Rebooting AI: Building Artificial Intelligence We Can Trust

Despite all the hype around AI, building an intelligence equal to—or exceeding—human intelligence is far more difficult than we thought. Professors Gary Marcus and Ernest Davis, well-known researchers in artificial intelligence, argue that we’re still a long way from creating fully autonomous cars or superintelligent robots. The reason is that the approaches currently used in AI development are too narrow to create genuine intelligence. The real world, by contrast, is extremely complex and diverse. How can we bridge this gap? What will be the consequences when we do? Inspired by human reasoning, Marcus and Davis explain what we need in order to take AI to a new level. If we focus on giving machines common sense and deep understanding—not simply teaching them to compile ever larger collections of data—we can create AI we truly trust with what matters to us: our health, our homes, the comfort of our loved ones, and much more. The authors offer a comprehensive and vivid assessment of modern science in AI and propose an inspiring vision of how a new generation of AI could make our lives better.