Too Fast, Too Furry
The car swerved. The passenger swore. I knew the AI had called me in a moment too late.
I checked the mirrors but realised the problem as I looked up thru the windscreen.
This AI couldn’t interpret the LIDAR readings of a group of people dressed as Pokemon.
I glanced at the dashboard. It was an African AI. Great with crowds but the Japanese ones had the edge with this kind of subculture recognition stuff.
I jammed the brakes on (causing the passenger to shout about his coffee). He had probably seen the “Human Intervening” warning light come on so he’d just get more and more verbal now. I muted the in-car microphone.
Turning the wheel and coming off the brake, I was able to avoid two Pikachu and find a gap in the group for the car to pass thru. It wasn’t a drift manoeuvre that an urban AI was typically trained for.
I knew it. The AI, I knew, knew it too. It had called for help in the available human driver community that it hoped had the experience and reflexes to react best.
The situation was OK. The car was safe. The passenger was safe. The cosplayers were safe.
The car AI switched back to full control. I got credit for helping out. The AI had learnt a new trick and I went back to the kind of driving I knew best.
I was back in the race.
Mario was about to overtake.
Then my mum told me to come down to dinner. Being thirteen still sucks.
I wrote this design fiction after watching my 9 year old driving with a Labo steering wheel and talking about if he’d ever actually drive on a road.
Combining human gaming capability with self-driving car technology as a safety feature is interesting tho.