Maybe I misunderstood the term, but populism seems to follow an awfully simple premise: people get angry when they think someone else is benefiting off of their adversity. It’s a term that seems to be widely applied to different situations, each time with different definitions. But it’s different than the misnomers we run into for other political terms -- this one feels like it is never addressed. Because the others certainly are; “that’s not what communism actually is,” “no, China isn’t socialist, it’s a market economy,” blah blah blah. Nevertheless, for all the revolutions, manifestos, theories, rebuttals, elections, coups, the underlying premise follows -- one group is systematically being preferred (or, at least that’s what it seems like to the masses), and another group is benefiting off of it. Until… a bright charismatic leader shows up with a star-studded plan (for better or for worse). A leader truly ready to challenge the status quo.

I wrote a piece a while back about how the NYC subway was beginning to look like a walk down Silicon Valley. Car after car advertising complex B2B enterprise AI software applications that had no relation to 99% of riders (unless, of course, the product works to automate their work…). The ad is often accompanied by a sad attempt at a joke and a component that is meant to breed affinity with the city; “the enterprise AI that works for the Big Apple.” I made the point that most Americans will start to feel a cumulative effect of all these mentions of AI, a death by way of a million paper cuts. And this can only feel like a felicitous entrance for populism and a leader (or whatever you’d like to call it).

And it has already become, improbably, one of the most bipartisan issues in America since beer. 75 data center projects have been blocked or delayed in the first quarter of this year, costing big tech roughly $130 billion, practically matching the value of projects delayed/blocked in all of 2025. The number of active opposition groups more than doubled, from 396 to 833, spread across 49 states. Gallup found that 71% of voters do not want a data center built near them, including 75% of Democrats and 63% of Republicans. As we all should know by now it’s rare when we come across bipartisanship, and to be clear, the real (by real, I mean not the specious disagreements we see online) disagreements usually come in discussing the correct means to an end, not exclusively the end. In fact, usually the end is the exact same. But the means to inhibiting the proliferation of AI have, for now, been equal in that they have not been fully fleshed out.

Data center protests go national as backlash grows | KSL.com

Most Americans, of course, will never actually live next to one. There are a few thousand of these buildings in a country of 350 million, and odds are you will go your entire life without seeing one up close. But, as my previous article argued, the anger arrives through the hundred news articles, the thousand subway ads, and the million mentions of a technology that, at the very least, could replace them. The counterargument is often “but they won’t replace them,” but that isn’t the premise I am arguing. They may very well never replace them, but you cannot argue it is an outlandish position to take, meaning the fear is justified (or maybe, tell Altman and Amodei to stop saying AI will take every job to appease the market). Low grade and cumulative, the sort of thing nobody would ever organize around and everybody subconsciously resents. A Lingchi resentment (凌遲, “slow slicing” or “death by a thousand cuts,” was a form of torture and execution used in China from around AD 900 until it was officially banned in 1905).

But even so, where are our leaders? If the resentment is there, why hasn’t anyone figured out how to put it on a poster behind a podium? Perhaps the answer lies where those cheeky incentives are. Those cheeky economic incentives.

The Cheeky Economic Incentives

AI companies spent more than a billion dollars on digital ads in the US across all of 2025, up 126% from the year before. According to Sensor Tower’s State of AI 2026 report, they spent $1.3 billion in the first five months of this year alone, a 48% increase yoy, against 167 billion impressions. Last year's number was beaten before summer. In the first quarter of 2026, OpenAI’s US ad spend was up roughly 800% yoy. Anthropic’s, on the other hand, was up 1,184%.

A friendly reminder for you all: both of those companies are burning cash… ChatGPT’s share of the assistant market fell below 50% for the first time in March, and by May it sat at 46% against Gemini’s 28% and Claude’s 10%. This seemingly outrageous advertising surge is a free for all among the four or five companies who realized they have no moat and their users can be replaced in a matter of minutes. The reason you cannot get away from it on the subway, in your feed, or during a podcast about the Knicks is because you are standing smack in the middle of their debt fueled plastic knife fight.

