AI is Rewiring the Economy

You either believe AI will displace jobs or you think its hype. I think that’s the wrong question. The right question is how does AI reshape the economy. 

AI will force us to reconsider commerce, consumerism and the norms of our economy. We will enter a world where consumers buy less stuff, but with much higher conversion. Middle-income consumer populations will have less disposable income as their jobs come under pressure from AI. Meaning consumerism ceases to be the driver of economic growth.

This won’t happen overnight. We’re only just seeing the earliest signs.

This economic shift happens in three phases: 

• First, AI commoditizes knowledge work, destroying consumer purchasing power. 

• Second, human behavior shifts from consumption to goal-seeking as AI handles routine tasks. 

• Third, new business models emerge that profit from human flourishing rather than human spending.

The “is it really taking jobs” is a bit of a lightening rod. But if you take at a minimum that junior level jobs are not appearing. This is worth your attention.

My thesis is this will be much much bigger in the next 3 years. Microsoft is laying off 1% of its workforce per month on average. If that trend continues, disruption follows.

Sam may have pricked the AI Capex bubble, but the 20-year trend line is much more interesting.

AI will reduce the cost of services like containers did to goods.

In 1956, hand-loading a ship cost $5.86 a ton. After containerization, it cost only 16 cents a ton. That’s a 97% reduction. This drove consumerism and mass production. The quantity of goods carried by containers soared from 102 million metric tons in 1980 to about 1.83 billion metric tons by 2017.

The consumer and consumerism became the engine of global growth. Advertising and sales became the key business model driving all of this commerce.

What happens when AI does that to services?

It’s already true that AI is capable of entry-level service work. ChatGPT’s latest benchmark showed that Claude 4.1 is 97% as capable as human experts in 44 distinct tasks like financial analysis, healthcare analysis, law, and more. This is limited to asks, not entire jobs for now. But that’s coming.

Service businesses typically report gross profit margins in the high-90% range, while goods-producing companies like automotive businesses achieve only 9-12% gross margins.

Law, consulting, and finance will become lower-cost to distribute, and the competition will shift to who can bring once rare, high-quality knowledge work to the mass market. Look at the earnings from Accenture and the big outsourcing shops, they’re in the middle of job cuts. And Tech too has stopped hiring. 

This service-ism has the potential to drive a new type of growth and potentially challenge the standard business model of internet-based businesses today.

The arrival of containerization is a fascinating parallel to who wins in AI.

1. The largest incumbents immediately recognized the potential.

2. The government and policy ecosystem immediately pushed back.

3. Infrastructure (ship builders, data centers) got a quick boost, but not long lasting

4. Spending on capex and talent balloons just as pricing comes under pressure

In shipping, what happened next was investors pushing for M&A (Winfsurf anyone?), vertical integration became normal (models buying apps), and the benefit comes to those downstream of the supply chain.

We’re about to remake the global supply chain of knowledge work.

And you can see this in where investment is heading. It’s not to office buildings for knowledge workers.

This will destroy the income of today’s western “consumers” as AI competes for jobs.

If nobody has a job, who buys all the stuff?

How does agentic commerce work if AI destroys middle-income consumer wealth?

AI will add anywhere between 1% and 7% to GDP between now and 2030, but it won’t do that with net new employment. 

Job losses are accelerating, and every other day, I hear stories of companies cutting headcount in departments by orders of magnitude not percentages. High-paid knowledge workers are losing the jobs and they’re the consumers from western nations who buy all the useless crap our global supply chains can advertise to them. 

In “digitization,” you could expect a 10% staff reduction. Then there’s Opendoor, aiming to cut at least 85% of its staff.

• About 100 million employees in the US are estimated to be knowledge workers

• 61.5% of total US employment was tied to consumer spending

If 85% of those lose their jobs, who is the consumer doing the consuming?

