Mindscape 109 | Jason Torchinsky on Our Self-Driving Future

Blog post with audio player, show notes, and transcript: www.preposterousuniverse.com/...
Patreon: / seanmcarroll
Mindscape Podcast playlist: • Mindscape Podcast
It’s easy to foresee that technological progress will change how we live; it’s much harder to anticipate exactly how. Self-driving cars represent an enormous technological challenge, but one that is plausibly on the way to being solved. What will be the unanticipated consequences when autonomous vehicles become commonplace? Jason Torchinsky is a fan of technology, but also a fan of driving, and his recent book Robot, Take the Wheel examines how our relationship with cars is likely to change in the near future.
Jason Torchinsky is a senior editor at Jalopnik. His writing has also appeared in venues such as Boing Boing, Muck Rack, and Mother Jones. He is a producer and occasional guest star on Jay Leno’s Garage, and has been the host of the KZbin series Jason Drives.

[@RayWaldin]

I have driven Teslas for over 4 years now, and I have to say, the lack of accurate information in this podcast is very disappointing, and at times borders on scaremongering. Saying Autopilot is dangerous because people aren't accustomed to relegating partial control to the car is not true as evidenced by the fact that traditional cruise control has been used by millions of people for over 30 years now. People don't forget to press the brakes when they encounter traffic. On top of that, adding traffic awareness and lane following isn't MORE dangerous than traditional cruise control. According to recent Tesla safety reports, drivers using Autopilot have one accident per 4.68 million miles driven, compared to the average driver (from NHTSA) who has one accident per 479,000 miles driven. That's a 10x difference. Seems like that would be a relevant fact to mention when discussing the safety of autonomous driving. I really hope you return to having better subject matter experts in future episodes. I've enjoyed many Mindscape podcast episodes but this one was not one of them. www.tesla.com/VehicleSafetyReport

[@mynameisbren]

I had to seek this episode out on KZbin so I could leave a comment. There is a lot more data available on many of the subjects touched on in this episode, and in most cases the data break away from what your guest asserts in his comments. From experience I can also say that he whiffed particularly on describing how level 2 automation works in cars today. He made it sound terrifying while it actually is a) easy to get used to as a driver, and b) saves lives. Single case anecdotes otherwise are not compelling evidence of the contrary. Like other commenters I'll look forward to a more balanced conversation about autonomous vehicles. They are clearly coming, and having "car guys" lament about it isn't really productive.

[@alexseioo610]

He is a layman. Not an ai developer, automotive engineer etc. Even Sean has, as a scientist, more expertise on this topic than the opinionated journalist.

[@anonamouse610]

It was amazing finally hearing a factual discussion about the current state Level 2 autonomous driving from someone who has done a fair amount of investigation and writing on the subject, vs a paid Tesla marketing shill, or worse yet, the stonk-pumper in chief himself, Carnival Barker Elon Musk.

[@jweber4811]

Jason, I just spent 2 hours writing you a long and thoughtful reply that I thought might be helpful to you. Unfortunately, while proofreading, I lost everything and I am not going to try to recreate it. Some of the point I made were that just because someone is smart and has knowledge about a topic does not mean they use their knowledge in an open, free and unbiased manner. And, if they only give one side of the story, their knowledge is useless because it is only one side of the story. (There is no such thing as bias truth.) What naysayers do is make people afraid to even trying autopilot which can cause people to die. The original reason Tesla started pushing for the use of autopilot was because one of their drivers died in an accident and, if the car had autopilot, Elon felt he would have lived. Elon did not want to sit around waiting and have more people die needlessly. Autopilot is not perfect but how many people need to die waiting for naysayers to be happy to give their blessing. Can you ever please a naysayer? I'm glad I did not wait for the naysayers I am one of those people whose life was saved by autopilot. Tesla does not claim that autopilot is perfect. Every time you use it they clearly tell you that it is in BETA version and you have the choice to use it or not. They also give you a thorough description of what autopilot is (that description is not the same as what you describe) and how to use it. Naysayers can come up with all kinds of reasons for not use something but negativity, for negativity sake, does no one any good and it can kill. I've put enough into this and need to end. I am sorry I lost my original reply to you Jason. I believe it could have help guide you in a more honest and positive direction. Good luck to you. I hope that some day you give autopilot a try. It could save your life.

[@anonamouse610]

This is a nicely balanced and thoughtful reply. However, the idea that Elon Musk initiated autonomous driving because of someone's death is about as believable as the Pierre Omidyar's creating eBay as a way to sell his girlfriend's Pez dispenser collection origin story. Musk is a DoD contractor, and autonomous driving (ie autonomous weaponry) has been a sponsored goal of DARPA long before stealing Tesla from Eberhard and Tarpenning was a twinkle in Musk's eye. Musk is an oft documented/self-documented (ie Twitter) pathological liar. As with the rest of Silly-Con Valley, I would take any 'heartstrings' origin story with a very large grain of salt.