You’ll be pleased to hear that the single largest retail advertiser of AI-related messaging between May 2025 and April 2026 was Best Buy. Best Buy, the data center B2B SaaS artificial intelligence company? Nope, it’s that Best Buy, the same old one that sells TVs and laptops -- they didn’t pull an Allbirds. App downloads in the US driven by searches containing “AI” are also booming, growing 651% in Health and Wellness and 536% in Financial Services between the first quarter of 2024 and the first quarter of 2026.

I also looked into one of (if not the) fastest growing spheres of influence in media right now -- podcasting (and arguably more importantly, its clips), since if any leader were to emerge, I would posit them to be a podcaster.

There are more ads inside a podcast than there ever used to be. The top 3,000 American shows’ ad load reached 8.75% of total episode time in the second quarter of this year, a record, up from 8.11% a year earlier (Magellan AI). Over that same period, podcast ad spending was up 23%, and total US podcast ad revenue will clear $3 billion in 2026.

Some of that money is obviously insurance and mattresses and therapy apps. But in the first quarter, the two largest categories in all of podcast advertising were Financial Services in first place and Business Services and Software in second, at $79.4 million. Arguably, a majority of both of these buckets are either AI or AI-related products and companies.

Magellan AI counted 1,297 brands advertising on podcasts for the first time in a single quarter, spending an average of $31,400 each, mostly on mid-roll placements running about 51 seconds. Host-read ads account for roughly 55% of podcast advertising revenue, and they command premium pricing. Why? Because host readings lend trust to the product. So the position an anti-AI host puts himself in is fairly easy to work out. AI companies are buying up a growing share of the podcast sponsorship market, and turning that money down means turning down not only the most reliable, but the fastest growing advertiser (ever) available to a show of that size. Hold the growth rate steady for a few years and the refusal is no longer a matter of passing on one or two or twenty advertisers, it is a matter of sitting out 99% of the market, where nearly all the money has gone.

That was a mouthful of data, but hopefully you’re starting to get the gist of it: there is an economic incentive that holds an iron grip around most creators and influencers. Nearly every voice you trust is currently drawing cash money from the subject it would otherwise be raising an eyebrow (or, more of an eyebrow) at. And I know some of you are thinking that your guy/gal is different, that they are honest, that they’d never. Unless you mean J&E over at Passing The Torch, you’re probably wrong. I doubt OpenAI is sending your favorite creator a note asking them to hush. It wouldn’t need to, and yes, I hear how that sounds coming from me (; . But money tends to do its work much earlier in the process, back at the stage where someone is sorting through the twenty things they could write about this week and deciding which two they’ll end up talking about. What happens to criticism that never gets written down? Well there’s nothing to suppress if nothing is there…

See the Seesaw coming?

Now to tie all of this back to Passing The Torch’s paramount thesis: the seesaw.

As technology becomes more capable and more inescapable, people turn economically rebellious, and that rebellion arrives as money spent on the physical, the analog, the friction, and the handmade. A key aspect of the thesis is that it runs on the technology broadly succeeding. Consequently, ads serve a double mandate in this scenario, being both the instrument of success and the breeder of consumer resentment.

So it may be that the absence of an AI Che is what allows the seesaw to work at all… Populism by definition needs a voice to attach grievances to, because after all you cannot vote out a problem that is simply not on the ballot (unless you, well, overthrow the government. But we’ve got no time to do that, the next season of Love Island comes out tonight!). Be that as it may, economic rebellion requires no author, no leader, no manifesto, no centralization at all. It is a trend (a general direction in which something is changing, developing, or moving over time), not a directive campaign. As the theory suggests, this drift will be almost subconscious and consumer driven, and no revolution will be necessary.

Che Wearing Che T-Shirt – Shirtstore

Though on reflection, there is nothing stopping us from being the AI Che Guevaras ourselves.

Stay curious, folks!

—J&E