OpenAI recently launched an AI-powered web browser named ChatGPT Atlas, prompting discussions among tech commentators about its potential to disrupt the existing browser landscape and whether users should consider switching from established platforms like Safari. This development was a central topic on a recent episode of the Equity podcast, where Max Zeff, Sean O’Kane, and Anthony delved into the implications of AI browsers and their viability.

Despite OpenAI's massive funding, a significant hurdle for new browsers has historically been the difficulty in generating revenue from the browser product alone. Sean O’Kane highlighted that many companies have attempted to unseat major browsers but ultimately failed due to this financial unsustainability. However, OpenAI, with its substantial financial backing, is in a unique position where it doesn't immediately need the browser to be profitable, allowing it to evolve over time.

Max Zeff, who has personally tested ChatGPT Atlas and other AI-agent-powered browsers like Comet, expressed skepticism about their current utility. He noted that at best, these browsers offer only a "slight efficiency gain." More often, he observed the AI agent "clicking around on a website," performing tasks that he doubted normal users would genuinely seek out, such as having an agent add grocery ingredients to Instacart. This raises questions about the actual value proposition for consumers and whether the "agentic web" vision truly resonates with everyday browsing needs, presenting a disconnect between the tech industry's offerings and user demand.

Beyond efficiency, significant security risks associated with AI browsers were also a point of concern. Anthony also pondered the broader implications for the "open web," suggesting that if AI interfaces and chatbots become dominant, the importance of traditional web pages might diminish as more browsing is mediated by these new tools. He also mentioned his personal experimentation with non-Google search engines due to the prevalence of generative AI content in search results.

Sean, who considers himself an "old head" when it comes to search and browsing, primarily uses traditional methods like Boolean searches for his work, which often involves navigating discrete parts of web pages. While he has not yet tried AI browsers, he acknowledges the ongoing evolution of search, feeling that Google's traditional Boolean search capabilities might eventually be phased out. The current consensus among the podcasters suggests that while the concept of AI browsers is fascinating and backed by substantial investment, their compelling use cases and widespread adoption by average users are not yet clearly established.