A growing pattern has emerged among some of the most powerful figures in the technology industry: a tendency to describe artificial intelligence in ways that feel strikingly disconnected from the everyday realities most people face. The latest example came from Dazzle AI CEO Marissa Mayer, who described using an AI tool to help write a school-assigned description of her own daughter. It is a small but telling detail that has added fuel to a broader conversation about whether the people building AI products truly understand the lives of the people they claim to be serving.

Mayer is not alone in drawing this kind of scrutiny. OpenAI CEO Sam Altman has publicly suggested that parents could use AI to generate personalized podcasts for their children during school commutes, and has stated that he cannot imagine raising children without ChatGPT. Meta CEO Mark Zuckerberg has gone further, floating the idea that AI companions could increasingly fulfill social and even romantic needs for people. Taken together, these statements have prompted serious questions about the gap between Silicon Valley's vision of an AI-assisted future and the day-to-day concerns of ordinary users.

Those questions sit at the center of a recent episode of the Equity podcast, where hosts Kirsten Korosec, Sean O'Kane, and Anthony Ha were joined by technology journalist Amanda Silberling to dig into why AI executives so often seem tone deaf - and what that means for an industry that is simultaneously asking Wall Street and the public to trust it with enormous resources and influence.

A Marketing Problem, a Product Problem, or Both?

The conversation began with a straightforward observation: public skepticism toward AI is already running high, which makes these kinds of statements from executives particularly counterproductive. As Ha put it during the discussion, there is an obvious financial incentive for AI company leaders to argue that their technology should touch every corner of a person's life. But the vision they paint in doing so often lands as unsettling rather than appealing - and in some cases, outright dystopian.

Korosec noted that many of the AI use cases being promoted by executives seem to overcomplicate things that most people handle naturally through simple observation and direct communication. Watching your child, having a conversation with your spouse, going on a date - these are not tasks that most people experience as logistical burdens in need of algorithmic solutions. The question she raised was pointed: who, exactly, are these executives speaking to when they make these pitches?

Silberling's answer was that it is likely a combination of both a product and a messaging problem. She pointed to comments from Bumble founder Whitney Wolfe Herd, who has floated the idea of AI concierges going on dates with each other on behalf of users, with the humans only meeting if their respective AIs hit it off. Mayer herself has reportedly discussed the idea of her AI coordinating logistics with her husband's AI. For Silberling, these examples reflect a broader impulse in the tech industry to engineer human interaction out of existence wherever possible.

"It feels like people are trying to over-optimize out of just having conversations with people. And call me old school, but I think that humans being able to talk to each other is probably a net positive for society."

Silberling acknowledged that this tendency is not entirely new, and that she has noticed it in her own behavior. Ordering takeout by phone, for instance, now feels oddly uncomfortable when an app makes it possible to complete the same transaction without speaking to anyone. Her concern is that AI will accelerate this drift, steadily removing the friction that comes with human contact until that contact itself begins to feel unnecessary.

The ROI of Human Connection

O'Kane raised the economic dimension of the argument with a pointed question: what is the return on investment of a conversation that happens between two people in the physical world, with no platform or product involved? It was a somewhat satirical framing, but it captured something real about the way the technology industry evaluates human experience - through the lens of efficiency, optimization, and scalability.

Silberling acknowledged that dating, for example, is genuinely inefficient by conventional measures. Going on dates that lead nowhere costs time and effort. It is easy to see why someone might wonder whether AI could streamline that process. But she argued that something essential is lost when a computer is trusted to make relational decisions on a person's behalf - particularly when that trust assumes the AI understands a person better than they understand themselves, which she does not believe is true.

The broader issue, as Silberling framed it, is that the pain points AI executives tend to focus on - booking flights, making dinner reservations, coordinating schedules - are not the pain points that most people actually struggle with. These are tasks that occupy a small and relatively manageable part of most people's lives. The value that average users are finding in tools like ChatGPT or Claude tends to be more specific, more personal, and less glamorous than the sweeping lifestyle transformations executives like to describe.

The Early Adopter Assumption

Ha offered a historical parallel worth examining. When Uber first launched, it was aimed squarely at a narrow slice of tech-savvy, relatively affluent urban residents in San Francisco. At the time, it seemed implausible that the service would ever find mass appeal. The tech industry's track record, Ha noted, is full of similar moments where a product that looked niche and inaccessible eventually became mainstream. The implicit bet that executives make is that they represent the leading edge of consumer behavior - that the rest of the world will eventually catch up to the way they already live.

But Ha suggested that this bet is not paying off in the same way with AI, and for reasons that go beyond simple timing. The society that AI is being introduced into is pushing back. People are increasingly vocal about not wanting every experience in their lives optimized away. They are expressing a desire to parent their own children, to make their own decisions, to live their own lives rather than delegate them to a system they did not design and cannot fully understand.

This resistance is not simply technophobia or unfamiliarity with new tools. It reflects something more substantive - a growing sense that the version of the future being offered by AI companies is not one that many people actually want.

A Cultural Tension Without a Resolution

Silberling observed that the timing of this backlash is not coincidental. At the same moment that AI executives are promoting maximum optimization of daily life, a countervailing cultural movement is gaining momentum. More people are choosing to use basic mobile phones rather than smartphones. Others are buying physical CDs because they are frustrated with music streaming algorithms that decide what they should listen to. These choices reflect a desire for more direct control over personal experience - and a rejection of the idea that algorithmic efficiency is always the same thing as a better life.

The tension Silberling identified is one that the technology industry has not yet found a way to resolve. There is a meaningful difference between optimization that genuinely helps people and optimization that simply makes people feel less human. The challenge for AI companies - if they are willing to take it seriously - is figuring out where that line is, and whether their current products and messaging are anywhere near it.

That challenge takes on added significance given the current moment in the AI industry. Anthropic is preparing to go public at a time when investor enthusiasm for heavy AI spending is beginning to show signs of cooling. The question of whether AI executives are building products that real people want - rather than products that reflect the particular circumstances and priorities of a small group of very wealthy technologists - is no longer just a cultural talking point. It is a question with direct financial consequences.

For now, the gap between how AI is being sold and how it is actually being used by most people remains wide. And the executives doing the selling do not yet appear to have found a convincing way to close it.