The AI Industry Has Never Told You What’s In It for You

Here’s a number the AI industry should lose sleep over.

In a Pew Research Center survey from June 2025, half of American adults said the growing use of AI in daily life makes them more concerned than excited. Just 10% said the opposite.

When Pew first asked in 2021, the concerned share was 37%.

A year later, it’s worse. In Pew’s June 2026 survey, 52% said they were more concerned than excited about the growing use of AI in daily life, and just 9% said the opposite.

And it isn’t because people haven’t tried it. By February 2026, about half of American adults were using AI chatbots, up from a third in 2024.

Using it hasn’t made them more hopeful: 71% now expect AI to mean fewer jobs over the next 20 years, up from 64% in 2024.

It isn’t just Americans, either. In 34 of 37 countries Pew surveyed this year, people were more likely to expect AI to cost jobs than to create them.

Read that again. This is the most powerful technology of our lifetimes, built by some of the best-funded companies in history, and the public is moving away from it.

I don’t think that’s because people misunderstand the technology. I think it’s because nobody told them what’s in it for them.

What the industry pitches

Listen to how AI companies describe the future they’re building, and you’ll hear the same handful of ideas.

Capability. We’re building artificial general intelligence: systems that can do anything a human can do, and eventually much more.

Speed. AI will help build better AI, which will build better AI still, until we reach an intelligence explosion.

Abundance. All of this will make the economy so productive that everyone gets rich. Elon Musk calls it universal high income.

One of the most detailed plans out there, AI 2040: Plan A, hands every citizen a dividend funded by auctioning permits to build compute and robots.

Notice what’s missing. Every one of these is about the machines, or about money. None of them describes what your Tuesday looks like.

Now listen to what people say back.

What people hear

I’ve been reading reactions to AI on Substack, where a lot of thoughtful people are working through this in public. Three themes keep coming up, and each one points to a hole in the pitch.

Why are you building this at all?

Marianna Ganapini put it bluntly: she doesn’t understand why US labs and VC firms are so obsessed with AGI. Is it China? Is it playing God? Is it just because they can?

Luiza Jarovsky went further, calling AI a non-essential technology, and arguing that AI companies are pushing society to accept an unacceptable level of risk for it.

You don’t have to agree with either of them to see the problem. If you never explain what a technology is for, people will only see what it costs. Risk without a reason is just risk.

My work is going to get worse

Arvind Narayanan made the sharpest case I’ve read about AI and jobs, and it isn’t about replacement at all.

As more knowledge workers become managers of AI agents instead of doing the work themselves, he argues, they inherit everything hard about being a manager.

You stop practicing the craft you trained for. You’re responsible for mistakes you didn’t make. Your attention splinters across dozens of things.

Human managers put up with that because it comes with higher pay, higher status, and the satisfaction of watching people grow. Managing AI comes with none of that. And nobody chose it.

Compare that with a small detail Rich Holmes noticed: The Browser Company’s Dia credits the people who shipped each change in its release notes.

AI probably wrote a good chunk of the code, but a human decided what shipped and when. People want to see that. They want to know a person is still in charge, and still doing the part they love.

Who controls it?

Ben Dickson worries that power is concentrating in the handful of labs and tech companies that can afford the GPU clusters, data centers, and power plants. He wants AI that can learn without consuming so much energy, so that more people can build it.

Justin Bonanno makes the same point from the user’s side: everyone should learn to run a language model on their own machine. Ownership is agency.

And even inside the industry, nobody agrees on who controls the pace. Nathan Lambert argued that slowing down the frontier sounds good in principle but is dead on arrival in practice. Who would decide which benchmarks get to improve?

Put these together and you get a snapshot of how the public hears this pitch: a risky technology nobody asked for, built by a handful of companies who control it, that will make my job worse and might even take it away. Then, maybe, a check.

That’s not a story anyone wants to live in.

The pitch they should be making

Here’s what I think the industry gets wrong. It keeps selling the machine. People want to know what changes for them.

So here’s the pitch I’d make instead, in three sentences.

If a human wants to do it, a human gets to do it. Machines fill the work nobody wants, and get out of the way of the work people love. Narayanan’s reluctant managers go back to their craft.

You get the things, not a check. Food, housing, healthcare, energy and the rest, produced in abundance by machines and available to everyone. Not money that rising prices eat, but the stuff itself.

Your town owns the machines that make your essentials, and you own yours. Not three labs, and not a central government. Local, community-owned infrastructure that answers to the people who depend on it.

Every resident owns an equal share of those machines. This idea is called universal basic capital.

Put together, that’s universal basic goods and services, powered by universal machine labor on demand, distributed by universal basic capital.

Now, THAT pitch explains how AI will impact your rent, your health, your time, and your town.

It answers all three objections: it says what the technology is for, it protects the work people care about, and it puts control close to home.

This also happens to be the world I think we should actually build.

Where this goes next

This piece kicks off a series called The World We Want. It starts with a simple question, “what world do we actually want to live in?”, and works through how machine labor could get us there without taking away the things that make us human.

Start with the overview: Machines Can Make Us Both Free and Safe.

Then read on for the details: why a human should get first pick of any job, how to point the machines at what’s scarce, why quests should replace corporations, how every town could make what it needs, why universal basic goods and services beats universal basic income, and what history says about how we get from here to there.

The AI industry has had years to tell you what’s in it for you. They failed.

Here’s my attempt... so we don’t throw out an abundant future with the AI company’s backwash.