What happens when the elite-human-capital jobs get automated away?

It's looking like AI will be no respecter of persons

Yet one fine day, in a fit of euphoria, after he had picked up the telephone and taken an order for zero-coupon bonds that had brought him a $50,000 commission, just like that, this very phrase had bubbled up into his brain. On Wall Street he and a few others, how many? Three hundred, four hundred, five hundred? Had become precisely that . . . Masters of the Universe. There was … no limit whatsoever! Naturally he had never so much as whispered this phrase to a living soul. He was no fool. Yet he couldn’t get it out of his head.
Tom Wolfe, The Bonfire of the Vanities, 1987

PwC partner income rises to $814,000 due to AI adoption, fewer staff
AFR headline, 16/4/26

How AI shrank a 40-person PwC consulting team to just six
AFR headline, 17/4/26

But this stasis is just the rancid icing on the spoiled cake of much longer trends. A new working paper from three economists explains how the U.S. job ladder has been breaking for 40 years. This decades-long problem doesn’t affect just Gen Z but has also stymied wage growth for Americans in their 30s, 40s and 50s. This could be why the midlife millennials I interviewed a few years back felt that they were physically in their 40s but economically in their 20s, unable to find a career that felt secure.
Jessica Grose, NYT, 22/4/2026

Imagine you are 18 and what my fellow Substacker Richard Hanania calls elite human capital.

You are of above-average intelligence and expect to join one of your society’s sense-making or decision-making institutions, securing a role that will (a) facilitate a ‘comfortable lifestyle’, (b) give you disproportionate influence over your society’s culture, economy and politics, and (c) provide a reassuring amount of social status.

Necessarily – but not sufficiently – you’re 1-2 standard deviations above the mean IQ. You’re in the 115-130 range – part of the top 15 per cent of the population in cognitive ability.

Now, let’s be realistic here, you don’t have a 130+ IQ. Some paths are closed to you. Even under the stern tutelage of the most unforgiving of immigrant tiger parents, you’re unlikely to make it to medical school.

But you’re still smarter than most. You should be able to secure a job befitting your intellectual abilities and life ambitions. You’ll be getting a serious degree – law, accounting, engineering, economics, commerce, business, computer science – from a top-tier university. Then you’ll join a law firm. Or a management consulting firm, ideally McKinsey, BCG or Bain. Or you’ll hitch your little red wagon to one of the Big Four professional services firms – PricewaterhouseCoopers (PwC), Deloitte, EY and KPMG.

You’ll be pulling down comfortable-lifestyle-enabling money from the get-go. Indeed, you’ll start on a salary that’s more than many of your compatriots will make at the peak of their careers. If you make partner, you could be earning 10x the median salary by your mid-thirties. A few years of earning that kind of coin and you’re edging into “fuck you” money territory.

At the very least, you can afford a lovely home in a desirable location, expensive clothes, luxury cars, pricey restaurants, private school fees and regular overseas vacations.

If you’re ‘encouraged’ to retire in your mid-fifties, you’ll be able to do so easily. Maybe you’ll sit on a few boards, perhaps you’ll take it easy after several decades of 60-hour-plus weeks and up-or-out work cultures.

At least, that’s the way things used to be. But it seems AI is now coming for human capital of all varieties.

Elites are shrinking
The articles referenced in the AFR headlines above are paywalled. But here’s the tl;dr: last Thursday, Australia’s business paper ran a story about PwC Australia lifting average partner pay despite falling profits. How did they manage that? With AI-led efficiency gains and reduced headcount, of course.

Last Friday, the AFR explored how the increased efficiency and reduced headcount came about.

PwC used 18 AI agents to replace much of a 40-person consulting team and cut a project’s timeframe from 18 months to 90 days. AI handled coding, testing and documentation, boosting speed and quality, while humans shifted to oversight.

I imagine The Wall Street Journal, Financial Times, Handelsblatt, Les Échos, Nikkei and the South China Morning Post have similar stories about elite human capital being displaced by AI.

AI is ‘spiky’ and struggles with some things, such as spatial reasoning or novel-problem solving. But it is superb at what lawyers, management consultants and professional service firm partners are selected for – verbal intelligence, pattern recognition, recall and synthesis.

