Google, in the face of employee backlash over its drone-imaging contract with the Pentagon called Project Maven, has put in place some ethical rules to guide its company’s artificial intelligence pursuits.
And the principles do show promise. But let’s be clear: The devil remains in the details.
It’s one thing to lay out a path to walk, a wish-list to fulfill. It’s another thing entirely to have the technological know-how to accomplish these goals.
For instance: Google can say all it wants that its A.I. pursuits ought to “avoid creating or reinforcing unfair bias,” and that employees “will seek to avoid unjust impacts on people, particularly those related to sensitive characteristics such as race, ethnicity, gender, nationality, income, sexual orientation, ability, and political or religious belief.”
But realizing that goal is quite another matter.
Bias within A.I. algorithms has been a head-scratcher of a debate within the technology industry for some time. Moreover, it’s not a problem that will be simply whisked away — no matter how strongly worded it’s willed.
“Silicon Valley is stumped,” blared one CNBC headline in March. “Even A.I. cannot always remove bias from hiring.”
And in November, global management consulting firm McKinsey&Company reported this: “Myths aside, artificial intelligence is as prone to bias as the human kind.”
Much as the technology whizzes are moving to rectify that — and they are — fact is, bias still exists. Truly, the idea of a bias-free machine seems almost as impossible to achieve as a bias-free human mind.
But at least Google’s going there; at least the company is recognizing bias as a factor and making clear its intent is to fight that power.
Also making the tech company’s A.I. ethics’ list is a directive to “be socially beneficial” with technological advances — and not, for instance, seek simply the dollar and ignore the betterment of humankind — and to put safety at the forefront.
“Be accountable to people,” is another one.
That’s a good one. It’s actually the one that brought Google to drop its contract for Maven. When thousands of company employees signed a petition requesting Google part ways with its military research on this project — and dozens more resigned their positions in protest — senior-level executives at the company listened.
Too bad Google doesn’t react so swiftly and decidedly when it comes to complaints from conservatives about censorship on certain company platforms, like YouTube.
But once again — the devil’s in the details.
Google’s list of ethics may be a step in the right direction. But a written wish list does not a successful implementation make. For that, much more work and technological discovery will be needed.
First appeared at The Washington Times.
