Greg Brockman, president of OpenAI, announced at a Sequoia Capital event that artificial intelligence has rapidly evolved from a supportive tool to the primary driver of software development. Speaking on Thursday, Brockman highlighted a dramatic shift in usage patterns occurring over a single month, noting that agentic coding tools moved from generating 20% of code to 80% within weeks. He emphasized that AI has transitioned from a novelty to the main focus of engineering workflows, prompting him to advise company founders to fully embrace these technologies. Brockman pointed to the evolution of Codex, OpenAI's code-generation platform, which has expanded from a specialized utility for software engineers to a system accessible to anyone working with computers. Despite this rapid expansion, OpenAI maintains a strict policy requiring human oversight. Brockman clarified that while AI handles the bulk of the writing, humans remain responsible for merging code into production. He warned against two extremes: blindly relying on AI without review or rejecting it entirely. Instead, he advocated for a balanced approach that leverages AI's speed while ensuring human accountability for quality and safety. Brockman's comments reflect a broader industry trend where major technology firms are witnessing AI dominate code creation. Google CEO Sundar Pichai recently revealed that 75% of new code written within Google is generated by AI and subsequently reviewed by human engineers. This figure represents a steep increase, rising from 25% in early 2024 to 50% last year. Similarly, Meta plans to integrate AI more deeply into its operations, with reports indicating that over 65% of engineers in its creation organization are expected to generate more than 75% of their committed code using AI assistance. The trajectory predicted by industry leaders suggests an even more aggressive adoption timeline. Dario Amodei, CEO of Anthropic, stated at a recent conference that AI will likely be responsible for 90% of code generation within three to six months. He further speculated that within a year, the industry could reach a state where AI writes essentially all code. Amodei has noted that at Anthropic, the integration of AI into coding is already substantially accelerating the pace of progress in building next-generation AI systems. These developments signify a fundamental change in how software is built. The role of the programmer is shifting from writing individual lines of syntax to overseeing, refining, and architecting systems driven by intelligent agents. As tools become more capable, the barrier to entry for software creation lowers, allowing a wider range of workers to participate in development. However, the consensus among these leaders is that human guidance remains essential. The goal is not to replace engineers but to augment their capabilities, enabling them to focus on higher-level problem-solving while AI manages the repetitive and complex tasks of code generation. The industry is currently in a transitional phase, moving quickly from experimental adoption to standard practice, with the expectation that human oversight will continue to be the critical safeguard as AI takes the lead in coding.