Ask Lemmy
A Fediverse community for open-ended, thought provoking questions
Please don't post about US Politics. If you need to do this, try !politicaldiscussion@lemmy.world
Rules: (interactive)
1) Be nice and; have fun
Doxxing, trolling, sealioning, racism, and toxicity are not welcomed in AskLemmy. Remember what your mother said: if you can't say something nice, don't say anything at all. In addition, the site-wide Lemmy.world terms of service also apply here. Please familiarize yourself with them
2) All posts must end with a '?'
This is sort of like Jeopardy. Please phrase all post titles in the form of a proper question ending with ?
3) No spam
Please do not flood the community with nonsense. Actual suspected spammers will be banned on site. No astroturfing.
4) NSFW is okay, within reason
Just remember to tag posts with either a content warning or a [NSFW] tag. Overtly sexual posts are not allowed, please direct them to either !asklemmyafterdark@lemmy.world or !asklemmynsfw@lemmynsfw.com.
NSFW comments should be restricted to posts tagged [NSFW].
5) This is not a support community.
It is not a place for 'how do I?', type questions.
If you have any questions regarding the site itself or would like to report a community, please direct them to Lemmy.world Support or email info@lemmy.world. For other questions check our partnered communities list, or use the search function.
Reminder: The terms of service apply here too.
Partnered Communities:
Logo design credit goes to: tubbadu
view the rest of the comments
That would require an actual AGI to emerge, which it has not and is not going to. LLMs are fancy text prediction tools and little more.
What we see in AI as an average consumer is like the RC hotwheels to a state of the art tank being used by big corps.
Just imagine that if an early LLM can fool an engineer into thinking it's sentient, what a state of the art system can do, one designed to predict the market, run propaganda bots on social media or straight up manufacture news stories with the footage to back it up.
The AI being used by big corporations is so advanced, it's one of the reasons countries have been trying to digitally isolate themselves. It's really not an if, it's a when.
I'm not sure you understand what AGI is, and why we're not going to invent it any time soon.
I do. I did get a little lost in the weeds with my point though, as I was talking in a more general sense about how AI is already powerful and dangerous - because AI safety is a subject in this thread.
The "AI" being used by big corporations is still fundamentally an LLM and has all the flaws of an LLM. It's not a hot wheels car vs a tank, it's a hot wheels car vs a $2 billion RC car
I'd like to get into how both me and OP are talking about how fast AI, not just LLMs, is scaling, and the potential it has across a variety of industries - most concerning to me is it's use by investment firms. But I need to go to the barber because I already have enough split hairs.
It is my understanding that the fundamental architecture (the general purpose transformer) is identical between the "AI" used by Black Rock and by OpenAI
If you have some evidence to the contrary I'd always appreciate the chance to learn.
But the transformer based architecture is fundamentally flawed: it will always hallucinate.
Are you assuming LLMs are the only way humans could ever try making an AGI? If so, why do you assume that?
There's more important shit than worrying about if an unproven sci fi concept will come to being any time soon.
Yeah, agreed. That’s not what I asked though.
This response is a bit of a misdirection since we all discuss shit that isn’t the most important all the time.
I agree that AGI is dangerous but I don't see LLMs as evidence that we're close to AGI, I think they should be treated as separate issues.
Given what I think I know about LLMs, I agree. I don’t think they’re the path to AGI.
The person I replied to said AGI was never going to emerge.
I had meant to say AGI would never emerge from our current attempts at creating them.
If people start developing a new more promising kind of "ai", we can talk about it ðen. For now, ð þing we call "AI" sucks and just steals.
So you're confident that AGI is not fundamentally possible? That would contradict basically every single scientist in the world and this is exactly why this issue is so difficult. Ironically, proving my point for the OP's question lol