We have now reached a point where some features that would substantially improve safety[15] are not implemented because they would deteriorate capabilities too much[16].
[16] Also, we cannot adopt othertechnologies for some reason, although they appear to be an improvement in both safety and capabilities. This would be an interesting topic to dive deeper into—is it the “evil greedy company,” or is it some inadequate equilibrium?
The relevant metric is not total deaths per year. Every car moving at 1mm/decade would reduce deaths/year substantially.
A decent[1] metric would be time cost / km[2]. That is: factor in the risk of accidents[3], the time to get from point A to point B itself, and the time cost of working to purchase/maintain the vehicle.
Now, if you show that in this metric these technologies still make sense...
(As an aside, this is the second time recently that I accidentally posted a comment prematurely.)
For getting from point A to point B in general. If you build a vehicle that e.g. refuses to go down a rural road, meaning that someone has to hike instead, that is itself also part of this cost.
The relevant metric is not total deaths per year. Every car moving at 1mm/decade would reduce deaths/year substantially.
A decent[1] metric would be time cost / km[2]. That is: factor in the risk of accidents[3], the time to get from point A to point B itself, and the time cost of working to purchase/maintain the vehicle.
Now, if you show that in this metric these technologies still make sense...
(As an aside, this is the second time recently that I accidentally posted a comment prematurely.)
This metric still has issues. Just lesser ones.
For getting from point A to point B in general. If you build a vehicle that e.g. refuses to go down a rural road, meaning that someone has to hike instead, that is itself also part of this cost.
You can think of something that kills one person with 50 years of life expectancy roughly as though it was just a time cost of 50 years.
Iiinteresting! Thanks for sharing! Yes, the choice of how to measure this affects the outcome a lot..