HeadlinesBriefing favicon HeadlinesBriefing.com

AI hype or threat? The fallout of calling LLMs next‑token predictors

Hacker News •
×

Critics on Hacker News decry the label “next‑token predictor” as a slur, accusing LLM supporters of celebrating the “solved” future of animation, Hollywood, coding and academia. The rhetoric mirrors political tribalism, rapidly with enthusiasts crowing that AI renders human creativity obsolete today. This zeal fuels a culture war over whether the technology merely predicts text or reshapes entire industries globally.

Leaders such as Anthropic CEO Dario Amodei and OpenAI's Sam Altman proclaim that degrees are irrelevant and that AI will soon “solve” many jobs. Pricing models charge $200 a month for premium access, while token fees run into hundreds of dollars. Companies eye these costs as a path to replace offshore labor, tightening control over the means of production.

The training data pool consists of virtually every publicly available text, scraped without consent unless sites opt out via robots.txt. Thousands of low‑paid annotators curate these datasets, and massive GPU farms consume enormous energy. As venture capital pours into the rush, the promise of solving climate or disease recedes, leaving a profit‑driven AI boom that concentrates power in a few hands.