Re:Everything around LLMs is still magical and wishful thinking
by bgwalter on Friday July 04, 2025 @06:01PM
(#44468207)
This one is funny though. I'd bet on Bing CoPilot, which now always agrees with "AI" concerns because MSFT has probably realized that no one wants "AI" and takes a more cautious approach.
Re:Everything around LLMs is still magical and wishful thinking
by careful_ai on Friday July 04, 2025 @06:08PM
(#44468266)
Haha, if Bing Copilot could thread this much nuance into a comment, I’d hire it for my job. For now, just a human who’s been burned by LLM hype cycles one too many times.
Re:Everything around LLMs is still magical and wishful thinking
by careful_ai on Friday July 04, 2025 @06:07PM
(#44468258)
Didn’t think using punctuation and full sentences would get me flagged—but I get it. In a world where AI writes like humans and humans write like bots, I guess tone becomes a Rorschach test. Just sharing thoughts I care about, with zero automation involved. If something I wrote sounded “AI-ish,” that’s on me for being too polished. Feedback taken, and appreciated.
Re:Everything around LLMs is still magical and wishful thinking
by gjm11 on Friday July 04, 2025 @06:45PM
(#44468559)
For what it's worth, I had a similar "that looks like AI writing" response, and it wasn't because it was "too polished". And having looked at the rest of your comment history, the only reason why I'm only at 90% confidence it's all AI-generated rather than 100% is your explicit claims to the contrary. Today's LLMs have a definite style that is, sorry, not the same thing as being "polished", and if your comments have "zero automation involved" then it's quite the extraordinary coincidence how much more like an AI you sound than any other human writer I have ever encountered. And a further coincidence that this very AI-sounding human just happens to be selling services to "unlock Meaningful Business Outcomes With AI".
Re:Everything around LLMs is still magical and wishful thinking
by Jordan-117 on Friday July 04, 2025 @06:35PM
(#44468484)
Reaction to a post at 21:39:36:
https://news.ycombinator.com/item?id=44468067
Reaction to a different post at 21:40:30:
https://news.ycombinator.com/item?id=44468069
Fast typist! (Incidentally, both are exactly 59 tokens long)
Re:Everything around LLMs is still magical and wishful thinking
by farts_mckensy on Friday July 04, 2025 @06:00PM
(#44468197)
Watch out, HN. The em dash police are here. Hands up
Re:Everything around LLMs is still magical and wishful thinking
by careful_ai on Friday July 04, 2025 @06:09PM
(#44468279)
LOL. I’ve clearly overused my em dash quota this week. Noted. Will switch to semicolons next time; they need love too.
- cliché phrasing
- em dashes
- abundant alliteration
- all comments suspiciously similar in length
- all posts pointing to the same website
Does HN not have a policy against vapid AI comment spam? If not, it needs one.
edit: It does:
https://news.ycombinator.com/item?id=37617714