I mostly agree, but human drivers are pretty terrible too so you don’t have to go too far down the long tail to be safer than the average person...
I completely agree. It is irresponsible of Musk to distort #ai progress. That has led to multiple fatalities. I would not sit in a driverless car today. The world has a long tail. Lots of failure modes. We need regulation. We need testing.https://twitter.com/Captsully/status/1154553807658999808 …
-
-
-
We have an intuitive understanding how humans make mistakes but not for
#ai. Too many failure modes. Lack of calibration. Lack of testing. Tesla ran straight into a concrete block. It hit a white trailer. it's dangerous for this reason.@Captsully - Show replies
New conversation -
-
-
I will assume a bias since you work for nVidia. But, regardless, what would you require to ride with a driverless car? Could you quantify it? If not, it's a game of goalposts. If you can, how does commercial aviation measure up?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I agree we need regulation, but I'd rather government move faster rather than car companies move slower. Every day we delay progress on self-driving cars is another marginal ~100 preventable deaths.
-
Personally, I believe even today's self-driving tech is already better than human drivers on net (even if AI makes mistakes humans don't make). But if one disagrees, we should be able to agree that a self-driving future will be safer than today - let's get to that future ASAP.
End of conversation
New conversation -
-
-
There is a lot of projection hype, no question, Elon is guilty of that, but Autopilot got other manufacturers to start offering driver assistance systems that were always technically possible but they never did before Tesla pushed the envelope,so overall a net gain in safety
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So many mixed feeling about this. I think this is kind of a classic academia vs industry scenario.
-
Like what other consumer technologies were super innovative but dangerous in the past?
End of conversation
New conversation -
-
-
so. perfectly. stated
@AnimaAnandkumar reliability. safety. data. you cannot shortcut your way to progress. it takes work and investment. and diligence. not marketing and grand claims. as a tesla s owner, very aware of the limitations in self driving. not betting my life on it. -
The same goes for healthcare
#AI or healthcare via smart devices. Needs proper regulation and prospective trials before launch.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.