1. Waymo? Hard to tell at the moment but they will be beating Telsla's ~12/100mMiles... somewhat. I would have liked to hear his specific opinion (I infer 'no different to the rest' from the vid though).
2. Humans vs Driverless is not the comparison I want. Some humans are terrible drivers. Driving is not a right that everyone gets - Licenses can be denied or taken away, and people are retested at certain ages. Some humans appear to me to be 10, maybe even 50 (! not hyperbole) times more likely to get into accidents than others. Perhaps we should be starting to tighten regulations. Be more willing to take 'licenses' away (from human, robot or company - doesn't matter) if risk contribution is >threshold. Some people are reckless and have reckless attitudes (as are some corps). I want to see 'Worst human drivers vs driverless' and 'Best human drivers vs driverless'. Appreciate the difficulty in scope, sample selection, possibility of bias etc. If we start to just deny licensing on probability of incident, then we can just slowly tighten further over time.
3. Arguably, all of this investment, research and anti-consumer behaviour is 'planting trees for the next generation'. So we can frame it like that to see some silver lining. We benefit from the 1800s similarly (similar period of state legislation being unable to keep up with innovation). But yes that does assume that there will be a feasible product and healthy industry at the end of it. Similarly, FoodDelivery & TaxiApps, in my opinion, will pay off. They have enabled the disabled now, and possibly improved the future (Yes, I appreciate the inefficiencies of those models).
I was going to submit the same link. Like Benn Jordan, I was an AutoPilot enthusiast. Was going to buy a Model Y with FSD, until a drive from CA to AZ nearly drove me off the road.
I think it can happen, but with current methodology, if has to be a fully mapped data set across known driving roads with car to car communication, with the model having some sort of truth data in relation to gps/compass information for self correction.
Simply mapping sequences of frames to decisions is never going to work without above.
To get fully safe, autonomous driving, we need to have agents that can on the fly adjust to operate on data sets that they have never been exposed to. This is not possible with current training methods.
2. Humans vs Driverless is not the comparison I want. Some humans are terrible drivers. Driving is not a right that everyone gets - Licenses can be denied or taken away, and people are retested at certain ages. Some humans appear to me to be 10, maybe even 50 (! not hyperbole) times more likely to get into accidents than others. Perhaps we should be starting to tighten regulations. Be more willing to take 'licenses' away (from human, robot or company - doesn't matter) if risk contribution is >threshold. Some people are reckless and have reckless attitudes (as are some corps). I want to see 'Worst human drivers vs driverless' and 'Best human drivers vs driverless'. Appreciate the difficulty in scope, sample selection, possibility of bias etc. If we start to just deny licensing on probability of incident, then we can just slowly tighten further over time.
3. Arguably, all of this investment, research and anti-consumer behaviour is 'planting trees for the next generation'. So we can frame it like that to see some silver lining. We benefit from the 1800s similarly (similar period of state legislation being unable to keep up with innovation). But yes that does assume that there will be a feasible product and healthy industry at the end of it. Similarly, FoodDelivery & TaxiApps, in my opinion, will pay off. They have enabled the disabled now, and possibly improved the future (Yes, I appreciate the inefficiencies of those models).