3 ms·
Can you be absolutely certain that self driving can deal safely with every possible road irregularities, obstacles, bad, incomplete, damaged or tampered road si
by squarefoot 3y ago
Can you be absolutely certain that self driving can deal safely with every possible road irregularities, obstacles, bad, incomplete, damaged or tampered road signs, and other unpredictable events? Self driving on public roads done right is damn hard; probably harder than sending a manned but completely automated spaceship to Mars and back.
- jjoonathan 3y agoYeah, it's a hard problem, but researchers are smart https://www.youtube.com/watch?v=NRdUjrOnKdw https://www.youtube.com/watch?v=NRdUjrOnKdw and human drivers do not set an infinitely high bar.
- CamperBob2 3y agoThat's not the right question to ask. The right question to ask is, "Can self-driving tech kill fewer than 30,000 people a year, as humans do now?" If you kill the R&D goose or smother it in overregulation, more people will die.
- paulryanrogers 3y agoOr if the R&D phase kills so many people the public loses faith in the whole idea ... then you get to wait a generation for people to forget
- mensetmanusman 3y agoR&D would have to be killing 30,000*%oftheirvehicles per year.
- paulryanrogers 3y agoWe're talking about perception. Regardless, Tesla won't release the data per mile driven and seems to be dodging reporting regulations -- if not daring regulators to respond. Feels to me like amateurs (backed by a small army) confidently building bridges and daring the community to try and stop them. (Except Tesla's army are fan boys, lobbyists, and lawyers.) So basically speed running a millennium of civil engineering, except ignoring the more expensive safety controls; like LIDAR.
- CamperBob2 3y agoWhy don't we see if anything like that happens before just assuming it will? The precautionary principle is a recipe for eternal stagnation.