<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Fri, Jun 24, 2016 at 3:44 PM, BillK <span dir="ltr"><<a href="mailto:pharos@gmail.com" target="_blank">pharos@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class=""><br>
</span>Yes. But the researchers are making the point that people won't buy /<br>
use a robot car that won't take all possible steps to protect the<br>
passengers.<br></blockquote><div><br></div><div>And the first time a self-driving car plows through a pack of pedestrians, Democrats in congress will stage a sit-in demanding that self-driving cars avoid pedestrians.<br> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
That must also be the cheapest / most profitable type of car to build<br>
as well. In the seconds available to analyse the situation, there is<br>
not enough time to make a value judgement of all possible<br>
consequences. Just sending a call for assistance and deciding how best<br>
to protect the passengers while minimising other damage is a difficult<br>
enough task</blockquote><div><br></div><div>Driving a car is pretty damned difficult task. The seconds it takes an accident to unfold are not enough time for humans to well-considered decisions, but it's a relative eternitity for computer. Plenty to time to choose colliding with a vehicle going in the same direction vs. a vehicle coming the other way, or colliding with a low wall vs. an abutment.<br><br></div><div>My point is that self-driving cars will have to make these decisions, and just because buyers want feature X doesn't mean buyers will get feature X.<br><br></div><div>-Dave<br><br></div></div></div></div>