Google's Sundar Pichai says we should be 'clear-eyed' about the dangers of AI, calls for regulation

Google's Sundar Pichai today called for AI regulation in an editorial published to the Financial Times. While he noted the virtues of AI, including Google's recent foray into the weather prediction in India, he also raised points about the historical misuse of technology in general.

For AI, that potential misuse isn't potential anymore. The rise of deepfakes has sparked concerns on online platforms. Powerful AI has made it so that privacy can be eliminated with a single photo pulling up details of a person and their social circles.

Pichai argues that when building AI going forward, we should be "clear-eyed" about what could go wrong. Government regulation is key in this, as he stated that the "market forces" should not be the deciding factor in AI applications, having failed so far. "Sensible regulation must also take a proportionate approach, balancing potential harms, especially in high-risk areas, with social opportunities," Pichai said.

Pending the establishment of government regulation, Google already created its own ethical principles back in 2018 and made these tools widely available for interested parties.

Malevolent AI has been a longstanding sci-fi trope. Superman battles Brainiac, the Avengers have their Age of Ultron, and who can forget the Terminators. Despite their sci-fi trappings, they illustrate a fear that people have had since the idea of AI first occurred. AI, when misapplied, can be malevolent. But the fear has proven misplaced. Rather than computer viruses gone rogue or sentient or whatever, it's humans misusing the increasingly powerful tools we've created for "mass surveillance, deception, and fraud. Regulation, sensibly applied, is the only hero that can slay that villain.

Google's AI is better at detecting cancer than doctors, says study