Dan Luu’s review finds Ed Zitron’s AI scepticism riddled with failed predictions

Dan Luu has published a comprehensive review assessing the predictive accuracy of Ed Zitron, a figure widely cited in the ongoing debate over artificial intelligence progress. The analysis concludes that Zitron is frequently incorrect in both his specific forecasts and the underlying logic used to support them. Luu, who previously reviewed the prediction records of futurists such as Ray Kurzweil, argues that Zitron’s record is particularly poor because he is wrong on both the outcomes and the reasoning, a distinction that separates him from many other commentators who may be wrong on the details but right on the broader trajectory.

The review highlights specific failed predictions regarding the financial health of major technology companies, including Meta, Google, and Microsoft. In a November 2024 talk, Zitron described these companies as "dying" and suggested they were thrashing around with AI because they no longer knew how to grow. Luu points to reported revenue and profit figures that contradict this narrative, noting that the ecosystems of these companies are not in a state of desperation. For instance, Zitron’s claim that Google’s user forecast for its Gemini model was "unrealistic" was disproven when the product exceeded that forecast by 50 per cent.

Luu argues that Zitron relies on a "gish gallop" strategy, a rhetorical tactic involving the flooding of an audience with a high volume of aggressive claims and flawed data to overwhelm refutation efforts. The analysis notes that Zitron often cites inaccurate third-party tracking numbers or misinterprets primary sources to support his anti-AI stance. For example, Luu details how Zitron’s projection of Anthropic’s revenue contained multiple counting errors, including miscounting months and including dates that do not exist, which skewed the final figure.

The review also touches on the style of Zitron’s argumentation, describing it as heavily reliant on anger and identity-driven positioning rather than coherent economic analysis. Luu compares Zitron’s approach to that of futurists who use numbers to create an aura of credibility, but notes that Zitron’s reasoning often fails to connect cited data to a logical conclusion. Former Google engineer Juho Snellman is quoted in the analysis, describing Zitron’s economic analysis as "absolute trash" and suggesting that the errors are often intentional deceptions masked by flamboyant writing.

Despite the volume of incorrect statements, Luu acknowledges that Zitron has a significant following, largely because his provocative stance resonates with those who feel AI progress is overhyped. The analysis suggests that Zitron’s appeal lies in his ability to articulate crisp, strong positions that leave no room for doubt, a strategy that drives engagement but often sacrifices accuracy. Luu notes that while some futurists have managed to salvage their reputations by claiming their predictions were simply delayed, Zitron’s pattern of repeated, high-confidence errors makes this defence more difficult to sustain.

The review serves as a cautionary tale for investors and institutions relying on popular commentators for market signals. By dissecting the specific failures in Zitron’s predictions regarding corporate growth and AI capabilities, Luu provides a data-driven counterpoint to the narrative that AI progress has stalled. The analysis remains a single-source assessment, but it offers a detailed framework for evaluating the reliability of high-profile tech commentators in a rapidly evolving market.