222
Reasoning failures highlighted by Apple research on LLMs
(appleinsider.com)
This is a most excellent place for technology news and articles.
You don't have to get all philosophical, since the value art is almost by definition debatable.
These models can't do basic logic. They already fail at this. And that's actually relevant to corpos if you can suddenly convince a chatbot to reduce your bill by 60% because bears don't eat mangos or some other nonsensical statement.
It's all connected, the reasons why it can't do basic logical reasoning are the same for why it can't replace human art.
It's because neither of those activities are mere pattern recognition and statistical inference, which is all LLMs will ever be.
LLMs and image generating models are completely different things. Outputting an image doesnt require or benefit from reason and logic (other than making the model "understand" the prompt). Drawing a three headed monkey isnt "logical" and doesnt follow "reason" but that's ok because art isnt about making photorealisitic images.
AI images could totally be useful as a tool in art. "But a computer made it! It's not art!" It's the same tired argument we heard about electronic music before.
But the fediverse seems to have such a hate boner for ANYTHING associated with AI (dont get me wrong, there is lots to hate. Mostly with tech-bro grifting...) that people are unable to see that these can be useful complements to human creativity.
Here's another example... People crying that when an image contains AI generated elements, or maybe a video game contains some AI assets. People fly into a rage and want to dismiss the ENTIRE work and throw it all out. Human art doesnt require 100% human hands to make. Go look at any famous painting by a renaissance master. Did you know a lot of these guys had whole workshops of lackeys filling in background details for them? Are we going to throw out all the raphael and rembrandt paintings because they had assistance from other uncredited people?
Same with AI. Why cant an artist spend MORE time on important details and let AI draw some happy little trees in the background?
I think you're reading too deep into what I was saying. Perhaps I wasn't being clear, my bad if so.
I'm not against AI tools to assist people's work. Using them for grammar/spellcheck, code completion and automated testing, artwork help for filling in repetitive background details/textures, automatically removing background details in pictures like dumpsters or people photo bombing, etc.
What I am against is the grifting, the near religious devotion by tech bros to AI replacing humans in all areas of life, and the fact that the groups and companies controlling almost all of the development of this tech are multi-billion/trillion dollar corpos that don't make all aspects of their tech open source and are 100% motivated by profit.
Sorry, my comment wasnt really directed at you specifically.... Just the fediverse general hate for all things AI
yours was the first that mentioned "art" which triggered me, lol
I think we are actually both on the same page... You have a reasonable view of the whole AI thing. It's rare on lemmy/mastodon
Thanks for your response. Yeah, I think the issue isn't the technology, it's who controls and owns it.
I doubt it would be anywhere near as controversial if it were all fully open source and run by public organizations and communities that were interested in bettering the human experience and reducing mundane work vs maximizing profitability.