Our Walter Bradley Center director Robert J. Marks is back with Jonathan Bartlett and Eric Holloway, explaining their choices for Top Ten real advances (“Smash Hits”) in AI in 2020. Readers may recall that we offered a fun series during the holidays about the oopses and ums and ers in the discipline (typically hyped by uncritical sources). Now it’s time to celebrate the real achievements and our nerds think that #8 is the big advances in practical self-driving cars, that is Level Four cars.
The car industry defines five levels of self-driving. Level Five would be Elon Musk’s robotaxis that earn money all on their own while the owners’ sleep (hasn’t happened). Level Four is the practical approach, as Jonathan Bartlett has explained: GM has been given a permit to test five driverless cars on streets in San Francisco later this year, under controlled, mapped conditions. It’s not an automotive Wild West. Anyway:
Our story begins at 13:31. Here’s a partial transcript. (Show Notes and Additional Resources follow, along with a link to the complete transcript.)
Robert J. Marks: “Daimler, Waymo and GM make big gains in level four self-driving cars.” What is Level Four and why is that not the ultimate self-driving car? And then tell us what some of these breakthroughs are.
Jonathan Bartlett (pictured): When people think about self-driving cars, a lot of people think about cars that don’t have steering wheels. You just get in it and you tell the car where you want to go and it takes you there. That’s considered Level Five self-driving, that’s where the car can take you anywhere and does not require any intervention whatsoever. But … there’s a lot of roadblocks to Level Five. At least in my opinion, we probably won’t ever be able to achieve a full Level Five.
But below that is Level Four where you basically define certain parameters for which your self-driving works. So for example, you might have an area of town that you’ve really well mapped and so your car is able to do self-driving in this area. Or below certain speeds you feel comfortable that your car has enough sensors on it that it can drive itself. So, basically, as long as you are within a certain set of parameters that are well-defined, then the car does operate by itself, it doesn’t require any intervention.
What we’ve seen over the last decade or so is that the people who have been pursuing Level Five self-driving are kind of nowhere. They’ve sometimes come up with little tricks and techniques that are interesting, but they haven’t gotten any closer to the actual dream of Level Five self-driving.
But what’s really taken off is Level Four self-driving. And when you see companies that have aimed at having defined areas for their cars to drive in, this has actually worked out really well. And the first company to do this was Daimler, they had a self-driving parking garage. And basically they had cars that would go and park themselves within the parking garage.
That’s because the garage itself had an intelligent infrastructure that had sensors and so the cars weren’t limited to just what they could see. So the cars could find locate a parking space with the help of the infrastructure and navigate there.
And then over the last year we’ve had lots of developments neighborhood taxi services that have been able to drive in low speed neighborhoods and things of that nature.
You can think of Level Four as being an engineering project and Level Five as being a philosophy project.
Robert J. Marks: That’s fascinating. That’s a great analogy, okay.
Jonathan Bartlett: Level Four projects really work well is because they’re defined, they have a scope, you know what the difficulties are, you know what’s going in your favor, and you can really design around a space. But Level Five is the idea that we’re going to have machines that basically think like humans and can participate socially with them with no problems, and that’s more of a philosophy project. And when people confuse their philosophy projects and their engineering projects, we have problems.
Here are the Smash Hits to date:
9 AI Success: Smarter cars for non-millionaires If your car is a recent model, an affordable aftermarket kit might transform it into a much smarter car. One possible risk is that a hacker could take over your car but, no matter what we do with AI, we must deal with security issues.
10 Smash Hit: #10 AI Success!: Translation gets faster and better. Machine translation, properly used, can help us communicate better. What’s made AI tech translation work so well is not that it’s perfect, but we’re going to have a second pass.
- 02:10 | Introducing Jonathan Bartlett
- 02:39 | Introducing Dr. Eric Holloway
- 03:11 | #10: Text translation (Microsoft – https://docs.microsoft.com/en-us/ai-builder/prebuilt-text-translation, Apple – https://apps.apple.com/us/app/translate-translator-ai/id1375535400, DeepL – https://www.deepl.com/en/home)
- 09:19 | #9: “We hit the road with Comma.ai’s assisted-driving tech at CES 2020” (Road Show by CNET), comma two at the Comma.ai shop
- 13:31 | #8: “Daimler, Waymo, and GM Make Big Gains in Level 4 Self-Driving” (Mind Matters News)
- 19:37 | #7: “Hacking AI: Exposing Vulnerabilities in Machine Learning” (Datanami)
- 22:35 | #6: “After Thursday’s Dogfight, It’s Clear: DARPA Gets AI Right” (Mind Matters News)
- Jonathan Bartlett at Discovery.org
- Eric Holloway at Discovery.org
- #10: Text translation (Microsoft – https://docs.microsoft.com/en-us/ai-builder/prebuilt-text-translation, Apple – https://apps.apple.com/us/app/translate-translator-ai/id1375535400, DeepL – https://www.deepl.com/en/home)
- #9: “We hit the road with Comma.ai’s assisted-driving tech at CES 2020” (Road Show by CNET), comma two at the Comma.ai shop
- #8: “Daimler, Waymo, and GM Make Big Gains in Level 4 Self-Driving” (Mind Matters News)
- #7: “Hacking AI: Exposing Vulnerabilities in Machine Learning” (Datanami)
- #6: “After Thursday’s Dogfight, It’s Clear: DARPA Gets AI Right” (Mind Matters News)