A profile of Joseph Weizenbaum, who created the first chatbot in 1966 but turned against AI, believing the computer revolution constricted our humanity
Computer scientist Joseph Weizenbaum was there at the dawn of artificial intelligence - but he was also adamant that we must never confuse computers with humans Mastodon: @Weizenbaum_Institut … , @skarthik@neuromatch.social , @joningold@mastodon … , and @bentarnoff@mastodon.social Twitter: @johnpmerrick , @yohannk , @erocdrahs , and @cabitzaf Mastodon: @Weizenbaum_Institut@social.bund.de : “Weizenbaum's pessimism made him a lonely figure among computer scientists during the last three decades of his life; he would be less lonely in 2023.” — Compelling long read about Joseph Weizenbaum, the Eliza Effect and the importance of his critique in current debates on technology. … Karthik Srinivasan / @skarthik@neuromatch.social : A must read longform on the pioneer of AI chatbot who became AI's main and earliest detractors, Joseph Weizenbaum (of ELIZA fame here at MIT) by Ben Tarnoff! — Weizenbaum's was a lone voice in the 1970s, and many of his views are as apt now as they were then. … Jon Ingold / @joningold@mastodon.gamedev.place : Fascinating article on the creator of ELIZA (which it turns out is pretty much the least interesting thing he did) and the limitations, and the cult, of AI. — https://www.theguardian.com/ ... “Living well with computers would mean putting them in their proper place: as aides to calculation, never judgment.” Ben Tarnoff / @bentarnoff@mastodon.social : I wrote an essay for the Guardian Long Read about Joseph Weizenbaum, whose work I find boundlessly useful for thinking about the pathologies of computerization and AI https://www.theguardian.com/ ... Twitter: John Merrick / @johnpmerrick : This is a brilliant and utterly compelling piece of writing by @bentarnoff on Joseph Weizenbaum, AI pioneer turned heretic, that takes in psychoanalysis, the birth of modern computing, and more. https://www.theguardian.com/ ... Yohann Koshy / @yohannk : This is such a rich and compelling piece of writing — @bentarnoff on the wild life and times of Joseph Weizenbaum, tech high priest turned AI heretic: https://www.theguardian.com/ ... @erocdrahs : “ We should never “substitute a computer system for a human function that involves interpersonal respect, understanding and love” ” this is a very good (long) read: https://www.theguardian.com/ ... Federico Cabitza / @cabitzaf : “The book [Computer Power and Human Reason] has two major arguments. First: there is a difference between man and machine. Second: there are certain tasks which computers ought not be made to do, independent of whether computers can be made to do them. ” https://www.theguardian.com/ ...
Context & Ripple Effects
The Guardian's long read arrives at the end of a year when Weizenbaum's founding paradox stopped being historical trivia: the man who built ELIZA in 1966 spent his last three decades arguing that computers must never be confused with humans, and the coverage notes he 'would be less lonely in 2023'. His loneliness had a specific cause — as the related profiles show, the field's other founders went the other way, from the utopian personal-computing vision of Avery Johnson and Warren Brodey's 1967 cybernetics lab to Jürgen Schmidhuber's pursuit of artificial general intelligence.
What makes the revival more than nostalgia is that today's products are running the exact experiment he warned against: an artist who gave Replika's bot Lila an AI friend modeled on himself treats a chatbot as a social being, while Emily Bender's work mapping what LLMs can and cannot do carries his distinction into the technical literature.
First-order effects
- Weizenbaum's core argument — codified in 'Computer Power and Human Reason' as a moral line between humans and machines — gets a fresh audience precisely because consumer chatbots now occupy the companionship role he insisted machines should not hold.
- His critique gives contemporary skeptics a founder's pedigree: Bender's limits-of-LLMs analysis and the writer's case that chatbots repackage ideas without lateral thinking both inherit the position that some tasks are unfit for machine delegation.
Second-order effects
- Companion-AI vendors like Replika face a legitimacy question their marketing doesn't answer: if the field's own creator drew a hard line against confusing computers with people, every attachment users form becomes evidence for the critics' side.
- AI pioneers on the AGI path, such as Schmidhuber, now share a canon with their most famous dissenter — press profiles of founders increasingly have to reckon with the anti-delegation tradition rather than presenting the field as unanimous.
Third-order effects
- If the pattern holds, the durable legacy is institutional rather than rhetorical: Weizenbaum's name already anchors a research institute, and the boundary he drew — which tasks society refuses to delegate to machines — is becoming the frame for governing companion AI rather than a footnote to chatbot history.
- The split his career embodies, between builders who personalize computing and critics who police its reach, looks structural: each generation of the field reproduces both camps, meaning AI's governance debates will keep citing a 1960s origin point.
The trend: As chatbots shift from novelty to intimacy, the field is rediscovering its founding critic's question — which tasks should never be delegated to machines — as the central line in AI governance.