The Agency Crisis

In conversation with Gurwinder Bhogal

Gurwinder is the author of one of my favourite Substacks, The Prism. Every essay he writes is exceptional, but I especially recommend The Perils of Audience Capture, The Intellectual Obesity Crisis, and The Pathologization Pandemic.

I’ve followed Gurwinder for years now and have always admired not only his writing, but how he conducts himself online. He is very measured, thoughtful, and selective about what he does. Which is why I was honoured when he agreed to have a conversation with me (and slightly intimidated about trying to keep up with his insights…)

We begin by discussing the loss of agency in the age of AI—people using AI to flirt for them, solve arguments, even write birthday cards, and slowly losing trust in their own judgement. Gurwinder describes AI as a “personality amplifier”, giving more agency to those who already have it, but taking more from those who already lack it.

Paid subscribers can access the full conversation, where we talk about how to resist audience capture, why long-term relationships can be a “life hack”, the pathologisation of not just ourselves but other people, and narcissism as a collective defence mechanism.

We conclude with practical ways to reconnect people. I give some small suggestions, but Gurwinder, as always, thinks bigger. It seems two things need to happen: we need intentionally designed spaces and incentives for social interaction, but also the bravery to just begin. The agency to go against all odds, against every disincentive, and make the first move. Maybe even by doing something really radical, like knocking on our neighbours’ doors, or inviting friends over for dinner.

We find ourselves returning to agency again and again, and how we can hold onto what makes us human. This has always been at the heart of Gurwinder’s writing—agency against ideology, against impulse and emotion, and, increasingly, agency against the machine. If you doubt yourself these days, or feel like you are losing trust in your judgement, I really can’t think of a better guide than reading Gurwinder’s work.

Here’s our conversation, I hope you enjoy. And please subscribe to The Prism here:

---

Freya India: Gurwinder, you wrote an incredible essay for UnHerd about what you call the “pathologisation pandemic” — people confusing sadness for sickness, particularly young women. Some of my other favourite pieces of yours include TikTok is Time Bomb, and Why Everything is Becoming a Game.

I think a theme here is loss of agency. We rely on the medical industry to tell us what is healthy or unhealthy, depending on psychiatric labels and diagnoses to explain ourselves. We allow algorithms to deliver us our personalities and opinions. We depend on dating apps to find us the right matches, trusting online metrics and scores rather than our own judgement. As we have both written about, this seems to come from—or cause—an external locus of control, this feeling that outside forces dictate our lives. Even if people recognise their lack of agency now, their first instinct is probably to search for influencers or podcasts to tell them how to fix it. Whereas I tend to think most of the answers we need, the wisdom we are looking for, is inside us already. We know what we need to do, what right and wrong is, but we have silenced our instinct and intuition, muffled it with all this noise.

I think this is a defining paradox of my generation: we have this lack of agency, this feeling of powerlessness, but also an outward obsession with agency. Young women often emphasise how independent and empowered they are, how they don’t need anyone.…but the evidence suggests the opposite, almost as if it’s a front for how helpless and out of control we really feel.

I worry about this with AI. My main concern isn’t so much losing human creativity or everyone having AI girlfriends, but that someday we won’t trust ourselves at all. I see a future where young people won’t trust any instinct they have without checking with ChatGPT first. Where they will ask AI to solve relationship problems, to calculate who is right in an argument, to make decisions for them instead of going with their gut. I genuinely believe people are already doing this, outsourcing not only their ability to write or work, but to decide, to act. I’ve seen a few examples of this lately, like people using AI to flirt for them, message their dating app matches, or write birthday cards. We already had a dependency problem before AI—we couldn’t think for ourselves without searching online, without asking forums for advice, without sources and studies…but now it’s getting serious.

