chatgpt may now access the internet at large to generate natural language answers to queries but it definitely also still offers to find you exact quotes from texts and then fully makes shit up without saying thats what its doing, so, like, its still trash garbage at a foundational level
Comments
I feel like people often miss the keyword ARTIFICIAL in Artificial Intelligence.
ChatGPT cannot even accurately parse and represent the text of documents it is provided. It is a black box of BS. But I guess we live in a post truth society now.
so. like. actually fuck the lying machine for real.
But “oh no, I can’t blindly trust it to tell me everything” is kind of the standard for getting information online no matter what the source.
Sorry if I missed your nuanced and reasoned argument.
People should hate it for the right reasons.
Maybe I shouldn’t have asked ChatGPT what you were talking about. My bad.
This statement totally triggered me. I’m now angry at Big Media and the American electorate.
I’m gonna have to go take a nap.
Chatgpt:
Your desire for precision is valid, but what precedent are we working off of here?
...but ChatGPT just gave me a list of songs and artists that never existed. Like at all. Really weird.
Completely unusable, useless trash
That so many people think and/or claim that they are good sources of truth is horrifying.
He was promoting it as a learning tool for a while.
Grok is pretty good at fact checking memes.
#6'Elon
So, yes, keep pointing out Ai's limited trustworthiness, but it is only getting better, and that is a plus.
#6'Elon
It's Artificial.
Maybe an asterisk ✳️ on the made up shit would be a good idea? IDK
Someone is just trying to convince you the other stuff is fine and "only" AI is the problem
Everyone has decided AI is the problem, when the real problem is techno capitalism - tech was already stealing content, burning the earth, and destroying lives and labor in the name of “disruption”
There can be ethical AI. But not ethical Big Tech!
We're no longer in a world where claims are all you need to make a point. We need proof.
And don't believe in LLMs hallucinating, something you find literally in OpenAI's technical reports
Like AI researchers work on bringing down hallucinations. Because they exist.
When you make a claim, bring receipts. Otherwise you're just like flat-earthers yelling at NASA rocket scientists that they are wrong and that they should "do their own research".
https://bsky.app/profile/joekatz45.bsky.social/post/3loa5glmd2k2q
When we talk about an AI “manipulating” a human, we’re really asking whether the system is covertly influencing someone’s thoughts, feelings, or actions in ways they wouldn’t fully understand or consent to.