What Technology Does to Who We Are
Timandra Harkness says personalising technology fills a gap in who we are. Christine Rosen says mediated life costs us real experience. Both can be right.
Key takeaways
- Timandra Harkness argues that personalising technology spread because our sense of who we are had already shrunk, not the reverse.
- Christine Rosen argues that mediated life replaces direct experience, a claim about value that no study can settle.
- A 2020 experiment by Hunt Allcott and colleagues found that deactivating Facebook for four weeks increased time with family and friends.
- Amy Orben and Andrew Przybylski found digital technology use explained at most 0.4 per cent of the variation in adolescent well-being across 355,358 young people.
The question is usually asked one way round: is technology making us smaller, lonelier, more distracted, less ourselves? Two books published in 2024 suggest the question itself may be facing the wrong way, or at least that it is two questions dressed as one. One asks what technology is doing to us. The other asks what we were already missing, that technology now happens to supply.
Both books are serious, both are careful in places, both were written by people who have spent years on the subject, and they are less opposed than their titles suggest.
Timandra Harkness and Christine Rosen
Timandra Harkness's Technology Is Not the Problem makes a causal argument, and states it in one sentence at page 16: "We don't have a shrinking sense of who we are because of personalising technology. We have personalising technology because of our shrinking sense of who we are."
The argument behind that sentence is historical. The old structures that told people who they were, such as class, church, trade and locality, have loosened. Identity has become something each person has to work out and display alone, and anything that has to be displayed also has to be recognised. Personalising technology, in her account, succeeds because it offers a cheap, constant, measurable form of that recognition. The device answers a need that was already there. She is explicit that she is not calling it harmless, and says so more than once.
Christine Rosen's The Extinction of Experience is about something else: what a mediated life contains. Her mechanism is set out at page 17, that "we relate to our world through information about it rather than direct experience with it". The title phrase is borrowed, and she says so, from the naturalist Robert Michael Pyle, who used it for the loss of wild places. Rosen extends it to a loss of human practices: waiting, handwriting, reading a face across a table, wandering a street without a destination.
Her book is cultural criticism rather than a review of evidence, and much of its argument is about value. At page 99, on waiting and idleness, she writes, "To be fallow is not the same thing as to be useless." That is not a sentence any study could confirm or refute. It is a claim about what a good life includes.
A study can tell you how often people talk to strangers. It cannot tell you how much those conversations were worth.
So the two books are answering different questions. Harkness asks where the cause lies, and locates it in a society that left people to build their identities alone. Rosen asks what is being lost, and locates the loss in the body and in shared physical presence. A reader can accept both answers at once, and most of the argument between the two books turns out to be about emphasis.
That matters for anyone weighing their own habits, because the two accounts suggest different remedies. If Harkness is right, a phone is a symptom, and putting it down leaves the underlying need unmet. If Rosen is right, putting it down gives back something specific, a set of experiences the phone had quietly replaced.
Where the two books agree more than they seem
One telling detail sits on Rosen's own praise pages. Harkness endorses Rosen's book there. Two authors who are supposed to disagree, one recommending the other, suggests that their difference lies in interpretation and values rather than in facts.
Harkness also writes passages Rosen could have written. Near the end of her book, at pages 296 to 297, discussing concerts played to empty halls, she observes that "isolated individuals, even connected via the internet, remain individuals, with all the vulnerability that brings." That is a sentence about the loss of shared physical experience, written by the author whose title says technology is not the problem.
And Harkness names a worry that runs in Rosen's direction. At page 243 she writes that systems which sort people into profiles risk "nudging us towards becoming more like the cluster to which we were allocated". That is the machine shaping the person, the direction her main thesis treats as secondary. She records it and leaves it there.
Rosen, for her part, concedes the limit of her evidence at the point where it matters most. At page 45, about to present the studies that carry much of her chapter on face-to-face contact, she writes that "large-scale studies of the long-term effects of mediated relationships on children are not yet available", and calls the studies she does have "a few intriguing new studies". That is the right weight for what they are.
So both books hold both directions somewhere in their pages. Harkness accepts that the systems can shape who we become. Rosen accepts that she cannot yet measure how much. Where they really differ is on which direction deserves the headline, and on whether losses that cannot be counted, such as boredom, friction and chance encounters, should count.
Technology mostly arrives to fill a gap that was already open. Once there, it can widen the gap.
That second difference cannot be settled by research. A study can tell you how often people talk to strangers. It cannot tell you how much those conversations were worth.
