Waymo (Google's Self Driving Car Company)

[Deleted member 25]

Why do you still need to have a wheel on a level 4 ? When are you expecting to take over ?

 

[Cashmere]

Why do you still need to have a wheel on a level 4 ? When are you expecting to take over ?

Because no one (currently) manufactures cars without steering wheels.

 

[Sunny]

Tangential -

Someone (Giannis?) argued earlier how self driving car can't react fast enough if a car in front turns without the indicator.

Well, here is a an robot that plays rock paper scissor and will beat a human 100% of time -

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[klier]

Just for fun, I've recently started noting how many near crash experiences I have per day due to other drivers' inability to face basic rules. I average about 4 per day over the last month, with a maximum of 12.

In this world I wouldn't trust a self driving car.

This is a serious contradiction, lol

 

[klier]

Real life is completely messed up environment. There are tons of different, specific, special situations going on constantly. Perhaps too specific & too little of them happening for AI to make a pattern out of it. And even when it does it is limited by basic rules - eg. what to do in dangerous situations - and can't eg perform theoretically the most optimal move ... meaning: it can't create it's own (more optimal) rules. Just like AI player @ game playing. It's set by basic rules.

AI is not set by basic rules. AI is self learning, like you and I.

 

[CrunchSlaughtered]

AI is not set by basic rules. AI is self learning, like you and I.

A little more than 20 years ago, at the dawn of my driving career, I was more a crash test dummy than a driver. For 5 years I destructed completely 7 cars and thanks to the self learning I have managed to reduce the frequency of the accidents drastically. I feel myself almost like an AI. I hope it is going to learn(or to calm dawn) faster.

 

[klier]

I hope it is going to learn(or to calm dawn) faster.

Don’t worry, it will, and then some..

 

[Cashmere]

Having crashed my drone into trees on multiple occasions this is game changing tech and its a great example of how LIDAR and ultra sonic sensors are NOT needed for full autonomy. I was not expecting to see this kind of tech before 2020. Insane to see this on a consumer drone in 2018. The magic is all in the software (Neural Nets being able to identify their surroundings). Confident we'll see self driving cars before the end of this year.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Cashmere]

Waymo is ready for primetime.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Cashmere]

"Self driving cars are stlll 10-15 years away"

2016: Waymo orders 100 vehicles
2017: Waymo orders 500 vehicles
2018 Q1: Waymo orders 2,000 vehicles
2018 Q2: Waymo orders 62,000 vehicles

Waymo’s fleet of self-driving minivans is about to get 100 times bigger
Google spinoff orders 62,000 Chrysler Pacifica minivans

Link: Waymo’s fleet of self-driving minivans is about to get 100 times bigger

 

[tristatez28lt1]

Undoubtedly the leader in autonomous driving.

 

[Deleted member 25]

Impressive.

 

[Sunny]

Undoubtedly the leader in autonomous driving.

Really? And I thought it was Eni


Meanwhile, here are some dumb asses wasting $2+ billion on AD - SoftBank to Invest $2.25B Into Cruise, GM's Self-Driving Car Unit. Only if they had consulted expert chonaka, he would have told him there is no market for AD and saved them $2b.

 

[Centurion]

That's a mega order. This underscores what I mentioned in another related thread, tech giants can dominate AD without manufacturing cars. All they have to do is to develop industry leading tech and before you know it, tech manufacturers will offer to pay for the license or beg for the said company to exclusively use their cars.

Congratulations to Chrysler for selling more minivans.

 

[Deleted member 25]

Do they build themselves the AD hardware ?

 

[Wolfgang]

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Uber in ‘discussions’ to get Waymo self-driving cars on its network

 

[Deleted member 25]

Lol after all that money spent on courts.

 

[Centurion]

Have you all notice that lately Google captcha are all about selecting images of traffic lights, cars, shop fronts or signs?

Google are using the masses to populate training data for their autonomous machine learning.

 

[tristatez28lt1]

Have you all notice that lately Google captcha are all about selecting images of traffic lights, cars, shop fronts or signs?

Google are using the masses to populate training data for their autonomous machine learning.

Damn, nice catch and thinking!! I think you’re spot on!! That would be genius and I could imagine Google doing that. What exactly would they gain though?

 

[Centurion]

Damn, nice catch and thinking!! I think you’re spot on!! That would be genius and I could imagine Google doing that. What exactly would they gain though?

If you want to teach a toddler the difference what a man and woman is, you just need to point out several examples of a man and woman, and the toddler will figure out the differences themselves e.g Men tend to be taller, have deeper voices, rarely wear pink, don't wear skirts, have short hair. A toddler will eventually get it right but the more examples you give them, the greater chance that they will make the right guess in real life. With autonomous driving the tolerance for errors will be 0.00001% so the more training data gathered, the better.

This is the beauty of machine learning. You feed it examples and it learns to classify or predict behaviour by itself. 10 or so years ago AI was less sophisticated and required continuous programming to improve accuracy.

Machine learning is already used by Google images to show you the right result. When browsing my smartphone photo can search for ice cream or a specific car model and Google Photos will return the correct images eventhough there is no meta data or contextual information. The image recognition is very impressive and is being re-purposed for autonomous driving.

Worth mentioning is that deep learning is very cutting edge. In the past you could use AI for an algorithm that can beat human at chess, but that's all the algorithm could do. If you wanted an algorithm for a different purpose you had start from scratch. With deep learning an algorithm can be taught or self-taught to do several things: e.g image recognition AI could become a cab driver, tour guide, stunt driver, traffic officer, racing driver, crime scene detective etc.