this post was submitted on 15 Jun 2024
33 points (59.8% liked)

Technology

59669 readers
3168 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 1 year ago
MODERATORS
(page 2) 40 comments
sorted by: hot top controversial new old
[–] [email protected] 2 points 5 months ago

So...GPT-4 is gay? Or are we talking about a different kind of test?

[–] mindbleach 2 points 5 months ago

Haha, no. A five-minute conversation is not a meaningful demonstration of the core concept. This is the same technology that cheerfully tells people to eat rocks. It's not even the newest and fanciest GPT. Everybody is aware of the failure cases for this approach. It does not take all that much interaction to separate an LLM from AGI.

Remember, a buuunch of people have been using this thing, for a while now. It can do a lot of clever eerily clever and capable things. It can also read "a room with no elephants" and draw a room with seventeen elephants, because you said the word "elephants."

The real point of the Turing test is to highlight that interaction is the only way we discern intelligence in other humans. If you can keep probing and never tell the difference - then the machine is at least as intelligent as the meatbag behind door number two. But it doesn't take much to fool people. Eliza is a relic and still had chins being stroked over PDP-11s "passing the Turing test."

You will know when an AI becomes AGI, because all of marketing's stupid promises will actually work.

[–] [email protected] 1 points 5 months ago

So it's good enough to fool people into thinking it's a human?

Cool. Anyway...

[–] [email protected] 1 points 5 months ago (1 children)

It does great at Python programming.... everything it tries is wrong until I try and I tell tell it to do it again.

[–] [email protected] 1 points 5 months ago* (last edited 5 months ago) (1 children)

Edit :
oops : you were saying it is like a human since it does errors ? maybe i "wooshed".


Hi @werefreeatlast,
i had successes asking LLaMA 3 70B with simple specific questions ...
Context : i am bad at programming and it help me at least to see how i could use a few function calls in C from Python ... or simply drop Python and do it directly in C.
Like you said, i have to re-write & test ... but i have a possible path forward. Clearly you know what you do on a computer but i'm not really there yet.

[–] [email protected] 0 points 5 months ago

But people don't just know code when you ask them. The llms fo because they got trained on that code. It's robotic in nature, not a natural reaction yet.

load more comments
view more: ‹ prev next ›