Comment on OpenAI says it’s “impossible” to create useful AI models without copyrighted material
lemmyvore@feddit.nl 10 months agoThis isn’t about scraping the internet. The internet is full of crap and the LLMs will add even more crap to it. It will shortly become exponentially harder to find the meaningful content on the internet.
No, this is about dipping into high quality, curated content. OpenAI wants to be able to use all existing human artwork without paying anything for it, so they can flood the world with cheap knockoff copies. It’s that simple.
towerful@programming.dev 10 months ago
Shortly? It’s happening already. I notice it when using Google and Duckduckgo. There are always a few hits that are AI written blog spam word soup
lemmyvore@feddit.nl 10 months ago
Unfortunately you haven’t seen the full impact of LLMs yet. What you’re seeing now is stuff that’s already been going on for a decade. SEO content generators have been a thing for many years and used by everybody from small business owners to site chains pinching ad pennies.
When the LLM crap will kick in you won’t see anything except their links. I wouldn’t be surprised if we’ll have to go back to 90s tech and use human-curated webrings and directories.
prex@aussie.zone 10 months ago
I wonder how many comments in this thread are ai generated. I wonder how many comments on Lemmy will be in 5 years time.
emptiestplace@lemmy.ml 10 months ago
It’s especially amusing when you consider that it’s not even fully autonomous yet; we’re actively doing this to ourselves.