This is what Tesla Autopilot sees using neural networks that take 70,000 GPU hours to train and output 1,000 tensors (predictions) at each timestep. Source: http://www.tesla.com/autopilotAI pic.twitter.com/eBzvnLJtRn
-
-
If a human was driving, there are chances they would also get fooled or atleast will be tripped to brake., not sure whats the point here??
-
If it can be fooled someone is there it's just a matter of time before a powerful enough projection fools it into thinking nothing is ahead of it when there is. This doesn't even require hacking. That's another issue.
- Još 7 drugih odgovora
Novi razgovor -
-
-
What’s the problem again?
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
One of the major differences here is that you only have to train one computer to recognize this type of thing and every single car in the planet then receives the update and avoids it forever more. Drivers still crash their cars constantly when it rains...
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
Tweet je nedostupan.
-
-
A neural network wont be as good as a human in every aspect. Its meant to be light years better than a human in a few aspects. The fact that it avoids the projection is a good thing, because that means it's doing its best to keep the people in the car safe
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Ok boomer
- Još 2 druga odgovora
Novi razgovor -
-
-
this just proves how good the AI is.. there is literally no point to this post
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
sorry pal, but AI cars will prevent a lot of accidents.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.