At least you’re likable

What are you worth when nothing is scarce?

Tech high priests like Sam Altman and Paul Graham are preaching taste (an idea that has actually been percolating in Silicon Valley for a while). Founders are taking etiquette lessons. Starbucks just announced it is rolling back automation and hiring more baristas because customers want to be served by humans.

Meanwhile, a growing number of people are grappling with what human work looks like in an AGI world of abundance. For most of human history, economic value came from what you could do and how fast you could do it. But if the machines are smarter, faster, and stronger across virtually every task, what’s left?

Alex Imas wrote a fantastic essay arguing that as AI automates commodity production and drives down prices, the scarce and valuable thing won’t be cognitive skill or even labor in general, it will be human presence itself. This view concedes various contested points to those most bullish about future AI capabilities. He assumes the AIs would be categorically smarter, faster, stronger in every task. The only glimmer of hope for our utility would come from humans retaining value not by outperforming AI but simply by being human.

What would this world look like? A pessimist may point to “Nosedive”, a 2016 Black Mirror episode, as a reference. In it, society has embraced a technology where everyone rates their interactions with others on a scale of one to five stars, and those ratings cumulatively determine socioeconomic status. The world is pastel pink and homogenous. People are sickly sweet and infinitely boring. Relationships are mediated entirely by rankings, rendering them transactional and flat.

I watched this episode at sixteen and – comparing it to the Darwinian arena of secondary school – found it to be both terrifying and believable. If Imas is right, what goes wrong in “Nosedive” is exactly what we need to think through. Revisiting it now, I want to understand why it is so dystopic. Is it that the world is homogeneous and flat? That it’s coercive and illiberal? That it’s anti-meritocratic?

Every field has its gods

The “Nosedive” world is manicured and pastel precisely because everyone is optimising for the same thing; there is a single master ranking. But that is not necessarily what happens when human presence is the scarce good.

Imas uses Girard to argue that in an era of AI abundance, demand for status goods doesn’t get satisfied – it relocates to what is scarce and desired by others (in his account, the relational sector). The natural follow-up question is: what will be desired by others and scarce? I think Girard can help us here too.

In my view, Girard’s most interesting insight isn’t that we want things because other people want them. It’s the distinction he draws between types of mediators. When the mediator is distant (a celebrity, a historical figure, a saint) desire is quietly aspirational. There’s no competition because you don’t share the same world. But when the mediator is close (a colleague, a sibling, someone of your world) desire becomes obsessive and self-reinforcing. The more someone like you wants something, the more you want it.

This suggests desire clusters into micro-climates, which is precisely what Bourdieu means by fields. Society isn’t one unified arena but a cluster of fields with their own rules, logics, hierarchy, and dominant form of capital. Status competition happens within fields, not across them. Which is why a junior academic doesn’t burn with envy toward a hedge fund manager, and a chef doesn’t lie awake thinking about a senator. And why the demand function for luxury goods differs in SF, London, HK and Abu Dhabi.

Seasonal Clickfarm Worker@ClickingSeason

The problem isn’t your hobbies. The problem is that you aren’t good enough at your hobbies. There are groupies for every annoying endeavor on earth. Video game groupies. Fly fishing groupies. Marching band groupies. She’s repulsed not by your hobby but by your mediocrity.

8:12 PM · May 6, 2026 · 221K Views

---

100 Replies · 66 Reposts · 1.6K Likes

So we might expect that in Imas’ post-commodity world, the relational sector won’t be one unified destination for desire — it will fracture into many fields, each with its own hierarchy and vision of excellence. Some hierarchies will be more portable than others (wealth functions this way now, as does, to some extent, fame). But there won’t be a master ranking. Value will be plural.

And the number of fields may actually grow. Before the internet, culture had to clear a high bar to exist. Whatever reached an audience had to justify its slot by appealing to enough people in the same place at the same time. The internet collapsed that constraint. Chris Anderson called this the long tail: when distribution costs fall toward zero, niche communities that couldn’t form locally become viable globally. This is how you get marble racing fans, ASMR listeners, and vexillology forums.

