Mustafa Suleyman cofounded the AI company DeepMind out of a desire to have as broad an impact on society as possible. AI, he decided, was the fastest way to do it.
Now Suleyman has launched DeepMind Health to build AI that can better diagnose disease, including systems that detect early-stage eye disease and help analyze mammograms. He’s also focusing on how such technology is used by medical clinicians. “The tech community is only just finally catching up in thinking about the ethical impact of these systems,” says Suleyman. For instance, will time-pressed clinicians simply defer to the AI’s top suggestions without critical evaluation? How will such systems be audited? And how can new medical findings take into account implicit biases in old data used to train the AI? “I think this is going to be the year when Silicon Valley and the technology companies come to really accept the incredible social responsibility that such great power carries,” he says.
Last year, Suleyman launched the DeepMind Ethics & Society unit to design systems that anticipate and direct algorithms’ decision-making processes and their impact on society.
“The big pivot that technology companies are going to make,” he says, “is to ask the question: How do we shape these algorithms so they represent the moral choices that we collectively elect to make?”