Somhairle Macleòid
banner
penalosa.link
Somhairle Macleòid
@penalosa.link
How does a non-programmer/non-sme evaluate whether an AI tool has safely built a login system? Correctly set up database transactions?
March 28, 2025 at 5:49 PM
Even for those there’s an element of judgment (has the LLM added a tracking tag that inadvertently doesn’t fit with the GDPR? If you can’t read & understand the output you’d have no way of knowing) but for other use cases there’s much more potential for things to go drastically wrong
March 28, 2025 at 5:49 PM
This only really works for things where there doesn’t need to be any human judgement of whether the LLM output is any good (or safe). I can’t really think of any areas of programming that this covers other than pure frontend marketing sites (where the output is mostly self-evident)
March 28, 2025 at 5:49 PM
Oh amazing! I’d love to get wrangler in this—I’ll put up a PR!
March 28, 2025 at 1:07 PM
How did you make this graph? I’ve been looking for a way to do this to track version adoption
March 28, 2025 at 10:12 AM
👀
November 9, 2024 at 8:30 PM
How was [redacted]!?
October 23, 2024 at 7:43 AM