Post #3161056
2025-12-07 01:16 UTC
Replies (3)
-
@micahilbery@mstdn.social 2025-12-07 01:20
Generative AI isn’t bad because it’s closed source. It isn’t bad because it’s made by bad companies. It’s bad because it was designed to “solve” the ultra wealthy needing other humans that dare work for pennies of that wealth to live. It was designed for bad intentions from the very beginning. There is no good generative AI no matter who tries to tell you otherwise. Yeah, generative AI is likely not going away, but don’t pretend that somehow makes it ok. It spotlights who all is bad. 🧵4/4
-
@aemstuz@mastodon.com.pl 2025-12-07 08:25
@micahilbery@mstdn.social E.g. you can imagine using LLM to automate social media moderation. You can imagine using LLM to filter out clickbait from news. Last but not least, you can use LLM to have code autocompletion on steroids. Yeah, I realize all the above problems have their root causes and we should fix these root causes first. But I doubt we will ever live in a world where all people are kind, nobody tries to trick you etc. We will also never get to the point where code doesn’t require some extra verbosity. I’m also concerned that people use the term “AI” when they mean “generative AI”. There’s a lot of AI, tailored for specific tasks (e.g. identifying tumors on CT scans). These may not require so many computing power to train and operate.
-
@shaedrich@mastodon.online 2025-12-07 13:46
@micahilbery@mstdn.social It's not even misinformation per se if we define it as "intentionally wrong information to harm or manipulate". Especially in software engineering, vibe coding tools like #GitHub's #copilot are fed low quality code from sites like #StackOverflow not because people were poisoning the well since 2008 but because coding is somewhat hard, so it needs time to master. And since AI isn't *understanding* things but just combining stuff it finds, it struggles so tremendously with it.