11.13-AutomotiveAI-033123

Friday, March 31, 2023

SmartDrivingCar.com/11.13-AutomotiveAI-033123

13^hedition of the 11th year of SmartDrivingCars eLetter

April 2023 Issue of the Dispatcher

M. Sena, March 29, “ ….In this April issue of The Dispatcher for 2023, I have tried to summarize the discussions on Automotive AI that we had during the 2023 Future Networked Car Symposium. Amidst the warnings about the negative sides of AI, there are some positive signs that AI is being put to good use for automotive safety. I hope you will spare some time to read Musings this month. If you have ever wondered where the “We’re in the railroad business” cliché came from, you will find the answer in Musings. There’s also a tip on where to have lunch when you visit Steamtown National Park in Scranton, PA this summer.….” Read more Hmmmm….. Another great issue. Enjoy the time you spend reading it. Alain

SmartDrivingCars ZoomCast 311 / PodCast 311 w/Michael Sena, Editor of The Dispatcher

F. Fishkin, March 30, “With Elon Musk and others joining the Future of Life Institute in calling for a pause in AI development, what’s next? The Dispatcher publisher Michael Sena joins Alain Kornhauser and Fred Fishkin from Sweden to discuss that plus a freeze on ICE, Ford, Bill Gates and more. Episode 311 of Smart Driving Cars.

0:00 open

0:45 Future of Life Institute calls for pause in further development of AI

27:00 EU Big 3 Countries want to freeze the ICE ban

37:00 Mind your own business but know who your customer really is.. railroads, cars.. mobility

1:04 :00 upcoming Smart Driving Cars Summit

1:04:30 Ford CEO says new electric truck will allow you to nap, use phone, on highways in good weather

1:04:57 Bill Gates posts about going for a ride in a self-driving vehicle in the U.K.

The 6th Annual Princeton SmartDrivingCars Summit

May 22-24, 2023 | Princeton University, Princeton, NJ

Understanding the Business for Driverless Work Vehicles & Finding the Entry Point for a Radical Innovation… Providing Affordable High-quality Mobility to Those Without a Personal Car.

The objective of the 6^th SmartDrivingCars Summit will be to put the eventual manufacturers of driverless passenger vehicles together with the eventual operators of transportation services to decide if there is a business to be made from delivering affordable mobility to a large segment our societies who are underserved by the current options: private cars and public transport.

It’s already happening with military and work vehicles, with Oshkosh and Robotic Research, and John Deere, Caterpillar, Komatsu and others manufacturing driverless vehicles that are getting the jobs done in military operations, on farms, mines, warehouses, seaports, and airports, jobs where experienced drivers are too few in number or where driver safety cannot be ensured. They will be at the Summit during the sessions on Driverless Vehicles in the Line of Duty to explain how they have made the transition to building tools that safely meet the market requirements for both affordability and performance.

After twenty years of development, it’s time for the companies that will build the vehicles and the organizations that will operate the services to take the leadership position. Sensor manufacturers and software developers have gone as far as they can go, and some of them have already had to drop out of the race. Car manufacturers and energy companies are making the same mistake they made twenty-five years ago when they said they had to wait for a battery breakthrough before they could build an electric car. There is a ready and sizeable market for an inexpensive transport solution that has the flexibility of a passenger car for people who cannot drive themselves or afford their own car.

We have focused too much on the equivalent of the battery solution, on the technology, and not on the need. At the SDC Summit in May, we will focus on the need and how to satisfy that need. Join us in Princeton to be part of making this happen. Alain

How safe is safe enough for Autonomous Vehicles

Phil Koopman, March 31, “A pressing question for deploying autonomous vehicles is: will they be safe enough? The usual answer of “at least as safe as a human driver” (e.g., Positive Risk Balance) is likely to be both too simplistic and much more complex than might be apparent. Which human driver, under what conditions? And are fewer total fatalities OK even if it means more pedestrians die? Who gets to decide what safe enough really means when billions of dollars are on the line? And how will anyone really know the outcome will be as safe as it needs to be when the technology initially deploys without a safety driver? This talk covers risk acceptance frameworks, what people mean by “safe,” setting an acceptable safety goal, measuring safety, accounting for uncertainty, a framework for deciding when to deploy, and a sampling of ethical challenges that must be addressed. The emphasis is not on how to build machine learning based systems, but rather on how to measure whether vehicles based on this technology will be acceptably safe for real-world deployment.….” Read more Hmmmm….. Excellent. I do think a more appropriate title would be: “the average driver” would own and use. Phil made it very clear that “The Operational Design Domain (“The ODD”) the fundamental frame of discussion, but the implied ODD for most of the discussion where and when the average person drives their car today.

