Will artificial intelligence cause significant damage in the coming years? In any case, this is the fear raised by Eric Schmidt, ex-CEO of Google.
Will AI end up killing people? If this scenario seems straight inspired by a science fiction film, Eric Schmidt, CEO of Google between 2001 and 2011, ensures that it could become reality. A time member of the National AI Safety Commission in the United States, he now shares his fears related to the rapid development of artificial intelligence models.
For Schmidt, AI will pose serious problems as it develops. Local authorities must ensure that the technology is not ” misused by malicious people “, he warned on the occasion of the Wall Street Journal’s CEO Council in London Wednesday, May 24, relayed in particular by CNBC.
A fiction soon to be real?
Eric Schmidt describes the ” existential risk as the main danger of AI in the future. The ex-CEO of Google defines this bias as likely to cause a ” very large number of people injured or killed. But then, how could a virtual technology affect reality? The specialist talks about new discoveries that could potentially affect the course of humanity.
” There are scenarios, not today, but in the fairly near future, where these systems will be able to find zero-day exploits (security vulnerabilities still unknown to developers, editor’s note) in the field of cybernetics, or to discover new types of biology “, he assures.
A fictitious threat today, but ” likely to be verified ” in the future. And Schmidt adds: When that time comes, our goal is to be ready to prevent these technologies from being used maliciously by malicious people. »
An endless debate
The CEO of OpenAI agrees with Eric Schmidt on this observation. In a note published Monday, May 22 on the company’s website, Sam Altman considered that the main danger of a general artificial intelligence (smarter than humans) was the ” existential riskI. A point of view to be nuanced which remains the subject of lively debate within the scientific community of AI.
Eric Schmidt ensures that the question must go beyond the framework of simple regulations. It’s a broader issue for society “he assures. However, he considers it unlikely that a national body in charge of AI will be set up in the United States.
The fears raised by Schmidt echo the eternal debate on the imminent arrival of an artificial general intelligence (AGI). On March 29, more than 1,125 names in global tech (including Elon Musk, Steve Wozniak or John Hopfield) co-signed a letter and called for a moratorium and a pause in the development of AI. A call quickly contested by other big names in the sector. A debate that is still far from over.
Do you want to know everything about the mobility of tomorrow, from electric cars to pedelecs? Subscribe now to our Watt Else newsletter!