On Tuesday, the Knight Foundation announced the Ethics and Governance of Artificial Intelligence Fund, aimed at working out the ethical conundrums around A.I. in the world — in other words, making sure that the machines don’t take over and create a dystopian hellscape. Thanks to expert backers like LinkedIn co-founder Reid Hoffman and eBay co-founder Pierre Omidyar, the fund’s initial commitment is $27 million, but it’s looking for more. Keeping us out of the Matrix is going to cost.
“We’re not going to live with artificial intelligence, we are living in artificial intelligence,” Alberto Ibargüen, CEO of the Knight Foundation, said in a video released on Tuesday. “Figuring out the ethical issues, figuring out who decides, who governs the way this information is distributed, these are essential issues to an informed community in a democracy.”
The sort of ethical questions the fund will answer are unclear, but A.I. ethics is a developing area of interest with many groups seeking to decide on a code. Among some of the issues that arise include racial biases: in September, a beauty pageant judged by A.I. led to only one non-white winner out of 44. Groups like the BSI recommend considering issues like these when designing systems. With A.I., it’s less about hard-coding in a “don’t be racist” rule and more about avoiding a system design that would allow racist results.
It’s the sort of issue that, if considered now, could help avoid a future where A.I. amplifies the same problems in human society on a wider scale. Google DeepMind co-founder Mustafa Suleyman said last month that, considering these biases now, A.I. could help us evolve as a species, as we remove the ethical issues that plague our world in its judgments. It’s a scary proposition: down one path lies a world like ours but worse, but down another lies a more utopian vision.
The Knight Foundation’s fund seems to be aware that it’s not the only A.I. organization in town, but it’s hoping that the focus will help it to stand out in an otherwise crowded industry. A similar group conducting research into A.I. ethics, the Partnership on Artificial Intelligence to Benefit People and Society, is supported by Google, Facebook, Microsoft, IBM, and Amazon. But the partnership’s mission statement focuses on broader “best practices,” which means the Knight Foundation’s fund could differentiate itself in its laser-sharp focus on ethical questions.