this post was submitted on 28 Sep 2024
432 points (98.2% liked)
Technology
59982 readers
4137 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- 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, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
IMO it's even worse than that. At least from what I gather from the AI/Singularity communities I follow. For them, AGI is the end goal - a creative thinking AI capable of deduction far greater than humanity. The company that owns that suddenly has the capability to solve all manner of problems that are slowing down technological advancement. Obviously owning that would be worth trillions.
However it's really hard to see through the smoke that the Altmans etc. are putting up - how much of it is actual genuine prediction and how much is fairy tales they're telling to get more investment?
And I'd have a hard time believing it isn't mostly the latter because while LLMs have made some pretty impressive advancements, they still can't have specialized discussions about pretty much anything without hallucinating answers. I have a test I use for each new generation of LLMs where I interview them about a book I'm relatively familiar with and even with the newest ChatGPT model, it still makes up a ton of shit, even often contradicting its own answers in that thread, all the while absolutely confident that it's familiar with the source material.
Honestly, I'll believe they're capable of advancing AI when we get an AI that can say 'I actually am not sure about that, let me do a search...' or something like that.
I follow a YouTube channel, AI explained, that has some pretty grounded analysis of the latest models and capabilities. He compared LLMs to the creative writing center of the brain, as in they're really nice to interact with, output things that sound correct, but ultimately are missing the capabilities of reasoning and factuality that are needed for AGI
He's the best for unbiased info when new models drop and news drops. Love that channel.
He's really good. I'm torn on the subject because the current AI hype is most certainly a bubble and a grift, but I find the technology fascinating. I do think there's potential for great things there, but the technology is almost exclusively in the hands of 3 companies and will have a terrible impact on everyone else. I enjoy just focusing on the technical details every once in a while
I feel the same way man, I'm so excited about this tech because of the few use cases that we will discover and will change our lives and cause a paradigm shift...but it will be controlled by someone who is ultra rich (a person or corp) so that's awful. And there is a ton of stuff that is a grift that will die off and gives it a bad name so I temper my excitement of it around most people because it's unpopular to root for the tech.
Every LLM answer is a hallucination
The hallucination is in the mind of the user – people fall for the illusion of talking to a creature.