Once you dig into the “AI industry” it becomes clearer & clearer what it really is. It is an extraction project. It has already extracted the ENTIRE DIGITIZABLE OUTPUT OF HUMANITY. Our ideas. Our art. Our science. Our math. Through a clever algorithm discovered in 2017, Jensen Huang and Peter Thiel’s protege Sam Altman figured out how to accelerate existing “generative AI” technology with GPUs. This was NOT an innovation in accuracy or “intelligence,” it was a way to process an old algorithm in parallel. OpenAI was staffed in large part by “Effective Altruists”—a cult started based on the doomer science fiction of Eliezer Yudkowsky on the website LessWrong. Peter Thiel was the initial funding for OpenAI, EA, and Yudkowsky. OpenAI successfully convinced the market, and themselves in many cases, that they had birthed some new form of life. They just needed to make it BIGGER and it was destined to become “superintelligent.” Based on Yudkowsky’s goofy ideas, AI was gradually anthropomorphized and deified—despite the fact that it was just software regurgitating the exhaust of human beings. The entire “alignment” movement is based on the presumption that chatbots deserve “moral standing.” It’s exhaustingly ignorant. Eventually OpenAI employee Dario Amodei, a dedicated EA adherent, became concerned that OpenAI was not RELIGIOUS ENOUGH about AI being sentient, so he broke off to form Anthropic and named his chatbot after a person. This false premise, that an algorithm was somehow producing “intelligence” was further amplified by training the output to sound as human, and addictive, as possible. The combination of a kind of religious fervor inside, a deceptive product, and a flood of influence operations on the outside fueled a wild flow of speculative investment in both OpenAI and Anthropic—and especially into NVIDIA. This eventually poisoned the entire technology sector which came to believe that the “race for AGI” was existential, so they committed TRILLIONS to achieving it. The data centers are all a function of a misunderstanding of what they are building. Generative pre-trained transformers (GPT) are NEVER going to be intelligent, much less conscious. All they do is take our stolen stuff and sell it back to us. They do not experience. They do not do anything—unless you send a string of text for them to process. Nevertheless, the entire financial industry, and now the federal government is pushing this dead-end project to engulf the nation in data centers and swallow the economy with “AI factories.” It’s downright insanity. All of that capital expenditure is already junk. It can physically never make money. It’s built for GPUs which are the wrong substrate for AI. Large language models are useful search engines for stuff we’ve already done. They are the compressed, digitized exhaust of humans with a clever language interface. That’s it. There is no moat. There is no real innovation. There is just an accelerationist delusion and no adults in sight. Chatbots are a software feature, a commodity, that has metastasized into a cult digging a $10 TRILLION DITCH. The reason this has gone so far is that rather than compete in a democratic capitalist system for the best technology product, this cabal of lunatics created an “internet alternate reality” that could “unilaterally change the world.” The current AI industry is the largest extraction project in history. It is strip mining humanity and selling it back to us as slop. “If you’re evil, you’re not bad, and maybe that means you’re kind of good.” —Peter Thiel