Ideally, we need AI to generate a proxy for the standard of living. In health, well-being or by dramatically creating deflation in the things that are beyond reach for many. If AI can make stuff cheaper to produce and buy, could it also help us as consumers live better for less?

This makes me wonder if the advertising, commerce, and aggregation theory model of 2000 to 2020 is fit for purpose for agentic commerce and an AI economy. And if they’re not, the business models that we assume are the right ones don’t copy+paste on to 2040 very well.

Which brings us to a critical question: if AI companies are destroying their own customer base, how do they plan to monetize? The answer lies in understanding how platforms evolve under economic pressure.

AI Platforms Will Evolve Towards new Business Models

Platforms tend to go through three phases (as described by enshittification, yes, it's a real thing, it has a wikipedia page)

1. Stage 1: The Golden Era (User-Centricity). High quality, free, but loss-making. Users love it. It grows incredibly.

2. Stage 2: The Shift (Business Customer Focus). Begins to serve ads, sponsored content, and focus R&D on monetization.

3. Stage 3: Full Enshittification (Profit Maximization). The Ads begin to dominate, your data is the product. 

Google Search is in Stage 3: Google started as the fastest way to find information on the internet, they even showed you the response time in the search query. Over time, they added ads, made it less obvious they were showing you ads, and added more as the top search results, so finding the actual top match was harder.

AI is in Stage 1: For free, or $20 per month, you can make an image of just about anything. AI is indistinguishable from magic. 

AI has a Unit Economics problem. Each new model requires more tokens for better outputs. 

Need more daily active usage = burn more tokens

Need better unit economics = extract more value from users

These demands are opposite.

Somehow, AI (as a platform or product) has to figure out a way to align the interests of daily active users with better unit economics. That rarely means increasing the fee you charge them directly. 

If AI becomes a platform, it will be the most precise advertising and commerce engine the world has ever seen. Able to drive incredible conversion, but perhaps at the expense of considering a product you never would have done before.

• AI Search converts at 10 to 40%*, vs Google at 1 to 2%. How will AI companies resist the pull of that revenue?

• Shopify found that Agents can help increase conversion when checkout is embedded. 

* These numbers might not last, and AI search presents fewer choices.

AI companies are burning billions to reach Stage 1 perfection - making their platforms irresistible to users whose purchasing power is simultaneously being destroyed by that same AI.

The irony is incredible. 

We created the perfect shopping platform as buying power begins to disappear.

User Centric AI Replaces Consumerism - And becomes the competitive dynamic.

 In a world of:

• Fewer consumers with money

• Perfect conversion from AI

• Abundant cheap services

Revenue comes from delivering outcomes, not selling stuff. The AI agent becomes a personal optimizer, not a personal shopper.

Capitalists tend to view AI through the lens of commerce.

But what do I need to advertise for if the intent can be matched to a product more closely? In a world where consumers have less disposable income, what if agents could align more closely with a user's best outcomes and goals?

This isn't idealistic thinking—it's economic necessity. When your customers lose purchasing power, you either find new revenue sources or go out of business. The companies that figure out how to monetize human flourishing rather than human consumption will define the next economic era.

The AI companies have an opportunity to balance the goal of engaged daily active users (generating revenue) with better unit economics if they can create an incentive structure for something other than shopping or viewing ads.

AI can reduce the cost of high-value services dramatically. This positions AI companies to monetize outcomes rather than consumption.

What replaces commerce as an experience?

Shopping is the solution to boredom.

Finding, selecting, then buying a thing creates a little moment of joy, because there’s nothing good on TV that can hold your attention long enough. We’ve optimized our entire economy and global supply chain around this fact. 

The internet itself is optimized to get you to consider and buy, stuff.

AI narrows the window of consideration.

It presents one or two answers, instead of all of them. 

If AI Agents take care of our weekly shopping needs, help us find food to enjoy, and take care of the services that eat our time (filing taxes, mortgages, applying for and managing schooling). We end up with more time to pursue higher goals like entrepreneurship, creativity, and health or wellness.

