this post was submitted on 27 May 2024
1101 points (98.0% liked)

Technology

59374 readers
7113 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 know how Google's new feature called AI Overviews is prone to spitting out wildly incorrect answers to search queries? In one instance, AI Overviews told a user to use glue on pizza to make sure the cheese won't slide off (pssst...please don't do this.)

Well, according to an interview at The Vergewith Google CEO Sundar Pichai published earlier this week, just before criticism of the outputs really took off, these "hallucinations" are an "inherent feature" of  AI large language models (LLM), which is what drives AI Overviews, and this feature "is still an unsolved problem."

you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 222 points 5 months ago (17 children)

In the interest of transparency, I don't know if this guy is telling the truth, but it feels very plausible.

[–] [email protected] 126 points 5 months ago (14 children)

It seems like the entire industry is in pure panic about AI, not just Google. Everyone hopes that LLMs will end years of homeopathic growth through iteration of long-existing technology, which is why it attracts tons of venture capital.

Google, which sits where IBM was decades ago, is too big, too corporate and too slow now, so they needed years to react to this fad. When they finally did, all they were able to come up with was a rushed equivalent of existing LLMs that suffers from all of the same problems.

[–] [email protected] 8 points 5 months ago (2 children)

The snake ate it's tail before it's fully grown. The AI inbreeding might be already too far integrated, causing all sorts of Mumbo-Jumbo. Also they have layers of censorship, which effect the results. The same that happened to chatgpt, the more filters they added, the more it confused the result. We don't even know if the hallucinations are fixable, AI is just guessing after all, who knows if AI will ever understand 1+1=2, by calculating, instead of going by probability.

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

Even saying they're guessing is wrong, as that implies intention. LLMs aren't trying to give an answer, let alone a correct answer. They just put words together.

load more comments (1 replies)
load more comments (12 replies)
load more comments (14 replies)