Comment on The Butlerian Jihad is NOT a warning against AI
p03locke@lemmy.dbzer0.com 3 days agoI don’t think it can simply be repurposed for liberation, that’s a naive interpretation that ignores all of the actual ways in which the current implementations of AI work.
LLMs are a tool, and all tools can be repurposed or repossessed. Open source models and LLM tools are everywhere. Pretending that ChatGPT and other corporate models are the only ones out there are exactly what the corpofacists in power want you to believe.
By ignoring this technology and sticking our fingers in our ears, we are allowing them to reshape out the technology works, instead of molding it for our own purposes. It’s not going to go away, and thinking that is just as foolish as believing the Internet is a fad.
Are parts of it overhyped and overblown? Sure. Rich assholes have a knack for taking some concept and hyping it up to a million and under-delivering. But, the technology is still promising enough that it’s worth exploring where it can be beneficial, and it’s new enough that we can make sure we become a part of the design.
audaxdreik@pawb.social 3 days ago
That’s just simply not true. Tools are usually quite specific in purpose, and often times the tasks they accomplish cannot be undone by the same tool. A drill cannot undrill a hole. I’m familiar with ML (machine learning) and the many, many legitimate uses it has across a wide range of fields.
What you’re thinking of, I suspect, is a weapon. A resource that can be wielded equally by and against each side. The pains caused on the common person by the devaluation of our art and labor can’t be inflicted against the corpofascists; for them, that’s the point. They are the ones selling these tools to you and you cannot defeat them by buying in. And I do very much mean the open source models as well. Waging war on their terms, with their tools and methods (repossessed as they may be) is still a losing proposition.
Time will tell. How are your NFTs doing? (sorry, that was mean)
Guilty as charged, I’m pretty strongly anti-AI. But seriously, watch that ad and tell me that the disorienting cadence of speech and uncanny, overly detailed generated images look good? Most of us have seen what’s on offer and we’re telling you, we’re tired.
Look, I do apologize, I’m very much trying not to be overly aggro here or attack you in any way. But I think discussions about the religious overtones and belief systems of the BJ are exactly where we’re at.
How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
This is a really interesting article. Gary Marcus is a lot more positive on AI than myself I think, but that’s understandable given his background. If I do concede that some form of AGI is inevitable, I think we are within our rights to demand that it is indeed the tool we deserve, and not just snake oil.
AI art still ugly, sorry not sorry.
myfavouritename@beehaw.org 3 days ago
No note other than to say this is a fantastic rebuttal.