This a huge challenge for any type of AI, not just driving. It's a form of uncanny valley. The closer to human performance, the higher the expectations/trust. Unless very close to human, this trust is still always broken, with bigger fall from pedestal the higher the performance.
6 in 100 drivers think it is ok to nap in a Tesla driven by “Autopilot”, a system that in fact should never be left unattended.https://twitter.com/IIHS_autosafety/status/1142219414668009472 …
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Darwin Awards nominees if they weren’t potentially hurting others.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
80 out of 100 drivers have fallen asleep behind the wheel of a regular car.
-
How many thought it was ok? (And, I suspect that like 93.4% of all statistics on Twitter, this one was made up...)
- 2 more replies
New conversation -
-
-
Survival of the fittest must continue somehow
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
There are strict regulations for using algorithms in healthcare, e.g. cancer diagnosis. That’s understandable. Yet in cars, where the consequences of algorithm errors can be even more dire, they apparently can be deployed freely.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So Marcus is going to solve this problem by extending deep learning with symbolic AI? Truth is, DL conflates perception with recognition. Unlike the brain, a DNN cannot perceive an object unless it can already recognize it, a fatal flaw. Marcus is out to lunch.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.