Perplexity AI CEO Sputters When Asked to Define "Plagiarism" Onstage.
As TechCrunch reports, Perplexity chief executive Arvind Srinivas seemed speechless when asked to define plagiarism — which is telling, considering that Perplexity has repeatedly been accused of it in recent months.
https://www.msn.com/en-us/money/other/perplexity-ai-ceo-sputters-when-asked-to-define-plagiarism-onstage/ar-AA1tkUhY?cvid=6ad4c3f94526435cb82a38a5b245de9d&ei=167
/nosanitize
#AI
It sounds like he froze for a moment, but then got around to the right answer. Perplexity does not present the information as its own, and it is the only A.I. I know of that actually cites its sources.
@ceorl I haven't been following the story, but this looks as if the claims have merit.
https://www.wired.com/story/perplexity-plagiarized-our-story-about-how-perplexity-is-a-bullshit-machine/
@ceorl It's complicated.
https://techcrunch.com/2024/07/02/news-outlets-are-accusing-perplexity-of-plagiarism-and-unethical-web-scraping/?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuYmluZy5jb20v&guce_referrer_sig=AQAAAE8oUFZaYi6SGnyMTWcT4EGwTjF81cN5jXpYbajZIUIY3Gtx56GaXmLOv7yUDG08_BqTD0NG0MlabrRxm9Tt1ZlWOvc94iC20AHhzlr3sKMlYGYMa1H48bLw5LMbHOWlDlcK4T4iZj8X6RSSFuvyKEWIQBVBLzM3DL5WZhYBXInT
/nosanitize
It is complicated :)
Yes, I have mostly been impressed with their citations and the usefulness of them. I've even read the source articles and correct them on their interpretations of them.
I don't approve of rogue spiders crawling sites that don't want it, and not citing the sources, but I think the other A.I.s that don't cite any sources are much closer to actual plagiarism.
Its more Perplexity is scraping websites without permission.
https://counter.social/@ecksmc/112668053984814066
they are ignoring the Robots Exclusion Protocol - thats the real issue
Yes I know. What is so sad to me is that runs counter to their stated policy, which I think is a pretty good approach.
Maybe it is an honest misconfiguration, maybe it is deliberate. I guess the courts will have to settle that.
I'm not even sure that ignoring a robots.txt file is illegal.
Don't think it is illegal more frowned upon "its not cricket old chap" 😆
I don't think its a misconfiguration either they know what they are doing - but still one of the better AIs out there for actual upto date info
From what I've read, perplexity is using a third party web crawler, so it's not so much an A.I. issue, but rather a minor technical issue, and and a matter of ethical willingness.
There's a way, but is there a will :)
@peterquirk @ceorl lol Perplexity crawled Condé Nast property? CN signed a contract with OpenAI. Lawyers at dawn, pistols drawn.
Condé Nast joins other publishers in allowing OpenAI to access its content.
https://arstechnica.com/information-technology/2024/08/openai-signs-ai-deal-with-conde-nast/
@peterquirk
I'm still not sure that plagiarism is the right word, presenting someone else's content as your own without giving credit. This seems more like copyright infringement or just scummy behavior, but not plagiarism.