3 ms·
But even with occasional issues like this one, it's already half as likely to cause accidents than a person[0]. At this point it'd actually be a worse for us to
by underyx 10y ago
But even with occasional issues like this one, it's already half as likely to cause accidents than a person[0]. At this point it'd actually be a worse for us to not test this on open road.
[0]: http://www.telegraph.co.uk/technology/2016/04/25/elon-musk-teslas-autopilot-makes-accidents-50pc-less-likely/ http://www.telegraph.co.uk/technology/2016/04/25/elon-musk-t...
- rgbrenner 10y agoLying with statistics. Compares cruise control with human drivers. But think about that for a second... Highways have an accident rate about 1/3rd of urban roads. Where is cruise control mostly used? Highways. How can you compare human drivers on urban roads AND highways with a system that is nearly always used only on highways. If you compare Tesla's stat with humans on highways only.. Tesla is actually less safe. That stat is complete bullshit.