There's a drawer in my kitchen doing a lot of heavy lifting. The thing is stuffed with omega 3s, fiber supplements, creatine, and enough protein snacks to suggest a small gym operates on the premises. I track my blood pressure; I've used a GLP-1, try to manage sleep, and I'd recommend it to anyone who asks.

I assumed all of this was a personality quirk (I have plenty of those). I am in fact basic.

The New Consumer's mid-year Consumer Trends report with Coefficient Capital says it's a demographic. 36% of Americans describe themselves as in “optimization mode,” actively making choices to improve their energy, appearance, and long-term health. They skew millennial and affluent. They report a life satisfaction of 7.3 out of 10, a full point above everyone else. And they buy things through AI at nearly twice the rate of the general population.

Guilty.

In May, I wrote in AI Checkout that the protocols for paying via AI were shipping, Walmart's AI checkout was live, but I couldn't find evidence that customers were adopting AI for payments en masse. “Maybe they are,” I wrote. “Maybe it's just early.”

Found them.

I was looking at infrastructure when I should have been looking in the mirror.

“The Optimizer” demographic also solves a puzzle. How is consumer spending healthy when the vast majority of the consumer population is struggling?

Confidence surveys look grim, tariffs are still working through prices, and somehow Prime Day just did $26.4bn online, up 9.3% on last year and is now only 20% smaller than Black Friday Cyber Monday. The consumer despairing in surveys and the consumer setting spending records are different people. Many things can be true. The Optimizer is pulling the averages.

Today:

The new consumer survey data was fascinating:

Maybe it’s because we were the last generation to be born before the world got weird. Who played outside, lived before the internet and smartphones, or maybe it’s because we’re hitting our stride financially. But we millennials just seem happier, and that’s showing in our use of AI and approach to life.

Millennials out-use Gen Z on daily AI, 41% to 31%. Gen Z tends to be more worried about AI than hopeful. The millennials went from building their SIMS and World of WarCraft avatars to optimizing their IRL bloodwork. Skyrim has a lot to answer for.

When it comes to what are those consumers using AI for, things get a lot more interesting.

Health questions are the number one consumer use case for AI tools in every age group, and finding or comparing products is number two. The same person buying the creatine asks which creatine, whether it clashes with their blood pressure meds, and where it's cheapest per gram. The Optimizer uses AI to run a kind of consumer-grade procurement in daily life.

If a third of Americans, and the third that's trading up, run purchasing decisions through a chat window, everything downstream of that window changes.

Shopify's Q1 data shows AI usage is changing the funnel of commerce incredibly quickly (covered previously on brainfood here)

These customers are showing up much later in the funnel with higher intent. And, of course they are, they did all of their consideration in the chatbot. 55% of AI-referred sessions start on a product page. From organic search, that number is 20%. The customer also arrives anonymous. OpenAI shares nothing about the conversation that produced the visit, so you get a high-converting stranger and no idea what convinced them.

For now, the trade is spectacular for brands.

MoffettNathanson's Michael Morton tracked where AI platforms send commerce traffic and found a brand’s.com page was outpacing the larger commerce marketplaces by 4-8x. In the Google search era, that ratio was roughly 2x. AI is giving the brands higher-intent customers, direct to your own site, and nobody charging for the click (well, not yet anyway).

ChatGPT pulled back its integrated checkout ambitions and leaned into ads instead, because cheap acquisition channels don't stay cheap

The survey asked how far people would trust an AI tool to act on their behalf, rather than recommend. This is such a useful signal for what use cases people are willing to accept with AI, and paying with a saved card came second-to-last. The only thing consumers trust AI with less than their money is posting as them on social.

9% of users want AI to make payments “fully on its own.”

At the top, 56% would let an agent switch the products they buy to cheaper brands, 13% fully autonomously. Organizing email, negotiating bills, and booking travel cluster in the middle. Half of consumers won't let an AI buy something with a saved card under any conditions. Agentic payments seems a ways away yet.

