Hey comrades I wanted to ask a question about generative ai. There was an interview recently on breadtube and I have linked the video if you want to watch the interview. Essentially from my understanding what happened is that a prominent youtuber was being evicted and evaded that because of a notice generated by chatgpt and like it is good she is housed but she went to like advocate for its use on the left while still talking about its damages which feels very wrong because it seems she knows the harms but she still wants to use it. My question is, aside running independent models and making our own models, what is the function a tech built upon stolen labour, polluting the earth, and destroying livelihoods. LLMs themselves yea can be good, ML can be good, AI can be good, but chatgpt? I don’t think chatgpt being the product the world it is can ever be good. It is built upon stolen labour and runs on scarce water of poor communities.
You must log in or # to comment.
LLMs are a dead-end technology that have become the Topic of every conversation because they’re being forcefully crammed into every corner of our lives by corporate interests desperately fighting against the tendency of the rate of profit to decline by opening new product horizons, as well as by Tech Morons who think they’re literally producing God.
The ultimate goal of the technology, as steered by the investors pouring a trillion dollars into it, is to automate information-capital laborers. Accountants, software developers, customer service, lawyers, doctors, translators, managers, and so on. All desk jobs. Is this bad? Not necessarily, but we live in a world where you will suffer and die without a job.
Further, the dystopic dream of automating all jobs isn’t some big win for the capitalists in terms of cost. LLMs use more resources and more money at a worse environmental impact to produce worse results than human employees, and it’s seemingly intractable from their design. The only victory capitalists take from their mass deployment is in some hypothetical great victory over labor. It doesn’t matter if it costs more and performs worse because it can’t strike or demand better wages or unionize or drag them out of their homes in the night or blow up their factories.
So in gestalt, LLMs are a tool for capitalists to crush labor, and the enormous expense inherent to their use means they have highly limited utility in any socialist system where they’d be responsibly wielded.
On top of that it doesn’t even automate jobs well, it only increases work of correcting it’s errors but it gives capitalists more control over lives of people they employee so it is a good deal for them. There’s a reason coal was adopted so fast when water wheel powered textiles were far more efficient. Coal gave capitalists more power.
this is more like atomization story tbh it basically worked as a better search engine +ctrl-c ctrl-v, but reduced the friction of finding local orgs/communities.
Things aren’t ontological.
aren’t “things” the most ontological, uhhh, things?
No
I found a YouTube link in your post. Here are links to the same video on alternative frontends that protect your privacy:
