We all want to avoid a Terminator-style future where the machines have conquered humanity, but how do you do that? This is just one of the problems Google's Advanced Technology External Advisory Council (ATEAC) was supposed to help answer. It won't, though. Following some very public drama, Google has decided to dissolve the board and look for a different approach.
The ATEAC was supposed to examine issues like military use of AI, implicit bias in AI algorithms, and how AI can support authoritarian regimes. Google attempted to bring together voices from across the political spectrum, probably hoping to head off another round of complaints that it ignores conservative viewpoints. However, the selection of Heritage Foundation president Kay Coles James for the eight-member panel drew intense criticism. James has been widely criticized for negative comments on trans people and climate change denial. The panel also included Dyan Gibbens, CEO of the military drone manufacturer Trumbull. Many Googlers have objected to the company working with governments on done AI, so this selection seemed particularly tone deaf.
Google intended to have ATEAC meet four times over the course of 2019 and provide feedback about Google's use of AI. Cracks started to show several days ago, shortly after the members of ATEAC became publicly known. Alessandro Acquisti, a privacy researcher from Carnegie Mellon resigned from the board, and others were forced to defend their decisions to remain. Meanwhile, an internal Google petition to remove Kay Coles James collected over 2,000 signatures in a matter of days.
Now, Google confirms that it has decided to end ATEAC, saying that it "can’t function as we wanted." This leaves Google to navigate the ethical issues of AI on its own. ATEAC was seen mostly as a PR move anyway, and there was no reason to keep it around with all the bad press. It's unclear the panel would have really helped Google navigate any sensitive issues, but it's probably better than trusting engineers to make all the decisions.