Jeremy D. Miller

I'm genuinely worried that the AI + LLMS friendly documentation is going to kill off the service model I had been depending on, and I've been getting some anecdotal evidence that that's a legitimate concern.

Then I got a slew of Kubernetes related problems and an unusually challenging messaging scenario from a 3rd client in less than an hour.

Mar 10 at 2:25pmWeb
Show threadNeal WebbMar 10
@jeremydmiller this is exactly what Tailwind ran into with their paid components. If the base product is easy enough to use and well documented, the LLM can be guided to fill in the middle and recreate the paid portion. That said, I wouldn’t completely write off your paid services, at least not yet. If people yolo a bunch of code trying to recreate what you’re offering, the question then becomes, how confident are they in supporting that LLM generated code?
Show threadNeal WebbMar 10
@jeremydmiller if it was generated slowly, with a human in the loop at each step, with careful review, there might be enough confidence there to support that code going to prod. If they just went hog wild though, that starts to feel sketchy (to me at least).
Show threadJeremy D. MillerMar 10

@nealwebb I don't think any of us know how it's all going to play out. It's clear to me that I need *some* sort of AI play one way or another, even if that's just marketing.

But I can think of plenty of things to hold back from the docs. And that by itself also makes me uncomfortable.