The reason I think this is because every time you ask an AI to double-check your text messages, to make sure your email sounds right, to help you flirt with someone, you lose a little more trust in yourself, in your own judgement. I mean, imagine a young woman asking her boyfriend every time she does something, before every email she sends, “Is this okay? Does this make sense?” Eventually someone would say, uh, you’re giving him too much power over you. Eventually she wouldn’t be able to do anything without him checking first, and granting permission. That would be a recipe for a very toxic relationship. And yet I worry that is exactly what we are doing with AI, training young people to never trust their own judgement, until they can’t act, can’t think, without its approval.

As you put it in a recent tweet: “Ironically, the more the world becomes automated, the more important self-reliance becomes. When every task can be mechanized except personal agency, success hinges on taking charge of your life and making good choices.”

So, somewhat ironically, I’m going to ask for your advice…or at least your take on this. If low agency is our central problem, how do we take charge of our life, as you say? How do we trust ourselves to make the right choices, and build self-reliance in an age of algorithms and AI?

Gurwinder: Freya, it’s an honour to have this conversation, as I’m a big fan of the order in which you hit keys on a keyboard. Like you, I’m concerned about the fate of human agency in the age of automation. To me, the greater danger of AI is not that machines will think for themselves, but that humans will cease to.

The age of automation doesn’t just endanger agency, it also makes it more important than ever. For most of human history, the limiting factor in what a person could accomplish was often their intelligence. But now that we can outsource intelligence to machines, the new bottleneck for most people will be how proactively they make use of all that external intelligence.

So our degree of agency will determine our “success” in the AI age. However, the success of the AI age could also determine our degree of agency…

In Plato’s Phaedrus, Socrates worried that the invention of writing would cost us our memory and wisdom, because the ability to record knowledge on parchment would keep us from storing it in our heads. It may be true that writing has made us dependent on writing, but it also gave us the printing press, the internet, search engines, CTRL+F, and many other superpowers. So, overall, writing took a little of our agency, and gave us so much more.

The question is, will AI turn out the same way? Will the agency it gives outweigh the agency it takes?

I’d say it depends how much agency one already has, because agency typically can’t be given; it must be grown. Let’s take the example you give: of someone using ChatGPT to write flirtatious messages for them. A low-agency person would simply adopt the first pick-up line the chatbot generated. A high-agency person would give the chatbot multiple prompts and then carefully select the best pick-up line from among them, and in so doing, learn what works. So the low-agency person would use AI to eliminate choice, while the high-agency person would use AI to increase choice. The low-agency person would grow more dependent on the AI to think for them, while the high-agency person would use AI to help them think for themselves.

I therefore see AI as a personality amplifier; it will give more agency to those who already have it, and take more from those who already lack it.

I wonder what the long-term consequences will be of this Matthew effect. In H.G. Wells’ novel, The Time Machine, humanity in the far future has evolved into two subspecies, the Morlocks and the Eloi. The Eloi live lives of automated bliss, completely dependent on the machinery maintained by the Morlocks, who tirelessly toil underground. Generations of labour have kept the Morlocks sharp and self-sufficient, while generations of idle leisure have atrophied the Eloi’s minds, so that they never care to realise that the Morlocks are farming them for meat.

Since both agency and its opposite will compound in the AI age, it’s wise for people to maximise their agency now, because, years or decades down the line, it could be the difference between being whoever you want to be, and becoming an Eloi.

So how do we maximise our agency? We must first consider why we give it away in the first place. The reason is phronemophobia: humans are naturally averse to thinking.

In 2014, researchers at Harvard and the University of Virginia conducted experiments in which they left participants in a room with nothing to do except think or give themselves electric shocks. After just a few minutes, many participants began to give themselves the shocks. They preferred being zapped to thinking.

We’re configured to avoid thinking because cognition eats up a lot of time and calories, which in our evolutionary history were scant resources. As such, the brain is not so much a thinking machine as a machine that tries to circumvent thinking—it is calibrated to ration rationality. (This is why even the smartest people are dumb most of the time.)

The side-effect of this aversion to thinking is that people don’t want to be left alone with their thoughts. They’ll spend hours doomscrolling news of horrific tragedies rather than introspect. As Carl Jung wrote, “People will do anything, no matter how absurd, to avoid facing their own souls.”

