ChatGPT
-
The interesting question is whether it leans liberal only because the data set on which it was trained leans liberal, or if there was some intentionality behind it.
The selection of which data to train it on was likely biased.
Not necessarily. There are plenty of other ways to introduce bias in an AI model.
-
... popular authors including John Grisham, Jonathan Franzen, George R.R. Martin, Jodi Picoult, and George Saunders joined the Authors Guild in suing OpenAI, alleging that training the company's large language models (LLMs) used to power AI tools like ChatGPT on pirated versions of their books violates copyright laws and is "systematic theft on a mass scale."
-
https://www.theguardian.com/film/2023/oct/02/tom-hanks-dental-ad-ai-version-fake
Tom Hanks says AI version of him used in dental plan ad without his consent
-
Let's pull this thread a bit further. We know the AI (deep fake) videos are here and will only get better, and they aren't going away. What if we also had AI-faked signatures on contracts that lie about the celebrity's contract to do the fake ad? Dangerous times we have entered.
-
Let's pull this thread a bit further. We know the AI (deep fake) videos are here and will only get better, and they aren't going away. What if we also had AI-faked signatures on contracts that lie about the celebrity's contract to do the fake ad? Dangerous times we have entered.
One solution is to insist on the contracting parties signing physical documents in blood. That way you get physical, biometric proofs right there.
-
-
That way you get physical, biometric proofs right there.
A while ago, I read about a company that was promoting pens that contained DNA within the ink to prove the signature was valid.
I believe that their original idea was to market it to people who were famous enough to sell their autographs or things like that.
-
@taiwan_girl said in ChatGPT:
That way you get physical, biometric proofs right there.
A while ago, I read about a company that was promoting pens that contained DNA within the ink to prove the signature was valid.
I believe that their original idea was to market it to people who were famous enough to sell their autographs or things like that.
Nathan Tardiff does the same thing with an $8 bottle. Has been for years. Each one has unique markers that don't break down over time.
He makes a red that literally binds to the celluloid cells of the paper. It's pretty damn impossible to remove the ink.
-
She took our jobs!
https://fortune.com/europe/2023/11/23/spanish-influencer-agency-earned-11000-ai-model-posers/
Spanish modeling/influencer agency created AI generated model to do the jobs of models and social media influencers, because they find real life influencers are too costly and too unreliable to work with.
-
Go to the 5:15 mark for the ChatGPT bit:
Link to video -
https://gizmodo.com/google-search-ai-overview-giant-hallucination-1851499031
Google Search Is Now a Giant Hallucination
Article with many examples showing failures in AI generate “summaries” provided by Google in response to search queries.
-
-
No idea if ChatGPT or some other generative AI is involved.
Link to video -
Good educational introduction to Large Language Model (LLM):
Link to video