So, what happens when societies that were once broadly middle class and then structured on a 20:80 basis (with lawyers, management consultants and professional services providers comfortably ensconced in the top quintile) undergo further Brazilification?

Those unable to secure a grad role at a fancy law, management consulting or professional services firm will likely receive roughly as much public sympathy as those upset their PhD in lesbian interpretive dance only led to a job as a barista.

But while non-elite human capital will struggle to sympathise with would-be Masters (and, nowadays, Mistresses) of the Universe, they should be able to empathise with them.

The post-WWII decades saw the creation of a world where there was a lot more opportunity to get ahead economically. This was followed by the neoliberal era, where the rewards for getting ahead became much greater, but the opportunities for doing so became more constricted.

In the glorious three decades following WWII, a man could drop out of school in his mid-teens and, via a blue-collar but reasonably paid job, enjoy a middle-class lifestyle involving home ownership, marriage and children.

From circa 1980, opportunities to “get ahead in life” began contracting for Joe Sixpack. At the same time, the rewards flowing to cognitive elites started increasing.

However, if you were sensible enough to “get a good education”, the (increasingly globalised) world was still your oyster.

But then the supply of university graduates began outstripping demand. If you went to the sort of university that law, management consulting, and professional services firms recruit from, you’d probably still be able to achieve a lifestyle involving home ownership, marriage and children. Otherwise, you’d be joining the ever-expanding ranks of downwardly mobile graduates and manning the espresso machine at Starbucks.

Thanks to AI, we seem to have reached the final stage of the process. Now, even the cream of the human capital crop – the ones with above-average IQs, the ones who went to good universities, the highly conscientious, highly competitive, gratification-delaying overachieving ones, the ones who grew up being told they were destined for big things, the ones who did all the right things – are struggling.

A new aristocracy
If there are ever fewer but increasingly well-remunerated jobs and growing numbers of people leading a near-subsistence lifestyle financed by gig-economy work and/or a modest UBI, that sounds a lot like a neo-feudal society to me.

I’m not the only one who’s had that thought. The New York Times recently published an article entitled We Are Witnessing the Rise of a New Aristocracy by Jennifer M. Harris

Here are the money quotes:

As A.I. transforms anything touching a keyboard, it will land first and hardest on the income ladder’s middle and lower rungs…

When more people are out of work, those who still have middle-class jobs — the marketers, the air traffic controllers and so on — will have less bargaining power, reducing their wages. A.I. can reinforce this downward spiral by lowering the expertise required for these remaining middle-class jobs…

The flywheel that turns wealth inequality into democratic backsliding is already churning. If we want to keep the “democratic” in democratic capitalism as A.I. takes hold, we’d better get to work.

Harris, a former Biden economic official, is chiefly concerned with the prospects of low-income and middle-income workers. She believes elite human capital, or at least old-money elite human capital, will be relatively insulated from AI.

Premiums for elite graduates with hefty Rolodexes full of powerful people, and tacit knowledge (like how to generate a laugh at a cocktail party on Park Avenue), aren’t going anywhere. Chatbots are no substitute for people who can call the right people when high-stakes deals go awry.

But regardless of exactly how soon and severely elite human capital is impacted by AI-driven automation, the broader point stands. There are going to be a lot fewer people invested in the status quo and a lot more with nothing much to lose. If Harris’s data is right, we’re now in the process of moving from 20:80 societies to 3:97 ones:

OpenAI and Anthropic, which are already operating globally, employ only a few thousand people. Microsoft employs more than 200,000, and Amazon employs 1.5 million. The picture that emerges isn’t of just a deepening of the current divide. The A.I. story is one of more extreme concentration of wealth — at most likely not more than 3 percent of households, the very few who hold ownership in these A.I. companies or in the mostly private firms financing them.

Harris ends her article by arguing deepening Brazilification isn’t “a foregone conclusion”. She argues the process could be reversed by designing AI to augment workers, taxing capital more like labour, and broadening ownership of AI-generated wealth through public or worker stakes.

I don’t know how politically feasible any of those proposed solutions are. But I’m sure many in the elite human capital class will be loudly advocating for such measures in the months to come.