Anything which disempowers or erases dignity should be handled with extreme caution, particularly when it foreseeably entails the loss of good work for large numbers of people. Such prudence is attained by developing our faculties by our own efforts, even when a vast and easy intelligence is within reach.

Zeynep Tufekci believes that artificial intelligence will soon efface several forms of evidence we depend on to trust one another, including the proof of effort1. She gives the example of a cover letter. Until recently, when she received a carefully written note she would take this as communication of respect for her time and would give the person a chance. Now these are all she receives. Unable to tell which are written by AI, she ignores them, drawing instead on her personal network – a sure way to shut out those not already well-connected.

AI can supercharge laziness (“Just use ChatGPT”) but at a cost to the corner-cutter. External analysis is useful but habitually offshoring the slow work which develops our capacity for thought will leave us childish and powerless. Jonathan Haidt describes the effect of being raised by screens in bleak terms. People, he says, are: ‘Coming of age in a confusing, placeless, ahistorical maelstrom of 30-second stories curated by algorithms’ and that ‘without solid knowledge of the past…young people will be more prone to believe whatever terrible ideas become popular around them’2.

Insight comes from working over difficult material ourselves. A psychiatrist I know once drew three concentric circles on an adhesive note. The first represented a ‘safe’ psychiatrist, who applies diagnoses and treatment strategies. One in the second circle, who understands and employs psychotherapeutic ideas and works well with the team, is good. The outer circle described excellence: a doctor who combines the aforementioned with a comprehensive knowledge of music and history, art and literature, anthropology and politics and film, the better to understand life and each patient. The front is too broad to command, but the effort to patrol it repels hubris and helps us grow.

In “Ship in the Sky”, Woody Guthrie sings of three children playing together. One’s father is a pilot, whose plane passes overhead. The next child’s father is a steel worker, who built the plane. The third, whose father works on the airport landing strip, says to the daughter of the pilot:
‘Don't be afraid when it gets dark and rains / My Dad'll bring your daddy back home again’3.

Good work is a vehicle for dignity and contribution which passes through communities and generations. 

Although artificial intelligence exceeds human capability in many domains, distributing tasks between people creates social ties and often a sense of dignity and purpose4. We should not optimise technology designed to replace this without good reason. Reducing friction in the process of consumption to the point where much of our population is unemployable and isolated is not a good reason.

Artificial intelligence and the companies that own and use it cannot be relied upon to protect the common good. The models that have been trained on vast quantities of written culture are increasingly controlled by shareholder-owned companies. Milton Friedman’s view that the function of corporations is to make as much money as possible for their owners5 is simple, self-serving and leads some executives to believe they are acting rightly in getting rid of as many employees as they find expedient. While Friedman continued that corporations should ‘conform to the basic rules of the society, both those embodied in law and those embodied in ethical custom’, this caveat is susceptible to the selective blindness of businesses which choose to ignore the negative externalities they create. These firms must be constrained by laws and forms of contribution enforced by states. These decisions are not straightforward. They should be made by citizens who are grounded in their communities and endowed with what Aristotle called phronēsis6, a habit of practical intelligence acquired through sustained effort and experience.

Technologies which stunt our minds, blunt our faculties, alienate us from useful work and make it more difficult to trust one another should not be considered ends in themselves. Charting our way requires us to draw on our whole selves: our orientation to life and history, our concern for accuracy, our common sense, our ability to see things in human terms, and our preparedness to admit error and change. In all this, one must know a great deal to avoid being made a fool of. Practical intelligence is personal, acquired by time and pain and paying attention, and is never merely artificial.

1. Tufekci, Z. “Are we having the wrong nightmares about AI?” June 4, 2026. Fulbright Distinguished Lecture. Kings College London. Lecture.

2. Haidt, J. End the Phone-Based Childhood Now. The Atlantic. March 13, 2024.

3. Guthrie, W. My Daddy (Flies a Ship in the Sky) [‘Ship in the sky’]. Bound for Glory: The Songs and Story of Woody Guthrie, Folkways Records, 1958 (recorded 1944). 

4. I owe the development of this point to a conversation with Sharif Kazemi, Seasonal Fellow, Centre for the Governance of AI (GovAI); Policy and Strategy Lead World Bank Group Development Impact AI Lab.

5. Friedman, M. The Social Responsibility of Business is to Increase Its Profits. The New York Times Magazine, September 13, 1970.

6. Aristotle. Nicomachean Ethics, Book VI.