Artificial intelligence is a hot topic this week, and not just because of the big ChatGPT news from the past few days. World leaders and tech executives met in Paris, France, at the AI Action Summit, where the development of safe AI was an ongoing concern.
The event ended with the signing of a joint agreement to develop safe AI, which the US and UK refused to sign. Even China, a regime known for its tight control of internet access and strong censorship practices, signed the agreement.
DeepSeek, the Chinese AI startup that stunned the world a few weeks ago, is an example of that censorship. The internationally available DeepSeek R1 model will censor itself in real time to avoid addressing topics sensitive to China.
The US and UK refusing to ink the agreement is certainly something we’ll keep talking about every time some sort of AI abuse comes to light, like the viral anti-Kanye deepfake AI video or when key individuals in the Western world manifest their worries about AI getting out of control.
Eric Schmidt, the former CEO of Google and currently an investor in AI, is one of the key individuals you’d listen to when they hypothesize scenarios about AI going rogue or being misused for harmful purposes. The former Google exec told the BBC he’s worried about AI being used in a “Bin Laden scenario,” where bad actors would rely on artificial intelligence to harm others.
“The real fears that I have are not the ones that most people talk about AI – I talk about extreme risk,” Schmidt said in Paris. “Think about North Korea, or Iran, or even Russia, who have some evil goal,” the exec said, warning that artificial intelligence could be used for weapons to create “a bad biological attack from some evil person.”
“I’m always worried about the ‘Osama Bin Laden’ scenario, where you have some truly evil person who takes over some aspect of our modern life and uses it to harm innocent people,” Schmidt said, advocating for government oversight over private companies developing advanced AI products.
“It’s really important that governments understand what we’re doing and keep their eye on us,” he said. “We’re not arguing that we should unilaterally be able to do these things without oversight, we think it should be regulated.”
The BBC notes that Schmidt agreed with the Biden administration’s export controls on microchips that power frontier AI products.
The DeepSeek AI products stunned the world because they were achieved largely without access to advanced AI chips. The Chinese startup used older hardware combined with software innovations to create a reasoning AI that could rival ChatGPT AI. That’s not to say that AI hardware chips do not matter or that Chinese firms can catch up without access to next-gen GPUs.
While Schmidt called for regulation and safe AI development, he also warned that over-regulation could stifle innovation.
Schmidt might not be involved with Google, whose Gemini AI is one of the most important ChatGPT contenders in town, but he is very interested in AI products. A CNCB report showed recently that Schmidt has been a big investor in AI products.
Since 2019, Schmidt’s family office, Hillspire, has invested in 22 AI startups, 13 of which occurred in 2022 alone. ChatGPT became viral in late 2022, turning AI into the main priority for most tech companies.
Schmidt reportedly invested more than $5 billion in these AI startups, including several important names: Anthropic, Holistic AI, and SandboxAQ.
