this post was submitted on 03 Oct 2026
472 points (98.2% liked)
Technology
88559 readers
3046 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I mean, it literally does though. The email sounded phony AF, especially knowing that the person's first language isn't English and it was overly proper sounding. I copy and pasted it into chatgpt and said "did you write this email?“ and it's response was" yes, based on the writing style and phrasing, I wrote this content."
I've been using ChatGPT a bit at work lately, since they've already paid for the subscription for everyone. I've learned that it is a bad idea to ask it any leading question, because it is likely to just give you the answer you want to hear.
Someone put in the intro to the book Frankenstein into an ai checker and it confirmed it to be 100 percent written by an LLM. It's not deterministic.
This is the first thing about LLMs always, and what most people can't seem to understand.
LLMs are a thing where you give an input and get an output, and it happens to correspond to reality much of the time. But there's no actual understanding there.
Exactly why it needs code to enforce repetitive results.
It does not remember things it made, so no. It doesn't work that way.
It doesn't remember but that won't keep it from answering a question as to whether it thinks an email was AI generated. It'll also be happy to confidently answer a question with a wrong answer to.
Doesn't work that way. Only if they programmed LLMs to sign the work like using a hash you would be able to confirm or deny with 100% accuracy