Pichai: Google "Losing" AI Coding Race to Rivals

Google CEO Sundar Pichai states his company is behind rivals in agentic coding. This is one AI capability that has become very important in the industry. Pichai mentioned this on the Hard Fork podcast. He said Google's models are strong in areas like text, multimodality, voice, and reasoning. However, they are not as strong in long-horizon coding tasks, where companies like Anthropic and OpenAI excel. What's interesting is that Pichai believes this isn't due to model quality. He suggests Google lacked the platform where developers work, which meant less data flow. He did express optimism that Google will improve in this area. Agentic coding is where significant money is being made. Anthropic's revenue surged with Claude Code, and OpenAI shifted focus to enterprise with Codex. The AI code tools market is expected to grow from $9.3 billion this year to about $30 billion by 2031. Pichai rejects the idea that Google's broad focus is a problem, stating they can handle multiple things due to their scale. He described the rapid pace of the field, saying "30 to 60 days looks like five years." Google's answer to this challenge arrived at I/O 2026 with Antigravity 2.0 and Gemini 3.5 Flash. Antigravity.