These Are AI's 'Most Obvious' Risks, According to Google's Former CEO

AI could toggle shape healthcare, the workplace, filmmaking and education — but its powerfulness could besides beryllium destructive successful the incorrect hands.

Google's erstwhile CEO Eric Schmidt elaborated connected the risks of AI successful a Monday section of The Prof G Pod podcast pinch NYU professor, entrepreneur and writer Scott Galloway. Though Schmidt highlighted that AI improves productivity for astir everyone, he besides acknowledged that the exertion could beryllium utilized to wounded quality beings successful the future.

"The astir evident 1 is their usage successful biology," Schmidt stated. "Can these systems, astatine immoderate constituent successful the future, make biologic pathogens that could harm many, many, many, galore humans?"

Related: Worried About AI Stealing Your Job? A New Report Calls These 10 Careers 'AI-Proof'

Today they can't, but location are "a batch of people" who judge that AI will beryllium capable to create harmful biologic materials successful the future, per Schmidt. These group are moving connected preventing AI from being utilized arsenic a biologic weapon, he said.

Former Google CEO Eric Schmidt. Photo by Shahar Azran/Getty Images

Schmidt besides pointed retired that AI opens the doorway to wide cyberattacks that could return down the full financial strategy of a country, for example.

"[AI] systems are truthful powerful that we are rather concerned that successful summation to democracies utilizing them for gains, dictators will usage them to aggregate power," Schmidt stated.

Schmidt was Google's CEO from 2001 to 2011. In August, he made headlines for stating that distant activity was the logic Google fell down the AI title astatine a Stanford University talk. He later apologized for misspeaking; Stanford took down the video of the talk, which garnered complete 40,000 views successful little than a day.

Schmidt isn't the only executive raising concerns astir AI. Sahil Agarwal, a Yale PhD successful applied mathematics who co-founded and presently runs AI information startup Enkrypt, told Entrepreneur earlier this year that arsenic AI systems get much complex, the anticipation that they incorporate implicit bias and toxic contented besides increases.

Related: Ex-Google CEO Says the Company Fell Behind AI Rivals Because of Remote Work. Now He's Taking It Back.