[@jweber4811]

Anon A Mouse Aaron you have a right to your opinion on Elon but I do hope you keep an open mind and perhaps, over time , your opinion becomes a little more positive. As far as my comment about autopilot, you are correct. In my original writing on the topic, I did go into it in more detail but, as I indicated, that was lost. On my second, much shorter and less thought out reply, I was not as clear as I should have been. I did not mean to imply that the car accident was the reason Tesla started autopilot. In fact, I believe the early stages of autopilot were already in place at the time of the accident. What the accident did was to encourage Elon to put more attention on autopilot so that it would be available sooner to help save lives. I did not mean to infer that this accident was the reason for auto pilot, I only meant to say that it was, along with other things, an influence to move forward quickly. Thank you forgive me the opportunity to to clarify my statement.

[@seanfitzgerald4207]

I really like the idea of programming heroism into the AV algorithms as an opt-in option for vehicle owners who are good with their vehicles being automatically used in situations to save lives. this reminds me of sitting in the emergency exit seats on airplanes

[@MichaelEdelman1954]

An entertaining episode with an entertaining guest, but I think there are several refutations to his appealing argument against Level 2 autonomy. The first is that Teslas with the autopilot function enabled have a somewhat lower accident rate than Teslas without the system enabled, and a much lower accident rate than the US fleet as a whole. A second point is that just about every driver has had the common experience of having their mind wander and finding themself miles from where they last remember being. In other words, we’re already doing exactly what he worries that automation will make us do. I’ve been driving a car with Level 2 autonomy for a few years now, and two things stand out to me. One is that the lane keeping allows me to concentrate better on everything else going on around me. The other thing is that automatic braking and cruise control “space keeping’ is a fantastic advance in safety. More than once the car has reacted to a car ahead of me braking before I have.

[@robinbrowne5419]

Interesting and fun. Thanks :-)

[@woody7652]

Thanks, Sean!

[@quintinkrivacek9800]

Cars are my first passion so this is quite a nice surprise

[@charliesteiner2334]

The first couple minutes bring up some interesting empirical questions about human behavior - Do people actually, on average, stay alert in semi-autonomous cars? - but seem to tackle them in a 100% armchair way. I mean, I know you said this was a fun episode. So it's probably not worth taking the arguments too seriously.

[@jweber4811]

Charlie Steiner Very well said Charlie. You are the kind of guy I was expecting on the show. Although i found it very difficult to listed to the entire show, I did make it through. I do hope Mindscape does this show again with someone more interested in presenting current information and more accurate information so this audience is not left with such negative and shallow information about EV’s in general and Tesla specifically.

[@ba0cbmft]

@40:36 As long as future actions are taken according to my wishes I am not relinquishing my autonomy since I am allowed to time-shift consent by agreeing to an acceptable set of contextualized outcomes now.

[@mbaske7114]

Things should get interesting once all traffic participants are fully networked. So that in a sticky situation, an autonomous car can make split second decisions based on the age, family status, criminal record, health condition, occupation and credit score of everybody involved.

[@steeneugenpoulsen8174]

How can someone that say they can predict the universe, think that the true self driving car would be unable to avoid EVER be in a situation where it has to choose who it kills? It's not rocket science here to predict if you will be in a situation where there is the danger. That problem for us human is that we are too limited and can't process the data at the level needed to handle driving cars, but a computer does not have that limit, it can see that there is one person that could force it to evade and the ONLY evasion option is into 5 people, so it simply reduce speed until it create a third option where no one dies. The room on wheel thing is also silly, since it wouldn't be worth selling those cars because of all the people that get car sick, you can do reverse seats in trains, because they have many seats, but a car would be too few seats to make half the seats pointless.

[@rhaglen]

Probably, very quickly after the autonomous cars have out-competed regular cars, will road-side businesses that depend on travellers disappear, making the road space between certain destinations become wastelands. Regular road trips will then become much harder if not dangerous, like hiking through the Australian outback, because you'll need to pack everything you need into your non-self driving car before you leave, and if you get stuck, you're doomed, unless you have an autonomous vehicle follow you with spare parts and possibly petrol.

[@einarwennmyr7564]

I'm disappointed in the lack of knowledge when critizising Tesla. They publish data that says that it is almost 10 times less likely to get into an accident when using Autopilot compared to the average american car. Why is this not mentioned or commented on? Secondly, if you go to sleep when using autopilot the car will stop automarically after a few minutes since you have to actively turn the steering wheel slightly for it to continue. You cannot go on for hours as stated in the intreview. Thirdly, the name Autopilot is taken from the aircraft industry where it works similarly to Teslas function. Personally I love to use Autopilot, it lets me forget all micro adjustments to eoad conditions and instead focus on the big picture.

[@TheReferrer72]

He has very good knowledge about this subject, though I don't agree with a lot of what he said at the start about Tesla. 1. that is the data they publish, and will be biased to highway driving. 2 . You can fool that system that was his point.

[@platonicdescartes]

I'm generally in favor of the automated car future, as it will reduce accidents on the road. Car accidents, especially in the US, are quite shocking when you look at the raw numbers. That being said, I still want the option to remain for self driving, especially as someone who loves cars and driving.

[@internationalfatherinlaw5585]

Sean what will happen, If there was a hole in space time? Please reply.

