Eric Schmidt plans to give $125 million to AI researchers

Eric Schmidt, former CEO of Google.

John Lamparsky | Getty Images

Eric Schmidt, the former CEO of Google, is plans to invest $125 million in artificial intelligence research through his philanthropic venture Schmidt Futures.

The funding, which will be administered through an initiative dubbed AI2050, will aim to support scholars working on “hard problems” in AI.

“AI will cause us to rethink what it means to be human,” Schmidt said in a statement Wednesday. “As we chart the course towards a future with AI, we must prepare for the unintended consequences that may follow.”

Schmidt, who has previously warned of the dangers of AI, pointed out that in the early days of the internet, no one expected social media platforms to disrupt elections and influence our lives, opinions and our actions to the extent that they do now.

“Lessons like these make it even more urgent to be prepared to move forward,” he said.

“Artificial intelligence can be a massive force for good in society, but now is the time to ensure that the AI ​​we build has human interests at heart.”

The AI2050 initiative will be co-chaired by Schmidt and James Manyika, Google’s new chief technology and society officer.

Payments are expected to be made to individual scholars over the next five years.

Berkeley scholar Stuart Russell and Rediet Abebe are among the first to be named fellows. Abebe studies how AI can help measure and mitigate socio-economic inequalities, while Russell studies probabilistic programming with the goal of improving the interpretability, demonstrable safety, and performance of AI.

Abebe told CNBC on Wednesday that she did not wish to be considered for any award or scholarship, but declined to say why.

Jon Crowcroft, a computer scientist at the University of Cambridge, told CNBC the money should be used to fund researchers trying to make AI more sustainable, adding that many tech giants are too focused on development huge AI models that are very expensive to train.

Crowcroft said he would also like to see more investment in making AI systems interpretable. As things stand, humans don’t know how or why many of today’s most advanced AIs make the decisions they do.

Schmidt chaired the US National Commission on Artificial Intelligence from 2018 to 2021. Last March, the commission found that the United States was significantly underprepared for the age of AI.

In a 756 pages reporthe warned that China could soon replace the United States as the global “AI superpower” and said there were serious military implications to consider.

Artificial intelligence researchers told CNBC last March that there was no need to impose strict regulations on its development at this stage, as the technology is still in its infancy and red tape will only slow progress in the domain.

James G. Williams