Will it Play in Peoria?
AI has a consumer desire problem and it starts with who it's built for: too much Palo Alto, not enough Peoria.
Forget the chicken in every pot — imagine an EV in every driveway and a robot in every kitchen.
Picture it: Peoria. 2036.
The 6 AM alarm wakes mom to the smell of chicken-apple sausage, pre-programmed into the kitchen robot for the week’s meals. Dad rolls the kids out of bed. The AI assistant pushed their homework to the school portal overnight, and the teacher signed off before breakfast. Both kids have soccer practice after school, and their uniforms were washed overnight. The laundry detergent is set to auto-replenish and trigger a new delivery by 3 PM.
Before anyone heads out the door, the family scans the daily cognitive report on the grandparents. They set up an in-home monitoring system two years ago so they could stay at home instead of moving into a senior living facility (Grandma is trending fine, but Grandpa is experiencing a little decline).
Mom heads to her hospital shift, where she’ll see patients all day and sit through onboarding for the agentic intake platform the administrator finally greenlit. Dad heads his job at the manufacturing plant, running a floor mostly staffed by autonomous machines.
Or do they?
Will it play in Peoria? That’s the old showbiz question that early radio and vaudeville producers would ask themselves to gauge whether a show would resonate with Middle America or only be a hit with urban aesthetes and coastal elites.
It’s a question we all have to start asking ourselves as a gut-check on AI’s ability to connect with the average consumer and deliver an experience that consumers deeply desire.
Many AI founders and investors are still building for the power users and tinkerers who just bought their fifth Mac mini.
Sometimes, we build a little too much for Palo Alto and not enough for Peoria.
Today’s agentic tools are as kludgy as they will ever be. They will get more user-friendly, more intuitive, and more obvious. And yes, we had to build the models and infrastructure before we could build for the masses. But their time is next and now. The night-shift nurse and the grandmother and the high school sophomore are all going to need to use AI in ways no one is building for yet.
But we risk making people hate AI before they have the chance to love it.
The backlash against AI is growing louder because it is being introduced with a bulldozer — often a literal bulldozer, as we build data centers in small towns — while failing to ensure that John and Jane Q. Public are the beneficiaries of its rapid deployment.
Does the single mother managing two jobs, three kids’ after-school activities, and the monthly planning that includes her child support and a complex co-parenting schedule feel that AI significantly improves her life today?
Probably not, and in many cases, she harbors a deep distrust of AI and worries about whether the company she works for will keep her around as they adopt more of it.
It is much easier to be anti-the current thing if the builders haven’t provided a use case that sparks joy or creates a significant improvement in your daily life.
The first wave and then the next.
It feels like AI is everywhere, but we’re still extremely early in the adoption curve for most consumers.
Work was the first frontier. Today, 69% of employees in leadership positions use AI at work. But as you move down the hierarchical ladder to managers and individual contributors, that utilization declines. The utilization gap for AI at work is that adoption is concentrated at the top tier of employees and even further concentrated in knowledge-based industries. AI has not as deeply penetrated service-based or deskless work as it has in white-collar jobs.
Employees who use AI at work do so because they feel they have to keep up and avoid being displaced by the very technology they’re being encouraged to use. For the working population, many AI tools are a necessary utility, so they do indeed play in Peoria — but begrudgingly.
To win over real people, AI has to create utility and delight outside of work, and the most obvious first frontier for this is commerce. 36% of people responded positively to using AI to automatically apply discounts at checkout, and a company like Phia has found success (to the tune of a $35.5M Series A) using AI to help shoppers find the best price across retail and reseller sites.
At least at first, a consumer’s willingness to adopt AI seems to be a combination of commerce, cognitive offload (how reliably can the user hand off something they’ve previously had to think about or track?), and the ability to give away the more mundane parts of their to-do list.
Consumer desire is our destiny.
The friction between consumer desire and distaste toward AI is where we should be focused next. Investors have committed over $500 billion to private AI companies worldwide, and the negative consumer sentiment about AI threatens to put all of these investments at risk. Just as importantly, we risk alienating the people and industries who could drive the next innovation wave of AI.
Pointing technology at real-world consumer problems and real-world industries is the path to making AI universally useful and perhaps even appreciated. AI has to be useful to the family whose schedules are an utterly chaotic stitching of school pickups and games and laundry. It has to be useful to the people who live paycheck to paycheck. It has to be trusted by the people who can least afford to lose their jobs. To make AI both useful and trusted is the path to making it beloved.
‘Will it play in Peoria?’ is the question that should launch a thousand code ships. It absolutely has to.
“‘Will it play in Peoria?’ is the question that should launch a thousand code ships. It absolutely has to.”
I just published a piece on this today re fashion tech. Its too true- the pendulum has swung too far silicon valley and soho