It is nice for those of us at the MITs, CMUs, Princetons, … of this world to be grappling with what is “safe enough” for a brand new technology that has yet to sell its first vehicle and is expected to deliver all (and more!) of the mobility opportunities and challenges that a technology that has endured all of the growing pains and survived all the pivots/wars/bankruptcies since its birth on January 29, 1886. To be fair, the discussion is not nearly as esoteric as the Trolley Problem. But we are really putting the cart in front of the horse. As Phil succinctly presented, if what we are trying to do is to have in our showrooms cars that we can buy that will accomplish all of the mobility that today’s car delivers to many of us and have one of this new technology’s leading selling points is that the mobility delivered will be “safer” that we, as a collective can do it better, then maybe this isn’t less esoteric of a discussion than “The Trolley Problem”. Steve Schladover has been saying this for years.. I’ll paraphrase. It ain’t going to happen anytime soon and add… I, with probability 99.9999 (or as many zeros as you wish), will not live to see it.

What this industry needs to do is to get real about what mobility attributes it might be able to deliver to whom when and where, and then to have the discussion about the safety constraint (the “enough”, the “>”). It is a constraint. (Yes, I understand the duality between constraints and objectives, but…) The objective is “mobility” ie. The ODD!

The objective is actually “quality of life” and the process that delivers “quality-of-life” is mobility whose one of many constraints is safety. I think!

Also Phil Koopman’s book “How Safe Is Safe Enough?: Measuring and Predicting Autonomous Vehicle Safety” and the hour-long video of his “20 minute presentation”: Alain

Pause Giant AI Experiments: An Open Letter

Open Letter, March 28, “AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research^[1] and acknowledged by top AI labs.^[2] As stated in the widely-endorsed Asilomar AI Principles, Advanced AI could represent a profound change in the history of life on Earth, and should be planned for and managed with commensurate care and resources. Unfortunately, this level of planning and management is not happening, even though recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control.

AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research^[1] and acknowledged by top AI labs.^[2] As stated in the widely-endorsed Asilomar AI Principles, Advanced AI could represent a profound change in the history of life on Earth, and should be planned for and managed with commensurate care and resources. Unfortunately, this level of planning and management is not happening, even though recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control….” Read more Hmmmm….. I simply want to challenge the premise… with human-competitive intelligence “. We aren’t anywhere close! And, whatever crap they produce, is way less interesting than the most banal cat video. All SiliconValleyBank-type smoke mirrors and hype. Alain

Missy Cummings’ Post

M. Cunnings, March 30, “Another wacky petition from the Future of Life Institute https://lnkd.in/eriQCixR Their last fearmongering petition had no impact and now they are at it again. These kinds of chicken-little moments are not productive, and as a mom of a teen, I can promise you that these kinds if sentiments only drive MORE people to use these systems. We need to stop these overly-alarmist, click-baiting tirades and instead start educating people on what AI is and isn’t. We also need to make all community college computer science classes free – the real problem is not ChatGPT or AGI, it is a profound lack of technical literacy.” Read more Hmmmm….. Excellent suggestion and problem statement. Alain

Ford just isn’t into the whole driverless thing anymore

Andrew Hawkins, March 30, “Five months after it pulled the plug on its autonomous vehicle division Argo.ai, Ford is continuing to divest itself from fully driverless technology. Today, the company filed a notice with the feds that basically retreats from its former position of wanting to deploy vast numbers of robot vehicles. The reason?

We believe the road to fully autonomous vehicles, at scale, with a profitable business model, will be a long one.

….” Read more Hmmmm….. What??? If the business model was profitable, what would be wrong about it taking a long time to go from zero to scale, especially when all other business models, except “F-150 ICE”, are on evolutionary roads that are unprofitable and those involving EVs, are trying to get off the ground against a profitable competitor. Talk about something that isn’t going to be easy. At least with driverless, no one is profitable.

But I digress… the challenge of Ford’s venture into driverless mobility is that Ford did not find a problem that achievable driverless mobility could serve profitably in the short run so as to enable its evolution to other more challenging yet more profitable problems. The objective is profitability all along the technological evolution of replacing conventional cars with driverless cars. Fordalong with the rest of the industry has focused on the potentially very profitable end state, rather than finding an evolutionary path that is profitable every step of the evolutionary way.

Henry Ford did not set out to build the F-150 Lightning. Ford Argo.ai should have taken seriously the profitability opportunities of MOVES-style mobility in moderately dense communities with substantial concentrations of households experiencing low car ownership rate.

The latent travel demand characteristics of these communities are for short trips to Walmarts, schools, medical facilities, churches, ball fields, … Trips that can be exceedingly well serve using local and secondary uncongested roads at low speeds and short walk access to & from dispersed kiosks and casual shared riding could be willingly realized. With modest driverless vehicles such Operational Design Domains could be safely served 24/7/350 at fare levels and vehicle productivities that would be modestly profitable. This evolutionary road can be increasingly profitable leading to enormous scale economies and profits,

Ford either didn’t see it or didn’t believe it or simply didn’t want it. Maybe GM or Tesla or Zoox or a Chinse company will see it and want it. Or, I’ll simply fade away into the sunset. Alain

INSIDER Ford CEO channels Elon Musk, saying the automaker’s next electric truck will drive itself while you sleep

A. St. John, March 24, “…In Elon Musk-like fashion, Ford CEO Jim Farley made bullish claims about self-driving tech and said the automaker’s next electric pickup will drive itself.

