The Human Premium Economy

01The Bifurcation Is Already Happening

The data on AI and wages is unambiguous in one direction: the premium for skilled, judgement-intensive work is rising. An April 2025 IMF working paper found that high-income workers' tasks are strongly complementary with AI, meaning AI amplifies rather than replaces their output — with productivity gains skewed toward higher earners, particularly those in professional, managerial, and knowledge-intensive roles. A software engineer directing AI agents ships more product. A consultant with AI research tools serves more clients. A designer with generative tooling produces at ten times the pace.

Meanwhile, the OECD's November 2024 report confirmed the other side of the split: roughly 27% of jobs in advanced economies face high automation exposure risk, with the sharpest pressure in data processing, customer service, and routine administrative work. The World Economic Forum's Future of Jobs Report 2025 projects 92 million job displacements by 2030, offset by 170 million new roles — a net positive on paper, but with a critical structural mismatch. The WEF found that 77% of employers are planning to upskill their existing workforce to work alongside AI, underscoring how many current roles will transform rather than simply expand.

The result is not mass permanent unemployment — it is structural change. The same force that moved 40% of Americans off farms and into factories, then out of factories into services, is now moving labour out of automatable cognitive work. The question is where it goes next.

02Where the Money Goes: Into Presence

The answer begins with how high earners already spend. The evidence is consistent across every major consumer data source: as income rises, spending shifts from goods to experiences at an accelerating rate.

The BLS Consumer Expenditure Survey shows the income gap is starkest in experiential categories. Households in the highest income quintile spend approximately $7,797 per year on dining out, versus $1,571 for the lowest quintile — a five-times gap that exceeds their 4.3× overall spending differential. Entertainment spending follows a similar pattern at roughly $5,100 versus $1,245. Total annual expenditure averages approximately $150,000 for the top quintile against $35,000 for the lowest.

This is not a passing trend. Bain & Company, in their 2025 Bain-Altagamma global luxury report, concluded directly: "After the shopping spree era, experiences and emotions have become the true engine of luxury growth." Three-quarters of consumers surveyed said they feel they get more value from experiences than from products — a number that rises sharply among higher earners.

When skilled workers get richer from AI leverage, their marginal dollar does not go into more software. It goes into presence — concerts, dinners, events, private wellness, hospitality, and the texture of being around other people.

03The Events Industry: Numbers That Demand Attention

The live events sector is not a niche. It is one of the fastest-growing industries in the world and structurally resistant to the very automation that is reshaping everything around it.

The signals within the industry confirm the premium shift. Average concert ticket prices rose to an all-time high in 2025 — approximately 45% above 2019 levels according to TicketHub industry data — while live sports attendance spending is up roughly 25% versus pre-pandemic, per Bank of America Institute reporting. The fastest-growing segment is experiential marketing and sponsorship, where global brands are investing heavily in event-based activations at a projected 14.28% CAGR.

The events sector is structurally labour-intensive in a way most industries are not. A stadium concert, a corporate summit, a food festival, or a private gala requires security, logistics, staging, catering, hospitality, transport, cleaning, audio-visual, and coordination — none of which scales down as the event scales up. Events are one of the few industries where headcount grows with revenue rather than being automated away from it.

The concert pricing signal is particularly telling. Ticket prices rising 45% above pre-pandemic levels are not inflation — they are a premium signal. People are paying significantly more to be in a room together. That willingness to pay will only increase as AI makes the digital world more abundant and the physical world more scarce by contrast.

04The Relational Sector as Absorber: Evidence It Is Already Working

The sectors that employ people in human-facing, physically present roles are already absorbing the labour market — not in theory, but in the data we can measure today.

Private education and healthcare — both deeply relational — account for just 17% of total U.S. employment but 56% of all employment growth between July 2023 and July 2025, according to Indeed Hiring Lab's August 2025 labour market update. Healthcare job postings are running 32.5% above pre-pandemic levels as of that same report — not because of a single-sector boom, but because of structural demand from an ageing population and increasing affluence.

