Automakers Are Rethinking the Timetable for Fully Autonomous Cars
May 20th, 2019A friend of mine is an AI researcher. I know that “AI researcher” is a nebulous term, but let it suffice to say that he is highly qualified in this area.
Back in April, I asked him what he thought about Tesla’s big self driving dog and pony show.
In short, I asked him: Is this going to work?
No.
He’s confident that it’s not going to work under the current paradigm:
Based on my understanding of the technology, I agree with this guy:
“It’s all hype,” said Steven E. Shladover, a retired research engineer at the University of California, Berkeley who has been involved in efforts to create autonomous driving for 45 years. “The technology does not exist to do what he is claiming. He doesn’t have it and neither does anybody else.”
“Unfortunately, it may be necessary for several people to die before regulators step in,” Shladover said.
https://www.chicagotribune.com/business/ct-biz-tesla-self-driving-technology-20190422-story.html
How about just self driving for freight?
My friend thinks the lack of autonomous systems being used for freight indicates that we are far away from a general purpose self driving capability:
But I wonder, why they don’t start by rolling out remote-controlled trucks? That seems like a step on the way toward autonomous vehicles. But for some reason, they never try do this. And, yet, the public is told self-driving cars are just a few years away.
My guess is that, if it happens, it will happen in the freight market first. Freight is a walk in the park compared to the robotaxi concept that Musk is talking about.
With freight, you can train the machine on a specific route and annotate, by hand, the areas that it’s not able to navigate properly. The machine wouldn’t be able to self drive without this training and annotation process.
That is far less sexy than what Musk is claiming is right around the corner, but I could see route specific AI systems being good enough to begin firing some truck drivers in the not too distant future.
As for the robotaxis zipping from one burbclave to another, I’m with my buddy on that one: Not happening anytime soon.
Via: Design News:
The popular notion of the go-anywhere, go-anytime, sleep-in-the-back autonomous car crumbled a bit in the last few weeks, as automakers admitted that the development of full self-driving technology is more difficult than expected.
Questions about the technology’s future reached full public view in April, when Ford Motor Co. CEO Jim Hackett acknowledged what had already become painfully obvious to much of the engineering community. “We overestimated the arrival of autonomous vehicles,” Hackett was quoted as saying by numerous news outlets. “Its applications will be narrow, what we call geo-fenced, because the problem is so complex.”
The admission came as a shocker to many in the public and in the media, essentially because it flew in the face of a growing belief that shiny new autonomous vehicles would soon be landing in dealerships.
Still, Hackett wasn’t the first to make such an concession. The auto industry had been scattering clues to that effect for months before the Ford statement. In November, 2018, for example, John Krafcik, CEO of Google’s self-driving car unit Waymo, had been even more blunt than Hackett. “It’s really, really hard,” Krafcik said during a live-streamed tech conference. “You don’t know what you don’t know until you’re actually in there and trying to do things.”
Krafcik went on to say that the auto industry might never produce a car capable of driving at any time of year, in any weather, under any conditions. “Autonomy will always have some constraints,” he added.
The comments by Krafcik and Hackett reinforced what many industry analysts had been saying for more than two years. “I agree with John Krafcik’s comment,” noted Sam Abuelsamid principal analyst for Navigant Research, which publishes an extensive annual assessment on the state of automated vehicles. “There’s no guarantee that we will ever have automated vehicles in the foreseeable future that are capable of operating everywhere, all the time.”
I’d be very interested to hear what your friend thinks about market and AI potential for the Tesla chip.
I mentioned Tesla’s chip, but he was pretty dismissive of the whole thing. My sense is that he’d slap Musk with a glove and challenge him to a duel, given the opportunity.
This is what I asked:
Do you think Tesla is going to be able to pull off full autonomy with this thing?
They talk about the ASIC they built in a fair amount of technical detail:
https://youtu.be/Ucp0TTmvqOE?t=4152
As I was watching it, I was thinking, “The software is the hard part.” Sure enough, later, Musk says, “All you need to do is improve the software.” hahaha Well, that’s all they need to do. 😉
The thing that makes me wonder if Tesla might actually be able to do it is that they have been fully capturing data from the wildcards, that is, the real world cases where the human driver had to take control because the AI failed.
No other company has that.
I guess the question is: Will that thing be able to infer that a new wildcard situation is similar to known wildcard X in the split second that it would need to do that???
They role out their Big Deal AI dude from Stanford at 1:52.
Yeah, I watched almost the whole shebang. Some pretty tense body language at times.
Since machine learning is the only thing that seems to be making any kind of difference at present in this and other areas of AI/autonomous operation, I thought their chip set-up might have multiple applications outside of Tesla’s primary aims for it and that AI researchers everywhere would want to see what it could do…but I’m far from knowledgeable on this topic.
I work in the automotive business specifically with Ford. Recently we had a meeting at a corporate office in California and we talked about the future of the automotive business for a few mins.
One of the things he said was that Ford is investing tons of cash into smart technology for whole cities. He didn’t get into detail but said basically said your autonomous or semi autonomous car will be simply taken over by the city when you get to a certain point. I asked, so you expect transportation for individual cities to control by a single corporation? Without hesitation he said yes.
Interesting
@brandon Did they mention 5G at all as being a part of that plan?
@Brandon.
Thanks for that. Welcome to the machine.