History doesn’t repeat, but it often rhymes. The current discourse around Artificial General Intelligence (AGI) bears unsettling similarities to another technological movement that promised to solve humanity’s problems: eugenics. Both wrapped ideological aims in scientific language, both created a false sense of competitive urgency, and both justified concentrated power through pseudo-scientific metrics.
The Scientific Veneer
In the early 20th century, eugenics took legitimate discoveries about heredity and twisted them into justification for social engineering. While Mendel’s peas and Darwin’s finches revealed genuine insights about inheritance, eugenicists made unsupported leaps to claims about human “fitness” that science couldn’t sustain.
Supreme Court Justice Oliver Wendell Holmes captured this pseudo-scientific confidence when he declared in Buck v. Bell (1927): “It is better for all the world if, instead of waiting to execute degenerate offspring for crime or to let them starve for their imbecility, society can prevent those who are manifestly unfit from continuing their kind… Three generations of imbeciles are enough.”
Today’s AGI discourse follows a similar pattern. While neural networks demonstrate impressive capabilities in narrow domains, advocates make dramatic, unsupported extrapolations about inevitable superintelligence. Scaling theories – the belief that simply making models bigger leads to qualitatively different, almost magical capabilities – mirror how eugenicists believed selective breeding would create qualitatively “superior” humans.
Measuring the Unmeasurable
Eugenicists created elaborate measurement systems – IQ tests, craniometry, family pedigree charts – that purported to quantify human worth. These metrics created the illusion of scientific objectivity while primarily measuring conformity to the researchers’ own cultural values.
AI capabilities benchmarks serve a similar function today. Tests like MMLU, TruthfulQA, and various leaderboards measure narrow capabilities disconnected from beneficial real-world applications. Like phrenology’s calipers, these benchmarks create circular systems where the measurement tool itself defines “success” rather than meaningful outcomes for humanity.
The Competitive Imperative
Perhaps most dangerous is how both movements weaponize fear and competition to accelerate development without democratic input.
Eugenicists warned of “race suicide” if nations didn’t implement breeding programs. This competitive framing justified extraordinary measures outside normal democratic processes.
Today’s AGI discourse similarly frames development as an existential race. National security arguments position AI development as imperative for survival against geopolitical competitors. Companies justify concentrating resources and power as necessary to “win” a technological race that will determine humanity’s future.
From Progressive Language to an Extremist Agenda
Both movements use forward-looking, even utopian language while producing deeply conservative outcomes. Eugenics presented itself as scientific progress while reinforcing existing social hierarchies. Its “progressive” policies in Sweden and the United States primarily targeted the poor, immigrants, and racial minorities.
Similarly, today’s “Techno-utilitarian Long-termism” in AI uses future-oriented rhetoric while concentrating present power. It justifies massive resource allocation to a few corporations, positions technical experts as uniquely qualified decision-makers, and treats democratic input as uninformed or dangerous.
Breaking the Pattern
Recognizing this pattern is the first step toward breaking it. We need not repeat the mistakes of the past. The alternative isn’t rejecting technological progress but democratizing it – ensuring that AI development serves human flourishing rather than abstract metrics or competitive imperatives.
The question isn’t whether AI will advance, but who will govern its development and to what ends. History shows us the dangers of leaving those decisions to self-appointed technical elites operating under the pressure of imagined existential races.
