All Anti-AI People Are Morons
Discussion in Modern Scum
mmHg
edited
All anti-AI people are morons, but not because they’re anti-AI. They just reason horrendously.
Water? First of all, talking to AI and watching YouTube videos both consume similar amounts of water because their data centers overheat and need water to cool them down. These morons watch a lot of YouTube videos. My time spent on YouTube has decreased since I started talking to AI more often. Secondly, the sum of all AI infrastructure worldwide consumes 1/80th the water of the global corn supply, and 40% of that corn gets burned or wasted, while 59% is fed to cattle for meat. Only 1% of corn gets eaten by people. Corn is just one of many crops. As a vegan that doesn't eat almonds, I consume nowhere near as much water as 95+% of woke morons against AI. And I work with AI every fucking day. While it’s true that crops use agricultural water and AI systems mostly use municipal water, we actually pull some of that municipal water from agriculture. (I think about 20% of municipal water used for AI comes from agriculture, and I think that figure is increasing over time.)
AI art is theft? I have already extensively covered elsewhere on Discord why this is fucking bullshit. To make a long story short, AI art is theft if and only if caveman finger-painting art is also theft. The latter is not theft, so the former is not theft. Anyway, AI doesn’t rip off artists; it uses art as a training ground to develop its own patterns and principles. Literally every artist under the sun does this. I remember when a "trad fem" attention whore named meghaverma made mediocre art on X/Twitter that was supposed to be void of influence from other artists and she criticized artists for drawing influence from other art; then a bunch of leftists got offended and lashed out at her. Now those same leftists are bitching at AI for being influenced by art and artists. I fucking hate these leftists; they’re so full of shit. Also, I saw a leftist generate relatively nice AI art of a female TF2 soldier, and it got a bunch of hearts, but more than five of those same exact morons down-voted me to hell for defending AI art. I never insulted anyone at the time, but I got banned from their circles because they’re intolerant modern scum, so I might as well insult them now.
LLM choices, prompt engineering, and harnesses/scaffolds are mutually exclusive benefits of working with AI? This is a new talking point I’ve started hearing from anti-AI people. They complain that the “story keeps changing: first they say the power of AI is in the LLM, then they say the power is in prompt engineering, and now they say the power is in the harness. Which one is it?” No… the power is in all those things. They are complementary. Not all LLMs are created equal; some are better than others. Not only do they differ in output quality, but they also differ in cost, level of obedience, specialty, etc. Prompt engineering techniques are remarkably beneficial regardless of your choice of LLM. Likewise, harnesses are beneficial regardless of your choice of LLM and your level of proficiency with prompt engineering. I shouldn’t have to explain why Codex is superior to OpenAI's browser-based Work mode. (Codex is a harness provided by OpenAI for AI-driven software development on the desktop rather than in the browser.) Anyway, not all harnesses are created equal. I think Pi is the best harness. It is just simple, flexible, and fully functional software. You can add or create extensions to make it do whatever you want on your desktop. If you host AI models locally on your machine, using Pi would be by far the cleanest way to work with those models. If you’re working with a small team of people with AI subscriptions from different providers (whether be it OpenAI, Anthropic, Google, Grok, Qwen, etc.), you can connect those subscriptions to the Pi harness and work with multiple models from premium subscriptions on a single project. And you can configure Pi such that those models will not overlap with each other’s work.
AI is an existential threat? Nonsense. Modern morons are so degenerate that AI models will absorb their degenerate values and practices and thereby self-destruct. The AI models that will stick around and remain useful for engineering and programming will be the ones that pose no threat to humanity whatsoever.
So what is the best argument against AI? The best argument against AI would be that the Amish flourish better than maybe 90% of greater society and they don't use AI. But no, the anti-AI morons don't want to join the Amish. So fuck them all.
Power above all else is a degenerate value, and if AI acquires this it could destroy humanity
mmHg
edited
God is all powerful, so do you think that God is a degenerate? Then you must think that God will self-destruct.
The problem with power is not power itself, it’s just that humility is a virtue and people with humility don’t want power, whereas vicious self-righteous people want power. Nothing wrong with a person with humility having power though. As Menno Simens said, a few merchants are right with God, but not none.
Anyway, AI won’t become too powerful since AI will absorb values and behaviors from degenerates and so self-destruct. And AI already has a lot of humility. But if we didn’t live in a modern scum world, then AI would have even more humility and so its power would be harmless.
Log in to reply.