What replaces shopping is goal seeking. And AI helps us achieve those goals.

If you look at some of the ultra-wealthy today, what they value changes. They’re less likely to doomscroll Temu for useless shit, and more likely to view longevity and health as a status symbol. 

What replaces that is AI experiences that compete to add value to the end user defined by their health, wealth or wellbeing (loosely grouped as flourishing). 

AI Could Expand Human Flourishing. 

The sectors most impacted by AI are the ones that impact the human experience. 

From AI will not make you rich.

Imagine if we could bring world-class healthcare, education, finance and creative services to anyone for almost zero cost. People able to afford these services don’t doomscroll looking for more stuff. They goal seek to flourish more. 

If there are no consumers, there can be no consumerism.

And if the gains from AI manufacturing only go to a small handful of people.

Can we challenge ourselves to define a new type of commerce and economics?

One where longevity, flourishing, and connection are fulfilled by the massive deflationary pressure of AI and robotics re-making our value chain, simplifying commerce away from our overloaded, dopamine-addled prefrontal cortex. 

And leaving us with an economy that seeks a new north star.

Not GDP, and GDF. Gross Domestic Flourishing. 

This won't happen overnight. Like all technology transitions, we'll go through a skeuomorphic phase - using new tools for old purposes:

1. Phase 1 (Today): We'll apply AI to existing commerce - better ads, smarter checkouts - but volume remains low because fewer people can afford to buy

2. Phase 2 (2027): Winners emerge by commoditizing knowledge work, creating abundance in services, while scarcity emerges in consumer spending  

3. Phase 3 2030+: Consumer-facing AI shifts from facilitating purchases to optimizing human outcomes.

But the growth companies are all scaling knowledge work not widgets. Cursor or Harvey scale engineering and law. Which tells you what your job is as an entrepreneur, business, or investor in the services sector.

So what does this mean for you? The transition is already underway, but most companies are still optimizing for the old economy. Here's how to position for the new one:

Your Job is to Scale World Class Human Experiences

Shipping and containerization are the perfect metaphor for AI.

The remaking of global supply chains foreshadowed a surge in consumer spending as the primary driver of Western GDP growth. Consumer standards of living were kept high by pushing consumer debt and asset ownership. But at the cost of lower quality products that gradually made us sick, and increased the costs of healthcare.

AI and robotics allow us to take creativity and the highest quality of services in knowledge work and make them available to everyone. The AI opportunity lies in scaling the ability to deliver flourishing and better outcomes for consumers (or, more accurately, flourishers).

In that world, growing revenue is about radical customer centricity. 

Louder for the people at the back.

Your revenue growth will come from improving outcomes for your customers.

Since AI agents will choose products based on outcomes rather than marketing, your competitive advantage shifts fundamentally

To unlock true consumer-facing AI-first experiences, we need to focus on bringing high-quality products to the mass market, where competitive advantage lies.

Advertising and commerce are the wrong mental model for that business. Don’t get me wrong. Short term, agentic commerce will happen. Agents will buy stuff. But my point is that it is a commodity. It’s a standards and timing question. Does OpenAI want to become a Merchant of Record, or will we embed Apple Pay in Agents? Pick one.

The bigger game is the new business model. For that, the right mental model is a personal shopper, a personal coach, and a doctor whose incentives are to help your customer become the person they want to want to be. Their better angel, not lesser demon.

This transition is already underway. The companies building for 'Gross Domestic Flourishing' instead of Gross Domestic Product get to define it.

Be that company.

ST.

Closing thoughts

• I have about 50% confidence in the above essay. It’s a nice story, but I can’t help but sense the lesser demons of slop and dopamine, ad-driven goo will fill our cortex for decades to come (save for some anti-slop renaissance).

• One day, I’ll start the AI Brainfood newsletter. But I’ll need sponsors and a team before then—hint, hint, those of you staring at your 2026 budget.