AI can make tech CEOs’ lives easier. What about everyone else?

TechCrunch's Equity podcast devoted a recent episode to a question raised by reporter Amanda Silberling: whether AI executives are too disconnected from ordinary users to build products that serve them. Hosts Kirsten Korosec and Sean O'Kane, along with Anthony Ha, spoke with Silberling about the pattern of remarks from industry leaders that land badly with the public. The conversation was published as an edited transcript, with Ha credited as the author.
The immediate prompt was Dazzle AI CEO Marissa Mayer, who Silberling described as the latest tech chief criticized over comments praising AI's usefulness. Mayer's example involved using an AI tool to help with a school assignment describing her daughter. Other cited examples: OpenAI CEO Sam Altman has floated the idea of parents making personalized podcasts for their children's drive to school and has said he could not imagine parenting without ChatGPT. Meta CEO Sam Altman — Mark Zuckerberg — has suggested that meeting social and romantic needs through AI will become increasingly acceptable. Silberling also pointed to Bumble CEO Whitney Wolfe Herd's musing about AI concierges whose AI dates another person's AI and introduces the humans only if the bots get along, and to a New York Times article in which Mayer imagined her AI coordinating with her husband's.
The broader context invoked on the episode: Anthropic is set to go public just as Wall Street begins losing its appetite for heavy AI spending, which sharpens the question of who these products are actually for.
Korosec framed the question as a choice between a marketing problem and a product problem. Silberling's answer was that it is a bit of both. The marketing failure, in her account, arises when executives at very valuable companies sit in a room trading pain points — booking flights, making dinner reservations — and mistake their own lives for everyone's. Those tasks are universal, but few people book flights weekly or find reservations a significant personal challenge. The value ordinary users get from tools like ChatGPT or Claude, she argued, is probably not flight booking.
On the product side, Silberling described an industry pushing toward a future in which basic human interaction is progressively engineered out. She offered her own generation as evidence: ordering takeout by phone now feels strange because doing it online is the default. Her stated view is that people talking to each other is a net positive for society. O'Kane's deadpan reply — asking what the return on investment is for a conversation in person — captures the tension the episode circles.
Dating came up as the sharpest illustration. Silberling conceded the inefficiency: going on many dates, wasting an evening, discovering a mismatch. One can see why someone would ask whether the AI could do the dating instead. But she argued something isn't human there, and that delegating the choice of whom to date outsources a decision to a computer that users assume knows them better than they know themselves — an assumption she doubts.
Ha offered historical framing. Uber launched targeting early adopters in San Francisco, and it was easy to conclude it would never spread beyond young urban professionals with money to spare. The tech industry has often succeeded by betting that it represents the leading edge and that consumers and families of the future will resemble its own employees. This time, he suggested, the bet has not yet paid off, partly because the society receiving the technology is different: people are voicing discomfort with its influence and resisting the idea that every experience should be optimized away. They want to parent their own children.
Silberling connected the executive talk to a countertrend: people reverting to flip phones, or buying CDs because they are tired of an algorithm choosing their music. The cultural tension, as she put it, is that no one has found the balance between optimization that helps and optimization that makes people feel less human.
This is a useful read for product teams, not just a media critique. The examples are mostly consumer-facing: parenting, dating, social connection. Executives are describing a future in which AI mediates intimate and mundane moments alike, and that framing may alienate the users such products need. A likely trade-off, in my reading, is that the strongest demos — a bot that books your flight, an AI that screens your dates — showcase convenience while implicitly telling users their own judgment is the bottleneck.
There is a testable version of the question. If the discomfort were purely about wording, better messaging would close the gap. If it is product-driven, usage patterns should show where AI actually earns its place and where it reads as intrusion. Korosec's question to the team is, in effect, a research brief: identify the pain points real users have rather than the ones executives notice in their own calendars.
The IPO and spending context raises the stakes. Public-market investors will want evidence of durable consumer demand, and the CEO quotes that play well in a conference room can travel badly to a mainstream audience. The episode does not resolve which side of the marketing-versus-product split matters more; Silberling explicitly says both.
Why it matters: For anyone building or marketing consumer AI, the episode is a reminder that the examples executives reach for can signal who a product is for. Developers and product teams may find that trust is shaped less by model capability than by whether the use case respects users' desire to do some things themselves — an inference from the discussion rather than a reported finding. As AI companies approach public markets, the gap between insider enthusiasm and mainstream skepticism is likely to be a commercial risk as much as a reputational one.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
On Equity, we discuss why tech CEO statements can seem so tone deaf.