Hunt Allcott and colleagues
There is one place where the two accounts can be set against the same evidence, and it is the question of displacement: does time online replace time with people, or add to it?
The strongest experiment on the question was run by the economist Hunt Allcott and three colleagues, and published in the American Economic Review in 2020. They paid people to deactivate Facebook for four weeks before the 2018 American midterm elections and compared them with people who kept using it. Deactivation reduced online activity and increased offline activities, including time spent with family and friends. It also reduced factual news knowledge and political polarisation, raised subjective well-being by a small amount, and left people using Facebook less even after the experiment ended.
Harkness argues, at page 250 of her book, that "there is scant evidence that more use of social media is the cause of less in-person social time." The Allcott experiment is direct evidence that, for one platform and one month, removing it increased in-person time. That is a point for Rosen on this question.
It is not the whole picture. It is one platform, one month and one country, and it measured what happens when a platform is taken away, not what happens as people take one up. Messaging used mainly with existing friends may well work differently. The fair summary is that online life can do both, and which one depends on how it is used, so a general claim in either direction goes further than the evidence.
The size of any harm is a separate question. Amy Orben and Andrew Przybylski analysed three large datasets covering 355,358 adolescents in 2019 and found that digital technology use accounted for at most 0.4 per cent of the variation in their well-being. That is a small association, and it fits the argument that harms are overstated better than it fits a sense of large loss.
Neither number settles the other question either. A small average can hide a larger effect in a minority of heavy or vulnerable users, and nobody has resolved whether it does.
Which account we find more convincing
On the question of cause, we find Harkness more persuasive. Her argument that the hunger for recognition came first, and that technology profits from it, fits the history she sets out and explains why so much personalising technology is built around being seen, counted and confirmed. It would also predict that taking the device away leaves the need unmet, though the experiment above found well-being rose slightly when one platform was removed.
On the question of what is lost, we think Rosen is describing something real that research is poorly placed to measure. The Allcott experiment shows that one kind of loss, time with other people, can be measured and does move. The rest of what she describes may never be measured, and that does not make it imaginary.
So our position, held with moderate confidence, is that both directions operate, and that the stronger one is the one Harkness identifies. Technology mostly arrives to fill a gap that was already open. Once there, it can widen the gap, by replacing some of the experiences that might have closed it. Neither book claims the opposite of that, which is why neither should be read as the answer to the other.
For a reader, the practical question is not whether screens are good or bad. It is which gap a particular habit is filling, and whether something better could fill it.
Common questions
Does technology make us lonely?
The evidence is mixed. A 2020 experiment by Hunt Allcott and colleagues found that deactivating Facebook for four weeks increased time with family and friends, while a 2019 analysis of 355,358 adolescents by Amy Orben and Andrew Przybylski found digital technology use explained at most 0.4 per cent of the variation in their well-being.
What is Timandra Harkness's argument?
That personalising technology spread because people's sense of who they are had already been weakened by the loss of older social structures, not the reverse. She states it at page 16 of Technology Is Not the Problem, and does not claim technology is harmless.
Sources
Harkness, T. (2024). Technology Is Not the Problem. HQ, HarperCollins. Edition of 2025. Read in full by this publication, 2026-10-06, with a reliability audit. p. 16, the causal reversal; p. 243, the cluster quotation; p. 250, the displacement quotation; pp. 296-297, the isolated-individuals quotation.
Rosen, C. (2024). The Extinction of Experience: Being Human in a Disembodied World. Read in full by this publication, 2026-10-06, with a reliability audit. p. 17, the mechanism, crediting Robert Michael Pyle for the phrase; p. 45, the concession; p. 99, the fallow quotation. Harkness endorses the book on its praise pages.
Allcott, H., Braghieri, L., Eichmeyer, S., & Gentzkow, M. (2020). 'The welfare effects of social media.' American Economic Review, 110(3), 629-676. doi:10.1257/aer.20190658. Four-week Facebook deactivation before the 2018 United States midterms; reduced online activity, increased offline activities including time with family and friends, reduced news knowledge and polarisation, raised subjective well-being by about 0.09 standard deviations, and reduced later use.
Orben, A. & Przybylski, A. K. (2019). “The association between adolescent well-being and digital technology use.” Nature Human Behaviour, 3, 173–182. Specification curve analysis, n = 355,358.
A note on scope: our position on cause is a judgement, held with moderate confidence. Whether a small average hides larger effects in a vulnerable minority is unresolved.