You're right. I should've said "mainland China". Personally, I hate it when people include Mexico when they're talking about central America, for instance.

I hope Kenyan driving culture is compatible with ours. How self-driving AI learning datasets are created. The 'intelligence' of AI is not artificial, it's just a condensed version of the human intelligence that created it.
https://www.bbc.com/news/technology-46055595

Each day, Brenda leaves her home here to catch a bus to the east side of Nairobi where she, along with more than 1,000 colleagues in the same building, work hard on a side of artificial intelligence we hear little about - and see even less.

In her eight-hour shift, she creates training data. Information - images, most often - prepared in a way that computers can understand.

The main ingredient lacking in Artificial Intelligence technologies is intelligence.

 

The main ingredient lacking in Artificial Intelligence technologies is intelligence.

I think the decision-making algorithms are the easy part. I might've mentioned this before, but I think the hardest part is imagine recognition. Something like "this is a bus, and it's size is X*Y, going at Z speed" and "this path here is the road, these are pedestrians, and there's a huge pothole there that needs to be avoided"

Yes, the driving algorithm would probably be massive, but it should be pretty much straightforward, given a clear set of rules. It's the object set and it's clear identification that pose the greatest challenge.

 

I think the decision-making algorithms are the easy part. I might've mentioned this before, but I think the hardest part is imagine recognition. Something like "this is a bus, and it's size is X*Y, going at Z speed" and "this path here is the road, these are pedestrians, and there's a huge pothole there that needs to be avoided"

Yes, the driving algorithm would probably be massive, but it should be pretty much straightforward, given a clear set of rules. It's the object set and it's clear identification that pose the greatest challenge.

Individual object classification and individual object motion physics are the parts that current 'AI' systems are very good at due to the massive learning data-sets from places like Kenya that mainly eliminate the need for traditional image recognition algorithms. The hardest part about perfecting self-driving is handling situations where normal driving rules don't apply and the possible future interaction of humans controlling those objects beyond the rules. At what point does the AI decide 'bad' intent to a driver, dismiss road rule and use its cars mass and energy to block or stop another car? In a lesser case does it use hints of aggression to position the car in heavy or chaotic traffic?

https://www.cbc.ca/news/canada/edmo...inct-to-save-kids-from-car-accident-1.1137795

 

By the time one realizes that one is in an accident.....any intelligence you have.......has no time to consider.

This is why the military trains. Emergency situations require immediate response.

Could AI be capable of running 100-200 simulations in 10 ms....of a local dynamic state? And then choose appropriate action to mediate?

How many people could avoid an accident.....IF when they realize there is an accident.....they could slow time down 1000 times.......to have time to think and regain control or mediate..... to avoid/lessen said accident? Not counting drunks and druggies.

We have not learned how to slow time.......but we can certainly speed it up.

AI isn't smarter.......it's just faster. We have processors that process large data sets now. And for pete's sake.....we have smart phones with a GB of memory now. These units are stamped out cheaply.

AI will come.......but it will be different than thought. Car AI. House AI. Clothing/body/health AI.

Maybe AI will replace insurance. AI...Active Insurance.

 

By the time one realizes that one is in an accident.....any intelligence you have.......has no time to consider.

You can only speak for yourself on this matter. In the few accident I've had over the years I was finding solutions or at least amelioration's to the crash as it was happening to the last instant. I'm sure computers in the future will be faster but they are still limited by the physics of motion to react in the face of uncertainty.

 

You can only speak for yourself on this matter. In the few accident I've had over the years I was finding solutions or at least amelioration's to the crash as it was happening to the last instant. I'm sure computers in the future will be faster but they are still limited by the physics of motion to react in the face of uncertainty.

 

What does this even mean? What happens when an industry is "taken by reform"?

It means that autonomous cars are the future and no matter how much we deny this, the industry is going to keep working to achieve this. While we see only a few names presently, once the systems and processes get streamlined and products start rolling out, even the newbies will join the ride.
Not just this, there will be other significant changes in terms of road/traffic rules/infrastructure and also the insurance industry. Its just a matter of time.

 

It means that autonomous cars are the future and no matter how much we deny this, the industry is going to keep working to achieve this. While we see only a few names presently, once the systems and processes get streamlined and products start rolling out, even the newbies will join the ride.
Not just this, there will be other significant changes in terms of road/traffic rules/infrastructure and also the insurance industry. Its just a matter of time.

It's the future if consumers want it and there is an advantage for consumers. Cars, trucks and most heavy machinery have high upfront costs and long lifetimes so even if they are ready today it will take decades for autonomous cars to be the majority of vehicles on the road. Those Ford F150s and Chevy Silverados will be around for a long time.

