AI for Humanity

CHAPTER 6Sustainable AI for Humanity

Because the stakes are so high and the irreversible situations are so easy to imagine we do need to somehow treat that differently and figure out a new set of safety processes and standards.

— Sam Altman

THE SEARCH FOR SUSTAINABLE AI WISDOM

“Don't Be Evil”: An Unusual Protest

The code name was Maven. The project was top secret and highly sensitive.

Project Maven was a US Department of Defense initiative that sought to use Google's AI to improve drone strike accuracy. The goal? Deploy computer vision algorithms to help Pentagon analysts identify, more effectively, friends from foes by analyzing drone footage.

Google kept Project Maven under close wraps. However, the project soon ignited a firestorm within the company. What started as robust exchanges on internal chats fueled heated staff meetings and even prompted some to resign in protest.

The unrest soon exploded into public view when nearly 4000 employees signed a petition demanding a clear policy stating whether Google would ever build technology for war.

The Project Maven saga forced Google into a precarious balancing act: harnessing powerful AI, pursuing profit, and ensuring ethical governance. The stakes were high. The company risked more than revenue. They stand to lose their top AI talent, who are appalled by even the suggestion of weaponizing AI.

While rivals like Amazon or ...