AI may do something analogous, but at the level of production rather than distribution. When it costs almost nothing to make content, tools, or goods for a small audience, viable fields could multiply again. But proliferation of fields doesn’t mean proliferation of winners within them – Rosen’s superstar economics may exist at every level of the long tail. The best fountain pen reviewer doesn’t share their audience with the tenth-best; the field is tiny and the hierarchy inside it steep.

Wait, it’s incentives? Always has been.

The rating system feels oppressive because people are visibly performing for approval and every interaction warped by a number hovering overhead. But in a way this is just incentives made legible: when something is rewarded, people do more of it.

“Nosedive” is right to think that if human presence becomes economically valuable, people will optimise for likability. They will become warmer, more attentive, more charming – because it pays. Markets have always done this. Literacy, coding, and the ability to read P&L sheets are all skills that developed and propagated because the market ascribed value to them.

Is this coercive? Incentives have the ability to completely shift our perception of the world. In “Time, Work-Discipline and Industrial Capitalism”, E.P. Thompson charts traces the shift from rural work, where work and life bled into each other, to the clock-governed discipline of the factory. Thompson argues that by rewarding punctuality, the market changed the perception of time in the general population. He frames this as a market pathology, but incentives reshaping human behaviour predates capitalism by millennia.

Literacy changed how our brains are wired. MRI data shows that learning how to read rewires the visual cortex and territory that once processed faces gets recruited for reading. This means literate people are measurably worse at certain face recognition tasks than illiterate people. The body responds, over time, to what the environment asks of it.

Every system has incentives. The USSR just replaced market signals with state signals. The desire for status, safety, and survival propelled things like the Stakhanovite movement. Religious incentives can be just as strong as worldly ones. Weber’s argument in The Protestant Ethic is that Calvinist theology (the belief that worldly success signals divine election) helped produce Capitalism. And believing in hell is good for GDP — the threat of eternal punishment turns out to be a remarkably effective incentive for hard work and honest dealing.

Arguably, we have always lived in a “Nosedive” world insofar as social skills have been rewarded. Among the Hadza hunter-gatherers, popularity is one of the two determinants of social status. In Louis XIV’s Versailles, proximity to the king was the only currency that mattered, so the entire French nobility spent their careers mastering etiquette, flattery, and social performance. Adage suggests that in our bytes-heavy world, the awkward engineer reigns supreme, but research shows the importance of social skills in the labour market has increased since the 80s – even in fields like tech.

Every major shift in how we organise production has changed the skills that get rewarded and changed human behaviour. Farming rewarded physical endurance; the factory rewarded the ability to tolerate repetition; the office rewarded cognitive horsepower. And people responded in kind.

A new meritocracy

Michael Young coined ‘meritocracy’ in his 1958 satirical novel The Rise of the Meritocracy, combining meritus (deserving) with kratos (rule or power). He invented the term to criticise what he saw as the hidden cruelties of a society where rule by the ‘deserving’ had replaced aristocracy. He worried that by hoovering up every talented person from working class communities and rewarding them for leaving, meritocracy would strip those communities of their natural leadership, their internal culture, and their ability to make sense of their own situation with dignity.

Imagine how horrified he was when centre-left politicians like Blair and Clinton started lauding the term as something we should actively aspire to build. The settlement of the 1990s gave meritocracy an unlikely coalition of defenders. The left abandoned class politics in favour of opportunity; the right abandoned deference to tradition in favour of earned success. What in part made this settlement stable was that the cognitive economy seemed to prove the model worked. It had the pretence of objectivity: outputs are measurable, credentials are legible, and “education, education, education” was a defensible rallying cry.

But the cognitive economy’s objectivity only went so far. Your code compiled or it didn’t, but the ability to write the code was never randomly distributed. IQ is 50-80% heritable. Conscientiousness – the trait that best predicts job performance – is around 60% heritable. Social skills are roughly as heritable as IQ. Thinkers as different as Rawls and Hayek agree that the natural distribution of talents is, from a moral point of view, arbitrary. No one earns their starting position.

You don’t have to believe a surgeon deserves her salary to justify paying it. The functional case for high pay is that it aligns private incentives with social needs. The market rewards outputs that are scarce and valued. And high pay for rare skills does two things at once: it signals where society needs people, and it incentivises them to get there.