[@WitzyZed]

Spoken like a true supervillain

[@ronjaurigue2673]

Yes! Episode titles screen! Excellent for browsing search results. 😌☕️

[@ulenrich]

We have strategies to bring self-driving cars to a stand still. And we will deploy them. No chance!

[@ulenrich]

I mean what Jason Torchinsky tells about in minute 53 about humans negotiate in traffic

[@Toocrash]

Discount on a protective cab ride

[@pumpuppthevolume]

I wonder if Sean has seen the Doug Demuro video about why buying the i3 makes absolutely no sense :P

[@WitzyZed]

q u i r k s a n d F E A T U R E S

[@pumpuppthevolume]

@@nihlify it's a neat car.... the price and better alternatives make it nonsensical to buy.... otherwise it's fine.... and it's pretty neat that they researched how to make the most eco friendly body.... I guess that's one reason to buy it... to support that research

[@Tanju132]

[car stops in the middle of the road] hmmn...there's something strange about SLOP sign

[@alexseioo610]

No facts, no data. Just opinions. A lot of opinions.

[@anonamouse610]

Great episode!!! Had enough of listening to outrageous claims/marketing pitches about cars self-driving cross-country to come pick you up in 2018, or 1 million robot taxis by 2020... So it was super refreshing to hear intelligent adults discussing the current state of autonomous driving (what's working, what the current limitations are/why its stuck at level two)-while also using some interesting reflection as a basis to look ahead and discuss what humans might want out of their personal modes of transportation in the future.

[@barefootalien]

Man, sometimes he sounds _so_ much like Michael J. Fox!

[@DavidRDavidRoss]

Self driving is a long way away. Don't let the dullards tell you different.

[@robsmith8119]

Very disappointing. I was looking forward to Sean’s usual standard of evidence based discussion. This was simply a guy saying he doesn’t like autopilot and rolling out a series of assertions and opinions, with no mention of any safety studies or suggestions on a better way to get from current levels of automation to what he would consider acceptable levels. I had hoped for a short update on some of the statistics and an open-minded discussion of what autonomous vehicles (which will come and will inevitably be a significant improvement over manually driven) might imply in terms of car ownership, reduced traffic, reduced pollution, changes in street usage, impact on other forms of transport etc. I suppose that is my own misinterpretation of the description but I’d not recommend anyone spend time on this episode.

[@jonathanbowen3640]

I found the introduction odd, going at length to say that it was a more fun episode than educational. I found it neither fun nor educational to be honest. What a shame. Its one of the most important and interesting topics of the day. He did say a few interesting things and made some nice observations, but most if it seemed to just keep go back to how much he enjoys driving.

[@ellusivegman]

first

[@jweber4811]

I am a faithful listener to Mindscape. I have to say, however, I was VERY disappointed with this show and found it to be a total waste of my time. Your guest said nothing new, nothing positive and nothing that gave a true picture of autonomous driving. Your guest obviously has a VERY, VERY, VERY strong bias on the topic. For this reason, the show was not educational - it was simply a hit job. Your guest showed no interest in presenting the topic fairly, truthfully or accurately. My guess is that his only interest was in selling his book to a bunch of "shorts". Autopilot is a wonderful thing. Is it perfect, no. But it is definitely not what your guest described. It is true that autopilot has its faults but it is an amazing technology that is getting better every day. Unlike what your guest (and shorts) try to assert, auto pilot on cars AS ON AIRPLANES does NOT mean that the car can do everything without human assistance. It would have been nice to hear your guest correctly and fairly talk about this. I hope in the future you will bring a guest on to give a fair and accurate description of autonomous driving to correct the false impression given by this show. Thanks for listening Sean. I look forward to your next show.

[@charlesalexanderable]

J Weber it was a good balance against Tesla's lies. Go watch their autonomy investor day presentation from last year and listen to their claimed timeline.

[@jweber4811]

@@charlesalexanderable Charles, you are funny.

[@kory6897]

I had to stop listening after his answer for whether self-auto-2 is currently safer than driving...was expecting some data and real researched answer but got "humans weren't meant to drive in the first place" - twice

[@TheReferrer72]

@Fourier21 humans were not meant to do a lot of things including math, reading eating hot food.....

[@jweber4811]

@Jason Torchinsky Which means nothing. There are a lot of people on the internet who think Barack Obama is not a US Citizen. Should we believe them? People who speak from a bias or quote others with a bias, who speak with some facts but without knowing or hide the full context of those facts or use old information can say just about anything on the internet. If you want to acknowledge it or not, there is a lot of this going on in your presentation. Here and there you provided some truth but, overall, it did not provide a full, fair and honest impression of what is really going on - I don't believe the impression you leave has anything to do with reality? In the past 6-months, have you even driven in autopilot for 20 minutes or more? I am now here to pick an argument with you. We can agree to disagree but most of the shorts I know complain about things they never experienced. (Hard to believe but true.) If you have not given autopilot a fair try, I hope you will. Jim Cramer and Sandy Monroe are two people I assume you know who shut up on their negativity long enough to try it and they are very happy with what they found. With an open mind, I suspect you will too. Good luck and remember, it gets better every day!