AI could surpass human intelligence or outperform humans in most tasks by 2030, with extraordinarily capable models that do things humans cannot do emerging by the end of the decade.

Altman notes that current AI excels at narrow tasks, but true AGI would require general reasoning, and he would be surprised if the milestone isn’t reached by 2030.

The race to scale AI infrastructure continues, with OpenAI, Oracle, and SoftBank expanding data center capacity under initiatives like Stargate to support more powerful models.

OpenAI’s sizable data-center development, including the 800-acre Abilene complex built with Oracle and SoftBank, aims to underpin AI model scaling and Stargate’s broader goals.

Altman’s stance sits within a broader tech leadership debate on AI timelines and impact, alongside other prominent figures and corporate bets.

Despite optimism, Altman emphasizes that human preferences and societal needs will shape how AI develops and is used, including what people find useful and how they work with technology.

Even with powerful models, deployment requires substantial infrastructure and energy, and public release of advanced systems may be constrained by resources.

AI could handle roughly 30% to 40% of tasks in the economy in the not-too-distant future, shifting from broad job destruction to automation of individual tasks.

This shift will have tangible economic effects, prompting workforce restructuring as AI absorbs a significant share of tasks rather than replacing all jobs.

Other industry leaders envision earlier timelines for AGI, with figures like Anthropic’s Dario Amodei and Elon Musk suggesting faster progress than 2030.

OpenAI’s Stargate initiative, in collaboration with Oracle and SoftBank, is central to addressing the automation shift and scaling capabilities.