Thinking Machines and the Future of Humanity
Within our lifetimes, AI will, by design, begin to behave unpredictably, thinking and acting in ways which defy human logic. Big tech companies may be inadvertently building and enabling vast arrays of intelligent systems that don't share our motivations, desires, or hopes for the future of humanity. Is it too late to change course and realize a human-centered future for artificial intelligence? Stick around after the event for a book signing with Amy Webb.
Chinese and US companies dominate AI technologies in the present and are set to control their future, says Future Today Institute CEO Amy Webb. But the idiosyncrasies of these two innovation powerhouses are resulting in dramatically different visions for the future of humanity’s relationship with AI. Webb explains these differences, and why they matter:
Artificial intelligence technologies are ravenous for data — the more data they can gather, the more effective they are. Amy Webb warns that we haven’t grappled with fundamental questions about our relationship to AI and data, and that we’re facing existential uncertainty as AI data collection works its way deeper into our lives.
What do you think?
The answer to “Who owns your face?” should be obvious, says Webb, but the rise of facial recognition software has complicated that question. What rights do individuals have to their own data when they walk through a store that utilizes facial recognition, for example? Webb says these are the sorts of questions about data collection that we, as a society, are ignoring.
Amy Webb has a radical proposal to create what she calls a “digital dividend.” Her idea is to use blockchain and a collective pot of earnings to give everyone a cut of the money that tech companies are making through data collection:
Would you use a Wal-Mart shopping cart outfitted with a biometric reader? The reader, in the handle of the cart, transmits the shopper’s vitals to in-store employees. Wal-Mart officials think this technology will make for a better shopping experience. It will also make the big box store more money (which is why they’re developing it). Amy Webb says there are inherent problems in this model.
There may be a place for biometric shopping carts in our society, but Webb thinks the financial incentives behind AI technologies are rushing the profound conversations needed in developing such a technology.
Although modest gains towards diversity have been made in recent years, the workforce developing AI technologies is overwhelmingly white and male. And most graduated from the same cadre of universities. Amy Webb suggests a compelling case for diversity in the AI technology workforce:
This excerpt has been lightly edited for clarity
Amy Webb: The people who built the [TSA screening technology] weren’t people who were familiar with enormous hair, or weaves, or bras with underwire...The system wasn’t trained to recognize me. Why? Because somebody like me wasn’t in the room when the data set was built, when the algorithms to use that data were built, when the testing was done, and when the learning was done. That entire chain of decision-making excluded me, which now means that when I go through this stupid machine at the airport I can expect someone to get very familiar with me. That’s a small inconvenience. We could probably spend three hours going through serious infractions in the criminal justice system, and [discover] serious ways in which this is being used to really hurt people.