We have to stop ignoring AI’s hallucination problem
We have to stop ignoring AI’s hallucination problem
www.theverge.com We have to stop ignoring AI’s hallucination problem
AI might be cool, but it’s also a big fat liar.
You're viewing a single thread.
207 comments
-
Don't let perfect be the enemy of good.
-
Generative AI is hardly "good" yet, either morally or as a product.
-
Yes wow its so not an impressive discovery at all, obviously for sure 1000%
-
It's very useful as a product, for creating images, stories, poems, code. Just because you don't use it doesn't mean it isn't good.
-
I use it for sure, I even pay for Gemini for its creative writing capabilities, but most LLMs are bad at many tasks they're advertised to be good at (coding being one of those things), plus they're largely based on stolen work and/or copyright infringement. They don't reliably do what they're claiming, and they are unethically developed. Hence, they're bad products, just objectively.
-
Are you also against copyright infringement when it's piracy? Because I'm not, so it would be inconsistent to judge open source generative AI models for doing something I don't consider wrong.
-
I consider piracy wrong when companies are stealing from creatives (like authors whose books are included with no credit or royalties) for the purposes of profit. I don't believe all piracy is always good full stop. I believe piracy is ethical if it allows for preservation of content that may otherwise not be preserved or maintained.
Also that was just one of my points lol. Most LLMs are still just bad at what they are claiming to be able to do.
-
-
-
-
-
"good"
-
207 comments