In a new interview with Bloomberg published Friday, Farley upped his company’s pressure on Tesla, saying the company’s next electric truck set to debut in 2025 is a “breakthrough product.”

“On the highway on a sunny day, you should be able to go to sleep in your truck or make a call or do whatever you want to do in your truck while it drives for you,” Farley told Bloomberg….” Read more Hmmmm….. Wow. You can’t make up this stuff. Totally, absolutely, irresponsible! Does he realize that he has now acquired the liability associated with every Ford Lightning crash involving a driver that dozes off?! And Ford will have to settle every Lightning crash. The “dog ate my homework” excuse is going to be: “I fell asleep ☹ Farley said I could. His fault, not mine.” Hope he’s ready for the consequences of his comments. Alain

Tesla releases FSD update, leaps forward in safety and customization

W. Johnson, April 1, “Tesla has released its latest update to the Full Self Driving software, version 11.3.4, which brings massive improvements to safety and customization, along with a vast set of new features.

…Tesla has released yet another update to the FSD system, bringing a slew of safety improvements, customization updates, and new features.

According to NotaTeslaApp, the most profound change is coming to the software’s highway functionality. On the technical side, Tesla has merged the “vision” and “planning” stacks and removed the four-year-old highway stack altogether. This improvement integrates Tesla’s famous multi-camera video network system and next-gen planner, allowing for more complex interactions, less reliance on painted lanes, and introduces an overall improvement in vehicle behavior.

The Tesla FSD highway improvement also improves the vehicle’s positioning in wide lanes, improves handling in significant turns, and improves the vehicle’s positioning during cut-in scenarios, all thanks to the mining of Tesla’s growing library of fleet data….” Read more Hmmmm…..

Sure, the name of this product is absolutely horrible; however, the process of letting at least some customers experience improvement in the product is commendable. Since 2014 I’ve owned a Mercedes with intelligent cruise control, lane centering and other “Level 2” features. During the ensuing 9 years, my capabilities and features were enhanced once… when I bought a used 2018 version of the 2014 S-Class. Plus, I’ve never been contacted by Mercedes about any fixes or other improvements in those comfort, convenience and safety features. Will I really stay loyal to Daimler and buy a 2024 model to get the one-time upgrades to these Comfort, Convenience and Safety features? C’mon Mercedes. Join the 21^st century. It is almost 25% over.

Tesla Teases New Product, Semi Recall, Q1 Expectations

R. Maurer, March 31, ” ➤ Delivery and production expectations for Tesla’s first quarter of 2023 ➤ PCE report below forecast ➤ Preliminary EV credit adjustments issued ➤ Tesla China teases new product unveiling ➤ Tesla recalls Tesla Semi ➤ Franz von Holzhausen discuss Tesla in Motor Trend interview:

• Tesla Chief Desig…

0:00 Intro / TSLA

0:36 PCE report

1:08 EV credit update

3:21 New product teaser

5:26 Semi recall

6:37 Franz interview

10:41 Q1 delivery and production forecasts

Read more Hmmmm….. Continues to be interesting. Alain

As heavy EVs proliferate, their weight may be a drag on safety

R. Arbelaez, March 9, “How safe are electric vehicles? It’s a question that keeps coming up in different ways in my role overseeing our vehicle crashworthiness evaluations.

When we first started testing electric vehicles in 2011, the question referred to fire risk. When damaged, lithium-ion batteries can ignite, and the fires can be extremely hard to put out. As a result, we have special procedures for EV testing, including monitoring the battery’s voltage and temperature and asking our local fire department to be on site.

Fifty-five EV crash tests and zero fires later, I’m still worried about these batteries, but for a more mundane reason than their potential to burst into flames. My biggest concern is how heavy they are and what all that extra vehicle weight means for the safety of people on the road, specifically occupants of lighter vehicles as well as pedestrians and bicyclists….” Read more Hmmmm….. EVs are nice, but…” Alain

Pause Giant AI Experiments: An Open Letter

AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research^[1] and acknowledged by top AI labs.^[2] As stated in the widely-endorsed Asilomar AI Principles, Advanced AI could represent a profound change in the history of life on Earth, and should be planned for and managed with commensurate care and resources. Unfortunately, this level of planning and management is not happening, even though recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control….” Read more Hmmmm….. I simply want to challenge the premise… with human-competitive intelligence “. We aren’t anywhere close! And, whatever crap they produce, is way less interesting than the worse cat video. All SiliconValleyBank-type smoke mirrors and hype. Alain

The rules of the road are about to change

B. Gates, March 29, “… The result was a memorable ride. The car drove us around downtown London, which is one of the most challenging driving environments imaginable, and it was a bit surreal to be in the car as it dodged all the traffic. (Since the car is still in development, we had a safety driver in the car just in case, and she assumed control several times.)…” Read more Hmmmm….. Wow… sounds like riding around in a Waymo vehicle 10 years ago. Where have you been, Bill? The fact that during your ride …” she assumed control several times “ means that this vehicle isn’t anywhere close to being driverless, let alone autonomous in anything approaching the London ODD. Sorry, but he sounds like Bush senior did when he visited the exhibit hall of the National Grocer’s Assoc. in 1992.