With AI generated articles, spam, research, images, tutorials all creating halucinations your AI generated x will be no better and will only create more untrustworthy nonsense that will only clutter the homemade communities, posts, and comments here.

Personally I'm sick of this AI crap that's plastered everywhere like microplastics and PFAS chemicals. The cons most certainly outweighs the pros and the 'tech panaceas' like blockchain, nfts, and llms mostly serves to prop up the rich class.

For the 2 years I've been here the new fediverse feels like what Reddit used to be and should have been before the enshitification and honestly I love it here. So please stop. We don't need a Reddit 2.0

you are viewing a single comment's thread
view the rest of the comments

[–] lambalicious@lemmy.sdf.org 1 points 1 day ago (1 children)

Assuming it's right.

99% of the time it's hallucinating, inventing two extra authors for the text (each with six fingers), misattributing a quote to at least one of them, and n otoriously missing the point of the article in favour of developing a more click-baity summary.

[–] muntedcrocodile@lemm.ee 2 points 20 hours ago

Go read the source and review the model page on huggingface and u will find thats not at all true for my implementation. The model im using provides a summary 60% identical to that of a human summary and >95% accurate in maintaining the original meaning.

Plus 99% is one hell of a an exaggeration.