In the final chapter, we look at directions for change. We identify three broad paths. The first is to set ground rules for companies to govern how they build and advertise their products. We think there is an important role for regulation here, while we acknowledge that regulation shouldn’t go overboard. The second path is the way in which we integrate AI into society. For example, what is the role of AI in education and in children’s lives in general? And in the workplace, will we use AI to replace jobs or augment them? We see many of these questions as social and political choices rather than inevitable consequences of the technology itself.
Our third suggested path is to focus on the demand for AI snake oil rather than ist supply. We show that over and over again, AI snake oil is appealing because those buying it are in broken institutions and are desperate for a quick fix. For example, schoolteachers, already overworked, have reacted badly to the disruption caused by students using AI to help with their homework. Unable to carry out the overhaul of their teaching and assessment strategies that AI necessitates, they have turned to cheating detection software. But these products don’t work and have led to a spate of false accusations of academic dishonesty, often with disastrous consequences for students.
We can’t fix these problems by fixing AI. If anything, AI snake oil does us a favor by shining a spotlight on these underlying problems. More broadly, we show how concerns about AI, especially in the labor market, are often really about capitalism. We must urgently figure out how to strengthen existing safety nets and develop new ones so that we can better absorb the shocks caused by rapid technological progress and reap ist benefits.