Experts aren’t unanimous about whether the AI-powered search startup’s practices could expose it to legal claims ranging from infringement to defamation—but some say plaintiffs would have strong cases.
New models are like really good at context, the amount of input that can be given to them has exploded (fairly) recently... So you can give whole datasets or books as context and ask questions about them.
Perplexity Plagiarized Our Story About How Perplexity Is a Bullshit Machine ( www.wired.com )
Experts aren’t unanimous about whether the AI-powered search startup’s practices could expose it to legal claims ranging from infringement to defamation—but some say plaintiffs would have strong cases.
We have to stop ignoring AI’s hallucination problem ( www.theverge.com )
OpenAI strikes Reddit deal to train its AI on your posts ( www.theverge.com )
iPhone owners say the latest iOS update is resurfacing deleted nudes ( www.theverge.com )
cross-posted from: https://sopuli.xyz/post/12670977...
What it's like to be a developer in 2024 ( sopuli.xyz )
Source
That Nim Flashbacks ( sh.itjust.works )