‘Population collapse’ makes AI productivity gains vital, says top Meta executive

One of Meta’s top executives has said that impending “population collapse” means AI-driven productivity gains are “desperately” needed to support society, as tech leaders seek to persuade a sceptical public that the powerful technology will benefit their lives.

Alex Schultz, Facebook parent company’s chief data officer, told the FT’s Future of AI summit on Wednesday that falling birth rates meant the world would have “a lot fewer people 50 years from now”.

“The world as a whole isn’t thinking hard enough about things like population collapse . . . We desperately need more productivity,” he said. “How are we going to care for elderly people? How are we going to put ourselves in a place where we can actually make the most of the number of people on Earth?”

The global population is expected to peak as early as 2064, while the working population is set to shrink more quickly due to longer lifespans and falling birth rates.

Schultz’s comments come as Meta races to compete with frontrunners OpenAI, Anthropic and Google, having spent billions of dollars on its AI ambitions.

Chief executive Mark Zuckerberg told the FT last month that its new model, Muse Wonder, had caught up with rivals to reach the “frontier”, while its consumer-focused Muse agent has been downloaded millions of times since its launch in September.

These developments have come amid broader concerns about the sustainability of the huge spending spree behind the AI boom, with Meta announcing plans to raise an additional $95bn to fuel its spending last week.

Meanwhile, anger around the building of AI data centres in local communities across the US has emerged as one of the reasons voters rejected incumbent Republican candidates in the US midterm elections on Tuesday.

Extolling the productivity benefits of AI, however, Schultz said the fast-developing technology was already “accelerating” software engineers’ jobs and saving “a tonne of time” for lawyers who were using tools to predict questions in court.

Within Meta itself, AI was removing “all the awful tasks for software engineers” on monitoring alerts in its data warehouse, where more than 10 per cent of tasks were now being resolved by agents without the need for human input, he said.

Nonetheless, he acknowledged that not all AI use had led to increased productivity, with some workers “burning tokens with no obvious value”. Within his own team, “less than half” were using the technology effectively, he added.

Productivity gains were improving with the accessibility of more recent models, Schultz said, adding that he was “pretty slow on the uptake” until OpenClaw popularised so-called vibe coding at the start of the year.

However, he pushed back against fears about the technology leading to widespread unemployment, saying that “having tools doesn’t eliminate jobs, it just changes them”. In addition, the technology could create new jobs due to increased productivity, he added.