AI is a cruel technology. It replaces workers, devours millions of gallons of water, vomits CO2 into the atmosphere, propagandises exclusively for the worst ideologies

Moreso even than cryptocurrency, AI is entirely nihilistic, with zero redeeming qualities. It is a blight upon the world, and it will take decades to clear up the mountains of slop it has generated in the past two or three years.

Bruh, what is the author talking about? How do they claim to take a material view of the world when they make completely unsubstantiated statements like these? First of, AI is a much broader category than LLMs or Diffusion models, and the fact that the author doesn’t seem to know that is genuinely concerning. Secondly, even for LLMs, I’ve been using it to summarise lectures and generate quizzes for me to study with. It is a genuinely valuable technology that is helping me and many other students to learn better.

Sweeping, unfounded statements like this only damages the credibility of socialists, especially when we pride ourselves on having a grounded, material worldview. I do hope this is an outlier among the community.

Almost certainly intentional, given that none of their Dall-e models nor their Sora models have a yellow tint like this. My guess is that it functions as a watermark equivalent or something.

This encapsulates my views pretty perfectly. Consumer AI (whether it’s LLMs, diffusion models, audio models etc) has been here for a while now, and is used by tens, if not hundreds of millions of people daily. Certainly, you could argue it’s overvalued or overhyped, but just as certainly it’s not valueless. And the open models coming out from China show there is a way to put control of AI in the people’s hands.

To just cede the entire field of consumer AI to liberal or right-wing groups would be a terrible strategic error, especially as Musk and co have gleefully weaponised AI to for their political purposes. At the very least, we must figure out how to counter this, and we cannot do so by shunning the technology entirely.

LLMs are pretty good at translation in my experience, often better than traditional translation services. But quality can vary highly between languages. English to French and vice versa is probably the best case scenario for Mistral models, so I wouldn’t expect the same level of quality for other languages, especially non-European and/or obscure ones.

Also, I do agree that getting more pro-communist text on the internet for LLMs to train on is something we could try and push for. Would certainly make for better data than all the liberal and right-wing stuff on the internet right now.