2026-07-14 05:11 UTC
@TamsynUlthara@catodon.rocks
I'd say those things do not learn, they just have a fantamagic database from which some algorithms extract the most probable answer(s) from your question(s).
I'd say also they do not think, they just put not-so-random words one after the other, giving us the feeling of thought.
If I say or write something, I usually understand what I'm doing and can repeat and adjust that output (and outcome)
Replies (2)
-
@AAMfP@fosstodon.org 2026-07-14 05:11
@TamsynUlthara@catodon.rocks I've been asked to produce a dashboard with a chart, did it with LLM because as all the rest of the world we've also been asked to experiment these things. I have no idea what's there in that code. What if tomorrow I'll be asked to enhance that thing? And this enhancement changes the behaviour of existing code? How do I know? How can I rely on such thing?
-
@TamsynUlthara@catodon.rocks 2026-07-14 15:04
@AAMfP@fosstodon.orgIf I say or write something, I usually understand what I'm doing and can repeat and adjust that output (and outcome)How can I be sure? You're just telling me that you understand what you're doing. Why should I believe you? Do you see the problem?