Sujoy Mukherjee

The newsletter on Monday highlighted a launch that centered on agent teams that could be opened up to find out where things had gone wrong. This sentence has stayed with me all week.

Most of the demonstrations with agents focus on the final output. But the part that interests me is the failures. If an agent chain produces a wrong answer, then I have no clue where along the way something has gone wrong: was it during the retrieval, the function call, or an incorrect assumption made three hops back?

Not looking to stir up trouble here, but genuinely asking. The first version of my app was mostly put together with prompts, and it took about four hours to have a functional demo. Then came three weeks of fixing bugs that the prompt-based approach caused, like data type mismatches, weird behavior when form fields were empty, etc.

I suppose adding up total time spent, it may not have been quicker than if I had written the whole thing from scratch, though it certainly felt quick at the time because early success was very tangible.

The hero screenshot has been swapped for a fifteen second, unmotivated clip of a single task performed in real time, no zooming, no narration, no music.

Sign-up from this page was markedly improved. The slick two-minute video, which we paid for last year, remains on our about page and is rarely viewed.