Anthropic’s engineering leader, Fiona Fung, says Claude Code allows her team to ship way more code.<\/p>\n
She said it has also made her employees, who now work mostly with AI agents, lonelier.<\/p>\n
Fung said the antidote is to create more opportunities for engineers to collaborate.<\/p>\n
Coding is famously a solitary job. AI agents have only made it worse.<\/p>\n
Fiona Fung, an engineering leader at Anthropic who runs the teams behind Claude Code<\/a> and Claude Cowork, said her team noticed that as engineers relied more on AI agents, they became more isolated.<\/p>\n “The thing that we found interesting on the Claude Code team is, after a while, we felt it could start being a lonely experience because we all started just working with our agents so much,” Fung said on the latest episode of “Lenny’s Podcast.”<\/p>\n To address the problem, Fung’s team tried to restore some of what agentic coding had begun to erase. They began organizing programming lunches, hackathons, and blocks of shared “maker time” so engineers could work near one another and learn from each other’s AI workflows.<\/p>\n “Everybody uses Claude Cowork. Everybody uses a flow so differently,” Fung said. “When we do pairwise programming, we actually learn so much from each other.”<\/p>\n Claude Code has quickly become one of the most used products in software development. In a survey conducted by Business Insider<\/a> of more than two dozen founders and venture capitalists, Claude Code emerged as the dominant AI coding tool insidestartups. Some founderssay it has become their default for complex engineering work.<\/p>\n Engineers are now spending more time directing agents, reviewing outputs, and orchestrating parallel tasks. The rise of vibecoding, where people use natural language to prompt and build software<\/a>, has also allowed nontechnical founders to create custom tools without hiring traditional engineering teams, giving rise to the “solopreneur.”<\/p>\n That can get lonely, however, and many founders swear by collaboration.<\/p>\n Fung said that is why her team has focused on creating more opportunities for engineers to work side by side, even though they use AI differently.<\/p>\n “Every time I watch someone work,” she said, “I learn something myself as well.”<\/p>\n