Former Google CEO Eric Schmidt expressed shock at artificial intelligence’s rapid advancement in software development, declaring his own obsolescence after witnessing an AI system generate an entire program autonomously—a moment that crystallized both the promise and peril of the technology he has championed for decades.“Holy crap. The end of me,” Schmidt said during a forum at Harvard University earlier this month. “I’ve been doing programming for 55 years. To see something start and end in front of your own life is really profound.” The admission from one of Silicon Valley’s most influential figures underscores a broader shift already underway: AI is moving from a helpful tool to a potential replacement for skilled human labor. At leading AI research labs, between 10 and 20 percent of programming work is already being done by AI systems, Schmidt revealed, with that proportion expected to grow exponentially in coming years.
AI’s bigger impact lies in automating business operations, not just code
Despite the dramatic demonstration, Schmidt argued that AI remains underhyped rather than overhyped, as the technology’s true economic impact lies beyond coding. Speaking with Professor Graham Allison at the John F. Kennedy Jr. Forum at Harvard University, he pointed to the automation of corporate work as AI’s biggest untapped potential.
The real transformation is happening inside companies, where AI systems are taking over billing, accounting, product design, delivery, and inventory management, Schmidt explained. These routine processes quietly consume billions in corporate spending, and automating them could fundamentally reshape business operations.
Artificial general intelligence could arrive by 2029, warns tech leader
Schmidt predicted artificial general intelligence—systems matching the smartest mathematicians, physicists, and artists—will arrive within three to five years. This timeline is driven by what he calls “recursive self-improvement,” where AI learns independently without human instruction.“The computers are now doing self-improvement. They’re learning how to plan, and they don’t have to listen to us anymore,” he warned at another recent event. Schmidt noted that AI already writes 10 to 20 percent of code at leading research labs like OpenAI and Anthropic.Looking further ahead, Schmidt forecast that within four years, AI will achieve autonomous learning capabilities. “Many people believe that there will be AI mathematicians in the next year,” he said, adding that the ability for computers to discover new facts and generate medical breakthroughs appears imminent.However, Schmidt emphasized the need for human oversight as AI approaches these capabilities. “Somebody’s going to have to raise their hand and say, ‘We just went too far,'” he cautioned. “I think there’s no higher duty than to preserve human agency and human freedom.”The former Google executive suggested Wall Street is underestimating the magnitude of AI’s impact on business automation and scientific discovery.
