He mentioned achieving AGI before mentioning that he’s becoming a parent soon.
New Life
Many people’s New Year resolutions involve going to the gym or budgeting better — but for OpenAI CEO Sam Altman, the new year means ushering in the singularity.
During a talk with Gary Tan, the president and CEO of the Y Combinator startup incubator that Altman once led and was subsequently fired from in 2019, Altman was almost flippant when suggesting that artificial general intelligence (AGI) would be achieved within the next calendar year.
“What are you excited about in 2025?” Tan asked his predecessor. “What’s to come?”
“AGI,” Altman immediately responded. “I’m excited for that.”
He then went on to drop the bombshell that he and his husband, Oliver Mulherin, are having a child.
“I’m more excited for that than I’ve probably ever been,” Altman said as Tan congratulated him. “That’s the thing I’m most excited for, ever, in life.”
Be that as it may, choosing to drop the baby news after jumping to say how stoked he is for AGI does suggest some unusual priorities.
Resolved
Notably, Altman’s proclamation about AGI’s imminent arrival in 2025 occurred just days after telling “20VC” podcast host Harry Stebbings that within five years, the technology will have come and gone.
“I think in 5 years, it looks like we have an unbelievably rapid rate of improvement in technology itself,” the OpenAI CEO said. “People are like, man, the AGI moment came and went.”
Altman went on to say that he thinks within that time frame, “society itself” will have changed “surprisingly little” — a statement he’s made before, even as other OpenAI insiders grow ever more concerned about AGI changing things for the worse.
Indeed, just a few weeks ago, the firm’s outgoing AGI readiness czar Miles Brundage wrote on his personal Substack that he’d quit OpenAI because he doesn’t think the company is prepared to handle the gravity of what it’s building. After his departure, OpenAI dissolved its AGI readiness team, which seems to illustrate the point perfectly.
With these back-to-back Altman statements normalizing an imminent AGI future, two questions arise: is human-level AI actually that close to being achieved, and if so, do we really want OpenAI to be in charge of it?
More on AGI: AI Bros Terrified Singularity Will Hit While Trump Is President
