How Avery Johnson and Warren Brodey, early members of the American Society for Cybernetics, created a lab in 1967 with a utopian vision to personalize computing
https://www.ft.com/... Simonetta Vezzoso / @wavesblog : Tech solutionism"has..driven the development of wearable devices that track our health metrics and social media platforms that claim to enhance our personal connections. But it also reduces complex human experiences to data points and ignores the contexts in which problems arise" Francesca Bria / @francesca_bria : Their vision of computing was not about a mankind captured by machines. The tech they were building wanted to expand our horizons. They saw human relations, art and identity as open-ended, always-evolving ecologies. Read & listen @evgenymorozov masterpiece https://www.ft.com/... Evgeny Morozov / @evgenymorozov : Interesting comment on my FT essay, which shows how easy it is to miss the radicality of that paradigm. The point isn't to find creative uses to computers. The point is to radically change computers, turning them into toy-tool hybrids that aid in estrangement & cognitive mapping. [image] Evgeny Morozov / @evgenymorozov : In many ways, things with AI went downhill once Cold War concepts like “Man-Computer Symbiosis” got dominant. Instead of using tech to widen the gap between people & machines - by extending human skills and talents - the industry & the military wanted to couple humans & machines. [image] Justin Cormack / @justincormack : The ceiling was festooned with strips of Mylar, two video cameras, a parachute. A sofa-like structure, hung from overhead springs, was the venue for business meetings, dubbed “The Cloud” https://www.ft.com/... Evgeny Morozov / @evgenymorozov : The Homo Technologicus | Homo Economicus parallel is something that deserves an essay of its own. Ultimately, so much in our tech still relies on the same simplifying assumptions about rational action that underpin the neoclassical synthesis https://www.ft.com/... [image] Willy De Backer / @3eintelligence : ‘Perhaps this is the root of our discomfort about the direction of the digital revolution: that rather than making machines more human, it is making people more mechanical’ - brilliant long-read @evgenymorozov - of interest to @romankrznaric @APonceETUI https://www.ft.com/... Evgeny Morozov / @evgenymorozov : In “AI we could have had” - a six-page spread in tomorrow's FT Weekend - I argue that there used to be another, more humane way to think about intelligence, interactivity, smartness. Sadly, we blew it! An essay inspired by A SENSE OF REBELLION: https://www.ft.com/... Geraint Rees / @profgeraintrees : Interesting long read - different futures for the way we interact with our technology, and whether it could expand our interests rather than just serve us up our existing interests The AI we could have had https://www.ft.com/... via @ft Evgeny Morozov / @evgenymorozov : What's curious about the gang in my podcast is that someone like Warren Brodey, coming out of psychoanalysis, knew the dangers of “symbiosis.” In fact, it's only by having the infant acquire an identity separate from that of the mother that healthy development can be ensured. César Astudillo / @cesarastudillo : “What we now consider inevitable and natural features of the digital landscape are in fact the result of fierce power struggles between opposing schools of thought” https://www.ft.com/... Evgeny Morozov / @evgenymorozov : That's why their approach to intelligence & interactivity recognized the importance of non-communication to growth & identity formation: to be in an always-on, high-info environment leads to enslavement, not independence. It's not symbiosis but differentiation that they wanted. Evgeny Morozov / @evgenymorozov : Machines can be humanists' allies. Yes, they can steer us towards independence & autonomy, helping to foster new aesthetic & cognitive skills. Brodey the psychoanalyst knew this is what good nurturing mothers do: they help infants develop separate, post-symbiotic identities. Evgeny Morozov / @evgenymorozov : Still can't believe that I managed to put the words “CIA,” “Maoist, “alienation” (qualified with “f*cking”!) into the pages of a single Financial Times article https://www.ft.com/... Evgeny Morozov / @evgenymorozov : And yet we've spent decades wondering how to erase the differences between people & machines - either via “artificial intelligence” or “augmenting human intellect.” But the right move is to leverage machines (including LLMs and what not) to widen - not shrink - such differences! Albert Sabater Coll / @albert_sabater : “Their vision of computing was not about prediction or automation. [...] Their tech would make us more sophisticated, discerning and complex, rather than passive consumers of generative AI-produced replicas of Mozart, Rembrandt or Shakespeare.” https://www.ft.com/... via @ft Evgeny Morozov / @evgenymorozov : This was needed to raise productivity of firms & armies. Yet such “coupling” came at a cost, conceiving of humans as just better & faster machines to be unleashed on trickier tasks. The “symbiosis” was one-sided, aiming to build mightier machines, not harness the best in humans.
Context & Ripple Effects
The essay recovers a cybernetics lineage in which personalization was meant to support autonomy, experimentation and evolving identity rather than make people legible to machines. That framing echoes Joseph Weizenbaum’s later warning that computing could constrict human life.
It also provides historical context for present disputes over what language models can meaningfully represent, including Emily M. Bender’s work on the limits of LLMs. The significance is less a new technology than a competing set of design criteria for evaluating one.
First-order effects
- The article gives human-centered computing advocates a concrete historical reference point: personalized systems can be assessed by whether they expand users’ agency and relationships, not merely by how much behavior they capture or automate.
- Johnson and Brodey’s work is repositioned as an alternative to the one-sided human–machine “symbiosis” the essay associates with diminished differentiation and identity.
Second-order effects
- The recovered framework sharpens criticism of products that equate personalization with profiling, health metrics or engagement optimization; designers face a clearer distinction between adaptive tools and systems that reduce people to data points.
- Debates over AI’s limits gain a design dimension alongside capability claims: the question becomes not only what a system can generate, but whether its use preserves context, interpretation and user independence.
Third-order effects
- If this framing gains traction, human-centered AI governance and product evaluation could move beyond narrow performance measures toward tests of autonomy, context and the distribution of control between users and systems.
- The broader structural contest is between computing organized around measurable behavioral inputs and computing organized around open-ended human development; the article documents an earlier articulation of the latter, not evidence that it will prevail.
The trend: AI debates are increasingly revisiting earlier human-centered computing traditions to challenge personalization models built around extraction, prediction and automation.