The price of this avoidance is steep. Living as a fugitive from yourself makes you forget who you are and relinquish who you could be. Since you’re never listening to your own thoughts you fail to develop perspective, refine your beliefs, and make long-term plans. You cease to be proactive and your life becomes a series of reactions to your immediate environment.

Even worse, the less you think, the harder thinking becomes. All the screentime we engage in to escape our own heads stops us from developing a rich inner world, and with such barren imaginations we become even more dependent on external stimuli to keep us occupied.

So how do we make it easier for people to inhabit their own heads?

The simple answer is practice. To overcome the strain of thinking, do more thinking.

When the Industrial Revolution made it possible to live lives without physical exertion, going to the gym became necessary to stay fit. Equally, now that the AI Revolution has made it possible to live without mental exertion, we need the mental equivalent of gyms to stay sharp.

For me, the best brain-gym is writing—it forces you to shut out distractions and listen to your thoughts. A particularly useful form of writing is journaling, where you basically keep a diary in which you routinely interrogate yourself. What did you learn today? What did you regret? If you had ten times more agency, what would you do? If you did the same thing you did today every day, where will you be in ten years? (For those looking for an intro to journaling, I recommend my friend Elisabeth Andrews’ course.)

A habit of journaling helps you understand what you want (and don’t want), and it nurtures your imagination and acclimatises you to thinking for yourself. Daily journaling also acts as continual feedback by which you can evaluate whether you’re moving toward your goals. All of this brings order to the mind, so that, like a well-tended garden, it becomes a place we want to spend our time in rather than escaping at any opportunity.

So that’s my convoluted answer to your question. Write, even though machines can write for you, because the purpose of writing is not just to produce writing, but to distil your thoughts, refine your beliefs, and maintain your agency.

If Socrates had only spent more time with his pen and parchment, perhaps he would’ve realised that the thing he feared would cost us all our agency might ultimately be the thing that saves it.

That’s so true, what you say about writing as a way to maintain agency. Writing has always made me feel more in control of my life somehow, like I have more command over my thoughts, decisions, and what direction I’m going in. I think it helps you stay on track.

That is, of course, unless you start writing solely to please an audience.

Speaking of which, you seem to resist social media influence very well. I’ve been following you for years, and you are just as measured now as you always have been. You seem very intentional and, like me, you take your time with your essays. I see all these tips and tricks on how to grow a Substack now and I pretty much do none of them—the only thing I care about doing consistently is writing. Some people have told me I’m not doing enough—that I should publish more, react to breaking news, tweet every day, start a podcast. But to what end? More subscribers but less pride in my work? Host a podcast but constantly cringe because it’s not really me?

I don’t know if you’ve noticed this, but there seems to be an assumption now that if you don’t maximise engagement by every means necessary, you are slacking—rather than maybe deliberately resisting, doing this careful dance to avoid the abyss gazing back at you. People mistake being vague politically for cowardice, instead of trying to avoid becoming a parody of yourself. Or they assume you aren’t getting many opportunities, not realising you might be turning some down for the sake of your soul.

As you put it, “having the wrong audience would be worse than having no audience”. You warn that “being someone often means being someone you’re not, and if you chase the approval of others, you may, in the end, lose the approval of yourself.” I completely agree. I think if you build an audience but you aren’t being authentic, there’s always some sort of debt hanging over you. Even if you get followers and fame, something has been traded for it. The more you achieve on the outside, the more ashamed you feel on the inside. Because, as Johnny Cash put it, “you’ve still got the devil to pay.”

Anyway I’m interested in audience capture not only from the perspective of public writers, but anyone, particularly girls and young women. You don’t have to be some influencer for it to happen to you. Girls are now shopping online for who to be, funnelled toward the same face, recommended not just products but personalities. Maybe something more like algorithm capture makes more sense here—adapting to who the algorithm “wants” us to become.

So how do you resist this, both as a writer and as someone online? You mention staying deliberately vague, but is there anything else you do in your personal life? And what made you recognise the need for resistance in the first place?