this post was submitted on 14 Aug 2024
-62 points (17.0% liked)

Technology

59653 readers
2807 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
you are viewing a single comment's thread
view the rest of the comments
[–] Telorand@reddthat.com 3 points 3 months ago (12 children)

This all hinges on the definition of "conscious." You can make a valid syllogism that defines it, but that doesn't necessarily represent a reasonable or accurate summary of what consciousness is. There's no current consensus of what consciousness is amongst philosophers and scientists, and many presume an anthropocentric model with regard to humans.

I can't watch the video right now, but I was able to get ChatGPT to concede, in a few minutes, that it might be conscious, the nature of which is sufficiently different from humans so as to initially not appear conscious.

[–] UraniumBlazer@lemm.ee -5 points 3 months ago (11 children)

Exactly. Which is what makes this entire thing quite interesting.

Alex here (the interrogator in the video) is involved in AI safety research. Questions like "do the ethical frameworks of AI match those of humans", "how do we get AI to not misinterpret inputs and do something dangerous" are very important to be answered.

Following this comes the idea of consciousness. Can machine learning models feel pain? Can we unintentionally put such models into immense eternal pain? What even is the nature of pain?

Alex demonstrated that ChatGPT was lying intentionally. Can it lie intentionally for other things? What about the question of consciousness itself? Could we build models that intentionally fail the Turing test? Should we be scared of such a possibility?

Questions like these are really interesting. Unfortunately, they are shot down immediately on Lemmy, which is pretty disappointing.

[–] conciselyverbose@sh.itjust.works 5 points 3 months ago (5 children)

Alex demonstrated that ChatGPT was lying intentionally

No, he most certainly did not. LLMs have no agency. "Intentionally" doing anything isn't possible.

load more comments (3 replies)
load more comments (8 replies)
load more comments (8 replies)