Rawls would accept this argument but insist it come with a condition. Inequality needs to be justified. His difference principle says it is legitimate only if it benefits the least well-off members of society. Hayek’s response would be: who decides? Rawls sounds reasonable in the abstract, but his system requires someone to judge whether inequality is actually benefiting the least well-off. That requires information no central authority has. Better off how? By whose measure? The difference principle assumes a universal definition of the good that doesn’t exist. And the attempt to answer these questions is ‘the road to serfdom’ – a slow slide from benign intervention to centralised control, paved with good intentions.

Young, Rawls, and Hayek could argue for hours and in a sense, Western political thought has been, by proxy, for centuries. What matters for our purposes is how it changes (or doesn’t) when the economy that gave meritocracy its objectivity starts to erode. To answer this question, I think we need to determine whether a post-commodity future of work would be legitimate (whether most people accept the system as fair) and efficient (whether the system produces good outcomes).

Relational skills are roughly as heritable as IQ. If the relational economy simply replaces one set of inherited traits with another, it isn’t obvious that it’s less fair or legitimate. The question is whether the nurture component of relational capital is more or less class-stratified than the nurture component of cognitive capital.

One view is that taste, cultural fluency, the ability to move through elite social environments correlate with background and so a relational economy is a uniquely classist one. I’m not so convinced. Figures like MrBeast, Clavicular, and PewDiePie are not performing class-mediated cultural capital, and yet they command immense amounts of relational capital and attention. A plural relational sector is likely one that produces status micro-climates that aren’t all gatekept by the same cultural formation. The relational economy may not be more legitimate than the knowledge economy, but it is far from obvious that it would be less so.

As for efficiency, the logic is that a free market in the relational economy rewards people for treating each other better, being more interesting, and winning in their chosen status micro-climate – and that is a good thing. I, like many others, do worry about when markets reward things we find unsavoury or morally bad. If desire clusters into micro-climates with their own hierarchies, some of those hierarchies will organise around things most of us would rather not exist (like AI girlfriends, the OnlyFans creator economy or AI bait).

But it isn’t clear that the discomfort is tracking something unique to the relational economy rather than just a general unease that attends any major reorganisation of what gets priced. When life insurance emerged in 19th century America, it was widely condemned as a profanation that turned death into a vulgar commercial transaction and mocked divine providence.

In any case, the current setup isn’t morally clean either. Will we look back and find it grotesque that we took the most credentialed minds of a generation and had them spend their twenties reformatting pitch decks at two in the morning? It’s not obvious to me that we end up somewhere worse with a relational economy.

The revenge of the jocks

I am dispositionally an optimist. Every transition in how humans have organised work has looked, from inside it, like civilisational catastrophe. From farm to factory to filing cabinet, each time our economic model shifted, the world got stranger and produced new indignities and, by most measures that matter, got better.

Who wins in this new world? I would put money on those who have strong social skills and are able to build parasocial relationships to accumulate cultural capital. People are starting to realise this, which is, as my friend Alys pointed out to me, likely why OpenAI bought TBPN. Major transitions tend to reward people who are early to understand the new logic of value, not necessarily those best suited to the old one.

It’s really hard to predict things. I’m not sure if we will see a world of strong AGI, the impacts of AI on the economy are highly contested, and I haven’t yet wrapped my mind around what would happen if agents become legal entities.

But if we end up in a post-abundance world and, as Imas says, the scarce thing becomes simply human presence (not brilliance or not credentials, just being there) then the advantage is be, in some sense, universally distributed. Everyone is irreducibly human in a way that no one was irreducibly good at calculus. This advantage likely won’t be evenly claimed. Arguably, knowing that humanness is now the game is itself capital – a form that will accrue first to those paying attention.

Despite my attempts, even this essay feels stuck in the past – all of my examples are arguably from the attention economy, mediated by a screen. In a world where human contact is scarce, we might see the re-physicalisation of work. If the relational economy delivers anything worthwhile, it could be more of us tending to each other. A cure for the loneliness the screen helped create is likely worth a lot of GDP in an abundant world.

What we can say is that the relational economy does not have to be the “Nosedive” dystopia. A free and plural post-commodity future of work is possible. This world wouldn’t be a utopia but it might be one where a teenager at the back of a classroom could sense, correctly, that there is more than one way to exist and be fulfilled.

Thank you to Alys, Zhengdong, Will and others for your thoughts, criticisms and encouragement.