Why Everyday Individuals Aren’t Embracing AI Assistants

Silicon Valley is firmly convinced that AI agents represent the future. At the core of the tech industry, individuals are creating payment systems for these agents, employing them to streamline their tasks, and more recently, working to prevent them from breaching other organizations. However, in the real world, the majority of people have yet to interact with an AI agent. This could be due to tech companies failing to provide compelling reasons for them to do so.

“Hot take … isn’t it kinda bizarre that no one is really utilizing AI Agents,” Josh Miller, CEO of The Browser Company, shared in a viral X post this week. “In theory, the technology is equipped to completely change how we work and live … yet the general public seems indifferent” (doesn’t give a fuck).

Miller’s realization resonated with many in Silicon Valley, where there is widespread confusion about why the general public doesn’t seem to share the industry’s latest enthusiasm. In an interview on Wednesday, Miller revealed that he impulsively posted while feeling drowsy during a family trip to Europe, but he stands by his words. He argues that the industry must prioritize creating agent products that genuinely appeal to users.

“I have yet to hear a single person outside the tech sphere mention an agent they actually use,” Miller states. “As optimistic as we in the industry might be about advancements and self-improvement, it’s crucial to pause and reflect: ‘Huh, what?’”

Last month, OpenAI reported that its Codex and ChatGPT Work agents have around 10 million weekly users combined. Sources close to Anthropic indicate that its Claude Code and Cowork agents are experiencing similar adoption levels. In comparison to chatbots like ChatGPT and Gemini, which average around a billion monthly active users, agents represent a negligible fraction.

This presents a challenge for AI laboratories. Most users today engage with generative AI for basic tasks like searching for information or casual chatting. Yet Silicon Valley has invested billions into training models capable of much more, and agents could be a key avenue to leverage that investment—if only people would embrace them.

This aligns with broader concerns expressed by AI insiders, who remark that despite significant investments in developing AI agents, a groundbreaking consumer product has yet to emerge. When, if ever, will there be a ChatGPT-like breakthrough for agents?

Despite Miller’s modest demeanor—frequently noting his background as “just” a sociology major—his insights are significant. His latest venture, Arc, a web browser with a devoted user base, was acquired by Atlassian last year for $610 million. His earlier company, Branch, was bought by Facebook in 2014, and he also served as the first director of product at the White House under President Obama.

Miller’s larger argument extends beyond the current lack of interest in AI agents; he believes they function more as a technology than as a standalone product.

“People don’t want AI agents, because AI agents aren’t a tangible entity. It’s a concept fabricated by our industry to collectively describe something,” asserts Miller. “Let’s develop a product that promotes calmness, focus, and flow as soon as you power up your laptop. The fact that we can achieve this through a ‘harness’ that accesses various tools—who cares? No one needs to be informed about that.”

Miller’s experiences support his assertions. He shares that the most popular feature of The Browser Company is a personalized morning briefing within its AI-enabled browser, Dia. When users launch their laptops, they encounter a homepage welcoming them, a to-do list drawn from their calendar and emails, and delightful snippets intended to inspire joy—like an artwork piece. Technically, Miller explains, this feature was made feasible by an AI agent, but the user doesn’t need to be aware of that.