So look, I know it’s April Fools today – my tempestuous Dad’s birthday. Today is also posting day, and I’ve been harping about AI and certificates and courses and so on… But on this day of chaos, let’s do something different, as I have a bit of that contrary spirit, too, part nationality (all Poles get it), part inheritance.

I’ve found and summarised this fantastic, scathing and – really real – critique of generative AI by Edward Zitron. You can read the original at https://loom.ly/cIpFCt8, I cannot recommend it enough.

Core Argument: Generative AI Is Overhyped and Fundamentally Unsustainable

Important points, in order of appearance:

Exaggerated Media Reports – Zitron criticises mainstream media for taking company claims at face value, blindly reporting unchecked numbers and thus contributing to the issue.

Suspect Growth Metrics – Zitron argues with reasonable proof that OpenAI’s growth data might be inflated and the company refusal to publish monthly active user figures suggests weak conversion rate to paying customers.

Weak Adoption – Zitron presents traffic data for other generative AI platforms — Google Gemini, Claude, Perplexity, and Microsoft Copilot — showing modest numbers that don’t support claims of widespread adoption or indispensable use, especially when compared to the much higher traffic seen by traditional publishers like The New York Times or CNN.

Lack of iPhone Moment – Generative AI has yet to produce a game-changing, universally essential product. Zitron asks: If generative AI vanished tomorrow, how many people would actually notice?

Burning through Money – Generative AI companies are burning through billions. OpenAI and Anthropic both lost around $5B+ in 2024 alone. Microsoft’s reported $13B annual “AI run rate” revenue is not profit and is dwarfed by its total earnings. Zitron argues these are not signs of a thriving industry but of massive, unsustainable expenses disguised as future potential.

Anthropic’s Forecasts are Fantasy – Anthropic’s projections — expecting $34.5B in revenue by 2027 — are labeled laughable. Their actual 2024 revenue was under $1B. Zitron dismisses their claims as “vibes and hysteria,” driven by charismatic but unreliable founders who make outlandish, unproven claims about future AI capabilities.

Deep Research is Mediocre – Zitron calls it a superficially impressive but deeply flawed product, with poor citation practices, uncanny and soulless language, and heavy compute costs — all for a feature that fails at its basic goal: reliable research.

The Big Picture: It’s all Hype – Zitron believes the generative AI boom is being driven not by innovation or public need, but by elite financial interests eager to automate labor and inflate valuations.

Dangers Ahead – Zitron warns of the environmental toll of massive compute demands, the social cost of job loss driven by illusory automation promises, and a growing disillusionment with Silicon Valley’s ability to solve real problems. He fears that this tech bubble will burst in a collapse even greater than the dot-com crash.

Conclusion: Zitron argues that AI industry is driven by hype, sustained by questionable data, and producing little of lasting value. Despite the rhetoric about world-changing technology, Zitron sees an industry that burns money, pollutes, and gaslights the public into believing its mediocre outputs are revolutionary — a con that may soon collapse under its own weight.