| > - Do the technically-savvy folks on this list think that a heuristic approach (i.e. a bunch of rules written by humans) could ever succeed at effective self driving in diverse conditions? Or is the machine learning + massive data approach more likely to solve it? The only solution that sounds plausible to me is a hybrid approach. Much like when building a car you can make the engine as complex as you like - but you keep things simple with the brakes. Aircraft autopilots sometimes have three separate processors, programmed by three separate teams. In a self-driving car, I'd expect a heavily audited core of code to take care of not hitting clearly visible stationary objects like concrete barriers and fire trucks even if other parts of the system relied heavily on ML. > Or do you consider it fundamentally unsolveable? I think self-driving vehicles will trade off some types of accidents for others. For example, you see a running child who disappears behind a truck, then a ball rolls out into the road? A skilled, attentive driver would anticipate that the kid might run into the road. But that needs fairly detailed scene understanding. Accidents like that might increase. On the other hand, a lot of accidents are due to lack of attention - and a machine can provide consistent vigilance. So accidents arising due to drowsiness and drivers fiddling with the radio could be expected to decrease. |
But then I realized that detecting these kinds of patterns (kid chasing ball, car next to you that is stuck behind a slow driver, bumper to bumper with a merge approaching, etc.) are precisely what machine learning is good at.
Arguably it could get better at it than any human because it draws not just on the situations one person happens to have encountered in their lifetime. Rather the model that gets uploaded to each car is built from the experiences of millions of cars on billions of miles of driving taking place (eventually) over decades.
Perhaps at some point the quantity of relevant data more than compensates for the greater reasoning and fluidity of a human brain.
It’s hard to know until someone’s tech actually gets there…