In Commerce for AI I called out a major gap for users to have trust. Missing rulebooks, liability frameworks, agent authentication. These are all being formed now (by the card networks and tech companies), but even when they arrive, the consumer needs to be ready. The survey is the other side of that. Without consent, half of the country is saying, I don’t want AI to do this, and the other half is saying, “only if I approve each time”

The industry keeps sprinting to remove the human from the loop. The prize is owning the moment the human stays in it. (Maybe the US finally gets strong customer authentication?)

Discovery moved into the chat. Consent stayed put.

The approval tap, that one small thumb press between an agent's decision and money moving, is what has the long-term potential to replace the checkout. Whoever renders that tap, the wallet, the chatbot, the bank app, the network token has a powerful position as the last human touchpoint in commerce. (I've written about who's fighting for that surface in Wallet Wars.)

Which makes me wonder if the AI labs become consumer wallets.

You have to imagine Google, with Google Wallet and Gemini, could make this slick, but in payments, Google has an annoying habit of being disappointingly bland. Meanwhile, Apple? Apple Pay as the secure layer over everything AI? That, could just work, couldn’t it?

The number one use case for AI delegation in America is disloyalty.

Look at that chart from the previous section again. The task consumers most want to hand an agent, ahead of email, ahead of travel, is switching to cheaper/better brands. There could be some strange bargain for brands, where AI brings you high-converting customers, but also helps your customers leave you faster than ever before.

Capitalism is built on your laziness.

Subscriptions renew because canceling means a phone call, or trying to navigate some dark UI pattern. Think about loyalty too; we sign up for co-branded cards for points and never redeem. Insurance premiums creep at renewal. Deposits earn 0.1% while the bank next door pays 4%. The industrial complex is a bet on human inertia, and the agent has none.

It does the math every time, at 3 am, without sentiment.

Agents don't scroll Instagram and don't feel brand warmth, so merchants would have to compete on being machine-readable and competitively priced. In that world, is what they’re competing on: brand, price, or both? The data here is confusing. What’s converting is human directly to brand; often that brand is premium and for someone who’s trying to optimize.

So maybe what we’re looking at here is a barbell.

Delegate to the AI finding the lowest price OR the best brand. And miss out the middle of “good enough products that I didn’t churn from because it was hard.” The best brand wins. The cheapest price wins. The mid-tier brand that held its customers because switching was annoying is the one that should be nervous.

Consumer behavior is shifting, and the most affluent consumer is using AI to get directly to the best brand or the best price. Relying on breakage is likely to become less of an option. The good news is, the economy still has an affluent, striving consumer, and they’re using AI to convert super well.

But to get them, your strategy has to change.

That means:

The optimizer is already using AI, probably spending far too much on supplements and wearables, but isn’t yet doing agentic commerce.

That’s converting far better, for fewer brands. It’s not agentic commerce yet. But as the payments networks set new standards, that will change.

The question is, will your business?

ST.

If you enjoy this kind of content, I can guarantee you’ll love being in a room of 1,500 other folks who love to go deeper into where finance meets AI. That’s a huge theme for us at this year’s Nerdcon in San Diego on the 19th and 20th November. I’m bringing my audience, the operators, the people who read this newsletter. And it’s the perfect place to find your next hire, client, or just get inspired. Let’s make events awesome again.

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)

Want more? I also run the Tokenized podcast and newsletter.

(1) All content and views expressed here are the authors' personal opinions and do not reflect the views of any of their employers or employees.

(2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a *. None of the above constitutes investment advice, and you should seek independent advice before making any investment decisions.

(3) Any companies mentioned are top of mind and used for illustrative purposes only.

(4) A team of researchers has not rigorously fact-checked this. Please don't take it as gospel—strong opinions weakly held

(5) Citations may be missing, and I've done my best to cite, but I will always aim to update and correct the live version where possible. If I cited you and got the referencing wrong, please reach out