market analysis

Premium: The Hater's Guide To The AI Bubble 3.0

2 months ago

The author introduces a versioning system for updates on the AI bubble instead of volumes.
The AI bubble is described as a psyop, melodrama, financial crisis, and exploitation of ignorance for trillions in investment.
LLMs are real but don't fulfill promises like replacing 50% of white-collar jobs, as claimed by boosters like Dario Amodei.
AI markets are driven by fan fiction and scams, with companies like OpenAI and Anthropic discussing potential rather than actual performance.
89%+ of AI revenues and 90%+ of compute demand come from OpenAI and Anthropic, based on money-losing subscriptions and unsustainable token burn.
Enterprises are capping AI spend after blowing through annual token budgets, highlighting lack of ROI.
The entire AI industry's revenue barely reaches $100 billion by 2026, with no profitability outside data center component sellers.
AI products are inconsistent, unreliable, and improvements are measured by increasingly useless benchmarks.
Generative AI is limited by probabilistic hallucinations, making every output suspicious and decisions arbitrary.
AI cannot replace most jobs; it only generates and summarizes at varying complexity, with CEOs disconnected from real work.
Media failed to scrutinize AI hype, which is inflated by vague claims like 'AI can build software'.
Data center construction is delayed, and AI compute demand is concentrated in Anthropic and OpenAI, which rely on debt and venture capital.
NVIDIA's growth depends on a few customers, with demand for components high but AI compute demand low.
The AI bubble differs from the Dot Com Bubble and is driven by 'Business Idiots' throwing cash at a dead-end technology.

Hasty Briefsbeta

market analysis

DeepMind Disbands AlphaFold Team, Pivots to Gemini

Premium: The Hater's Guide To The AI Bubble 3.0