I'll elaborate.

[gargron@mastodon.social]

RE: https://mastodon.social/@Gargron/115737483111161917

I'll elaborate. I believe we're in an AI bubble. AI companies are pushing the overton window on AI discourse. They want the most extreme anti-AI sentiment to be "Sure, it's an overhyped technology right now but there will be reasonable applications down the road". I want to see pushback on this. It doesn't matter if I can disable some AI feature in the settings. The fact that Mozilla is jumping on this bandwagon is deeply disappointing to me.

[xinit@mastodon.coffee]

@Gargron Just like with blockchain, so much of this AI push is a "solution" in search of a problem.

[colman@mastodon.ie]

@xinit @Gargron LLMs and associated technologies have useful applications, just not any of the ones being pushed by big “AI”.

[xinit@mastodon.coffee]

@Colman ethically trained LLMs could be useful in some instances, but those don't seem to exist.

ML > LLM

@Gargron

[xchaos@f.cz]

@xinit I thing text shortening - excerpts, abstracts, and so on - may be essential for effective search indexing. Basically, removing extra information is like doing image thumbnails. It is increasing resolution and trying to make up for the missing information, which is dangerous...

Also machine translation seems to be relatively useful information.

Whatever practical and ethical use there can be, it is not the one which leads to exponential demands for computing use.

@Colman @Gargron

[xinit@mastodon.coffee]

@xChaos Search was working GREAT with pre-LLM Google. You could actually FIND things. Now, with whatever these "AI summaries" are, Google is effectively useless on its front page.

@Colman @Gargron

[memoria@wetdry.world]

@xinit @xChaos @Colman @Gargron