The hospitality sector is following the same arc. Researchers at EHL (École hôtelière de Lausanne) have consistently argued that as AI handles operational efficiency, the premium on genuine human warmth and personalised service rises rather than falls. The Hospitality Net's widely-read 2025 analysis put it plainly: "Human labour may assume a premium status as personalisation, authenticity, and empathy become highly sought-after commodities."

The "feeling economy" framing from Roland Rust's research, published via EurekAlert, captures this dynamic precisely: "As AI is becoming more able to think, human intelligence is deemphasising thinking in favour of feeling and interpersonal relationships." The economy is not becoming less human — it is becoming more human in the areas that matter to paying customers.

05The Starbucks Signal: Why Automation Lost

The most instructive micro-case is Starbucks. After years of aggressive automation investment — removing baristas, tightening processes, installing ordering kiosks — the company reversed course in 2024. CEO Brian Niccol announced a strategic return to coffeehouse warmth: handwritten names on cups, ceramic mugs for in-store orders, improved seating, and more baristas per store. The data showed that these human touches measurably improved customer satisfaction and drove revenue, because the product was never just coffee. It was human contact, atmosphere, and the feeling of being a regular.

This pattern extends across every high-touch sector. The cafe that stayed warm beat the cafe that got efficient. The restaurant that trained its floor staff better beat the one that installed order tablets. The hotel that invested in its concierge team retained guests the algorithm-optimised property lost. Presence is the product.

06Why This Works Economically: Baumol and Non-Homothetic Demand

Classical economics gives us two interlocking frameworks that explain why this transition is durable rather than temporary.

Baumol's Cost Disease (now a feature, not a bug): William Baumol observed that sectors resistant to automation become relatively more expensive over time — not from inefficiency, but because the opportunity cost of human time rises alongside general productivity. In the AI era, as software, content, analysis, and logistics become cheap, the relative price of human attention rises. A skilled event coordinator, attentive server, or trusted personal trainer becomes more valuable in real terms as everything automated becomes cheaper around them.

Non-Homothetic Demand: A landmark 2021 Econometrica paper by Acemoglu, Lashkari, and colleagues showed that the dominant force in historical structural change was income effects, not price effects. As people get richer, they don't buy proportionally more of everything — they shift spending toward sectors with higher income elasticity. Agriculture has low income elasticity (you can only eat so much). Experiences, care, and human services have high income elasticity (there is always a better restaurant, a more attentive doctor, a more memorable event). Their analysis found income effects account for over 75% of observed structural change patterns historically.

David Autor's task-based framework completes the picture. Automation historically substituted for routine middle-skill work while complementing abstract, creative, and interpersonal tasks — producing labour-market polarisation. AI intensifies this: strong gains at the top, pressure in routine cognitive roles, and resilient or growing demand in in-person services where the human component cannot be separated from the value.

Daron Acemoglu and Pascual Restrepo's research on automation and labour demand provides the important cautionary note: displacement without new task creation is genuinely dangerous. The shift to a relational economy is not automatic — it requires policy infrastructure, career pathways, and wage support to ensure displaced workers can actually access the growing sectors.

07The Map of the New Labour Market

The post-AI labour market has a clear three-tier structure:

Six out of ten jobs people hold today did not exist in 1940. The next wave of new roles — experience designers, provenance certifiers, human-AI collaboration facilitators, community curators — will be just as hard to predict and just as real. The most common objection is "not everyone can be creative." The reply is that you don't need to be Picasso. You need to be the person whose involvement makes the experience feel like it was made for someone, by someone.

08Claude vs. Codex: Where the Two Analyses Differed

This article was independently drafted by both Claude and Codex before being synthesised and fact-checked. The comparison is instructive.

The synthesis: both are right. The market opportunity in events and experiences is real, measurable, and growing at over 11% annually. But it will not automatically translate to decent wages and upward mobility for displaced workers without deliberate policy choices. The economy will route labour into the relational sector regardless — the question is whether that sector becomes the foundation of a new middle class, or a more comfortable version of precarity.