Car automation scenario - kill owner

A good car AI should choose to kill you over killing several others.

Defining conditions where a car AI can decide to kill it's owner on purpose. It's a pretty sensitive topic, but it's going to be one of the most critical UX challenges for autonomous cars.

The scenario:
When a collision between two speeding, autonomous cars is imminent, the car AI systems communicate with each other and figure out the path of least fatality. If crashing a car to avoid colliding with the other saves more lives (but kills the driver), the AI can choose to do that, much like an ethical human driver would do. Fortunately, AI has much faster response time than humans so these decision are quicker and reliable.

To talk more about this, email me at me@arun.co.

***

I actually do a whole lot of UX & very little UI, so you might want to read the case studies on my portfolio instead. I occasionally post on Instagram & Twitter too. 🙂