The AI Dilemma

Conclusion

In the introduction to The AI Dilemma, we acknowledged that the technology in the wrong hands is dangerous but in the right hands is beneficial to all. Furthermore, trying to change its trajectory is more difficult than it might seem. As we saw with the Moral Machine experiment, what people want self-driving cars to do is not universal. It changes depending on context, geography, culture, and global zeitgeist—just to name a few factors.

Therefore, what responsible technology means to us right now may not hold over time.

We also cannot control Triple-A systems by figuring out how they work or reverse-engineering their intentions. They are machines. Machines don’t have intentions. However, as researcher Iyad Rahwan says, “we can judge ...