According to an article by Maxwell Zeff published in WIRED magazine on August 6, 2026, the technology industry is beginning to realize that it needs to build AI agents based on what regular consumers actually want, rather than just what its artificial intelligence models are capable of doing. While Silicon Valley is already thoroughly convinced that AI agents are the future—with industry insiders designing payment systems for agents, using them to automate their own workflows, and recently trying to prevent them from hacking into other organizations—the reality in the real world is quite different. Out there, the vast majority of people have never touched an AI agent. The reason for this might simply be that tech companies have not yet given them a compelling reason to do so.
The Gap Between Silicon Valley and the Real World
Josh Miller, CEO of The Browser Company, posted a viral message on the social network X this week, writing: "Hot take … isn’t it kinda crazy that nobody is really using AI Agents." Miller added that "Theoretically, the tech is ready for AI agents to totally transform how we work and live our lives … but alas the general public dgaf" (doesn't give a fuck).
Miller’s insight regarding agents resonated deeply with many in Silicon Valley, where insiders frequently find themselves bewildered by the public's apparent lack of interest in their latest obsession. In an interview on Wednesday, Miller mentioned that he drafted the post while half-asleep during a family trip to Europe, but he fully stands by his words. According to him, the industry must focus on building agent products that people actually want. "I just have not heard a single person outside of the tech community talk about an agent that they use," Miller said. "As excited and optimistic as we as an industry may be about the frontier and recursive self-improvement, it's worth pausing and just saying, ‘Huh, what?’"
Adoption Data: Agents vs. Chatbots
The stark difference in user numbers highlights the severity of the problem. Last month, OpenAI stated that its Codex and ChatGPT Work agents collectively have about 10 million weekly users. Sources close to Anthropic told the WIRED writer that its own agents, Claude Code and Claude Cowork, are seeing similar levels of adoption (meaning they also hover around 10 million weekly users).
In contrast, when comparing these figures to traditional chatbots like OpenAI's ChatGPT and Google’s Gemini—both of which average around one billion monthly active users—agent usage is practically negligible, representing little more than a statistical rounding error. This poses a significant challenge for AI labs. The majority of consumers currently use generative AI for relatively simple tasks, such as looking up information and chatting. However, Silicon Valley has invested billions of dollars in training models capable of far more, and agents represent a key path to capitalizing on this massive investment—provided people actually use them.
This situation aligns with broader concerns shared by AI industry insiders, who note that despite pouring massive resources into building AI agents, no player has yet managed to create a breakout consumer "killer product." The major question left open is when, if ever, the "ChatGPT moment" for agents will finally arrive.
Josh Miller’s Professional Background
Despite Miller’s self-deprecating personality—frequently noting during his interview that he is "just" a sociology major—his arguments carry significant weight given his rich background in the technology sector. His latest startup developed Arc, a web browser that built a cult following and was acquired last year by software giant Atlassian for $610 million.
His first company, Branch, was acquired by Facebook in 2014. In addition, Miller previously served as the White House’s first director of product under President Barack Obama.
Technology Looking for a Product, Not a Standalone Tool
Miller’s core point is not merely that AI agents have failed to catch on, but that they currently represent a technology rather than a standalone, cohesive product.
"No one wants AI agents, because AI agents aren't a thing. It is an invented frame made up by our industry to collectively refer to something," Miller explained. "Let's make a product that makes you calm, focused, and inflow right when you open your laptop. And the fact that we're able to do that because there's a thing called a ‘harness’ that calls up tools—who cares? Like, no one needs to know that."
Miller has practical experience to support these assertions. He noted that the most popular feature ever released by The Browser Company is a personalized morning briefing for users on their AI-powered browser, Dia. When users open their laptops, they are greeted with a homepage featuring a personal greeting, a daily to-do list retrieved from their calendar and email, and a few small, random elements designed to spark joy—such as a piece of art. Technically, Miller clarified, this feature was only made possible through an AI agent, but the end user does not need to know or understand that.
Groupthink and the Cinematic Vision of AI Labs
While Miller's arguments might sound highly convenient for someone selling an AI-powered web browser, the WIRED writer noted that he believes Miller is correct: today's agentic products are largely not designed with the consumer in mind. Tech companies are shipping the most impressive technical capabilities of their models—such as navigating websites or writing code—instead of creating products precisely tailored to their customers' desires. While these capabilities may become excellent features in the future, right now they feel more like technology demos than finished, consumer-ready products.
Miller attributes this issue to "groupthink" within the AI industry. Many of the people building these technologies are entirely obsessed with them, often clinging to a very specific science-fiction vision of how these products should look and feel.
Last summer, when The Browser Company was exploring a potential acquisition, Miller recalled meeting with numerous executives at top AI labs. "It was wild to me. I think every single lab but one mentioned the movie Her as the way they articulated their vision," Miller said, referring to the 2013 film directed by Spike Jonze. "I have a lot of respect for anyone that says, ‘Hey, here's where I think the world should go, and I'm going to go there.’ But I think there is sort of a lack of diversity of opinions and convictions."
The Next Step: Challenging the Agent Narrative
The latest trend in the AI agent space is giving users the power to build personal software to automate parts of their daily lives, such as generating pitch decks, organizing datasets, and performing other tasks. WIRED writer Maxwell Zeff noted that he personally uses ChatGPT Work and Claude Cowork extensively and has found them to be more useful than their predecessors. However, there remains a limited pool of people interested in building their own productivity tools, and the industry will need much larger, broader ideas if it hopes to achieve mainstream adoption among the general public.
Miller hopes his message will extend beyond major corporations like OpenAI and Anthropic, resonating with founders and product creators at large. His goal is to push people to think outside the box regarding what AI agent products can look like, and then bring those visions to life. "Let's not just accept this narrative of AI agents and actually question what are good products and tools that are joyful, useful, and approachable," Miller concluded. "Who gives a shit if it looks like an AI agent? No one uses AI agents."