots of cases with LL

> A lot of the claims of usefulness evaporate when tested In your personal experience? Because that's been my personal experience too, in lots of cases with LLMs. But I've also been surprised the other way, and overall it's been a net-positive for myself, but I've also spent a lot of time "practicing" getting prompts and tooling right. I could easily see how people give it try for 20-30 minutes, not getting the results they expected and give up, which yeah, you probably won't get any net-positive effects by that.