Some of the most profitable companies in the world are joining forces today to create a new nonprofit.
Amazon, Google, Facebook, IBM and Microsoft want you, as well as federal regulators, to know that artificial intelligence can and should be used for good, not evil. The initiative is creatively dubbed the Partnership on Artificial Intelligence to Benefit People and Society, or Partnership on AI, for short.
The tech giants certainly have their work cut out for them.
While few question the potential benefits of AI — how it can be used to scan through troves of data to detect cancer and automate driving to reduce fatalities — a lot of very smart people, including at the White House and the FTC, have been examining how artificially intelligent systems could go awry.
In order for AI systems to work, they must analyze massive amounts of data, but before that data is processed, it’s profiled and categorized by humans, with all their biases and prejudices in tow.
The White House outlined the discriminatory potential of big data in a report released at the beginning of this year. A ProPublica investigation this summer revealed how judges make sentencing and parole decisions by relying on AI software that is proven to be racially biased. And last July, a computer scientist at Carnegie Mellon found Google’s ad-delivery algorithm served men more ads for higher paying jobs than women.