 

The hardest part about perfecting self-driving is handling situations where normal driving rules don't apply and the possible future interaction of humans controlling those objects beyond the rules.

https://www.dailymail.co.uk/science...rless-cars-interpret-traffic-like-humans.html

 

https://www.dailymail.co.uk/science...rless-cars-interpret-traffic-like-humans.html

Test it in real Taiwanese traffic.

This system will still use data-sets obtained from human interpretation (maybe from the same people in Kenya) of reactions so it still won't be 'intelligent' on its own. It will have all the same weakness and limitations of Deep Learning classifiers used in other applications.

https://towardsdatascience.com/hype-disadvantages-of-neural-networks-6af04904ba5b

 

It's the future if consumers want it and there is an advantage for consumers. Cars, trucks and most heavy machinery have high upfront costs and long lifetimes so even if they are ready today it will take decades for autonomous cars to be the majority of vehicles on the road. Those Ford F150s and Chevy Silverados will be around for a long time.

I agree. Nobody expects regular vehicles to disappear. Maybe, initial phase would see driverless vehicles for business purpose only.

 

It means that autonomous cars are the future and no matter how much we deny this, the industry is going to keep working to achieve this. While we see only a few names presently, once the systems and processes get streamlined and products start rolling out, even the newbies will join the ride.
Not just this, there will be other significant changes in terms of road/traffic rules/infrastructure and also the insurance industry. Its just a matter of time.

I'm still baffled by how that is taking something by "reform". Are you sure you don't mean taking the industry by "storm"?

 

I will become a believer in true self-driving cars when they can handle Asia style traffic. It's a much less structured environment of small motor bikes, commercial trikes, tiny trucks, pedestrians walking into roads even with crosswalks, roads with no marking, huge trucks on 1.5 lane roads , etc ...

It looks like chaos until you understand the dynamics of human fear and survival to see the small clues of give and take instead of road rules.

That’s why China is aiming to introduce self-driving cars to the market by 2020.

https://interestingengineering.com/why-chinese-artificial-intelligence-will-run-the-world

 

That’s why China is aiming to introduce self-driving cars to the market by 2020.

https://interestingengineering.com/why-chinese-artificial-intelligence-will-run-the-world

The growth of artificial intelligence in China may become the revolutionary development of the current century. “Deep learning is the single greatest invention so far in the Era of Discovery, which was led by the US,” says Kai-Fu Lee, former President of Google China, “but since the deep learning breakthrough, we’ve already entered the Era of Implementation where what matters is execution, product quality, speed, and data. And that’s where China comes in.”

How true do you think their aim is? I'm not a skeptic of Chinese intent, I'm a skeptic of “Deep learning" being the AI breakthrough.
https://arxiv.org/pdf/1801.00631.pdf
https://medium.com/@GaryMarcus/in-defense-of-skepticism-about-deep-learning-6e8bfd5ae0f1

Noise like patterns that fool most DL image classifiers.

 

I'm still baffled by how that is taking something by "reform". Are you sure you don't mean taking the industry by "storm"?

No, I used 'reform' to emphasize the change AI is going to bring in the industry and 'improve' it. The impact will come later and probably would take the market by storm.

 

Here's a market segment for this technology that I had not foreseen :

https://www.foxnews.com/auto/self-driving-vehicles-will-turn-cars-into-brothels-on-wheels-study

Self-driving vehicles will lead to a rise in car sex, according to a new study.

 

You're probably right.

It seems that the porn industry is always ready to exploit every new technology. I remember when the Internet first came out in the 1990s, the porn industry jumped right in and most of the web pages were somehow related to adult sites.

 

Apple Co-founder is NOT confident about seeing SDCs in the future.

“The major concern Wozniak seems to have about the reality of completely self-driving cars is that roads were built by humans, humans are imperfect, and the cars are currently unable to account for this factor.”

The weird thing is, he owns a Tesla and according to rumors, Apple might be refocusing on car production.

http://www.thedrive.com/news/24921/...e-wozniak-doesnt-believe-in-self-driving-cars

 

Apple Co-founder is NOT confident about seeing SDCs in the future.

“The major concern Wozniak seems to have about the reality of completely self-driving cars is that roads were built by humans, humans are imperfect, and the cars are currently unable to account for this factor.”

The weird thing is, he owns a Tesla and according to rumors, Apple might be refocusing on car production.

http://www.thedrive.com/news/24921/...e-wozniak-doesnt-believe-in-self-driving-cars

The weirdest thing is everything/everyone is in a denial mode but the technology is still being worked upon. Its becoming a rat race.
SDC's are going to work, they might just not roll out for the general public soon enough. The infrastructure keeps changing, its keeps remodeling itself with time and technology. Problem is we are not sure what we really should do, since SDC's are pretty much in research phase.

 

When will we have self driving cars? NEVER