Elon Musk: Tesla Autopilot | Lex Fridman Podcast #18
Questions on ElonsFAQ asked in this interview (2)
Transcript
0:00Lex FridmanThe following is a conversation with Elon Musk. He's the CEO of Tesla, SpaceX, Neuralink and a co founder of several other companies. This conversation is part of the Artificial Intelligence podcast.
0:13Lex FridmanThe series includes leading researchers in academia and industry, including CEOs and CTOs of automotive, robotics, AI and technology companies. This conversation happened after the release of the paper from our group at MIT on driver functional vigilance during use of Tesla's autopilot pilot.
0:32Lex FridmanThe Tesla team reached out to me, offering a podcast conversation with Mr. Musk. I accepted with full control of questions I could ask and the choice of what is released publicly.
0:43Lex FridmanI ended up editing out nothing of substance. I've never spoken with Elon before this conversation, publicly or privately. Neither he nor his companies have any influence on my opinion, nor on the rigor and integrity of the scientific method that I practice in my position at mit.
1:01Lex FridmanTesla has never financially supported my research and I've never owned a Tesla vehicle. I've never owned Tesla stock. This podcast is not a scientific paper. It is a conversation.
1:14Lex FridmanI respect Elon as I do all other leaders and engineers I've spoken with. We agree on some things and disagree on others. My goal, as always with these conversations, is to understand the way the guest sees the world.
1:27Lex FridmanOne particular point of disagreement in this conversation was the extent to which camera based driver monitoring will improve outcomes and for how long it will remain relevant for AI assisted driving.
1:39Lex FridmanAs someone who works on and is fascinated by human centered artificial intelligence, I believe that if implemented and integrated effectively, camera based driver monitoring is likely to be of benefit in both the short term and and the long term.
1:55Lex FridmanIn contrast, Elon and Tesla's focus is on the improvement of Autopilot such that its statistical safety benefits override any concern of human behavior and psychology. Elon and I may not agree on everything, but I deeply respect the engineering and innovation behind the efforts that he leads.
2:16Lex FridmanMy goal here is to catalyze a rigorous, nuanced and objective discussion in industry and academia on AI assisted driving, one that ultimately makes for a safer and better world.
2:30Lex FridmanAnd now here's my conversation with Elon Musk. What was the vision, the dream of Autopilot when in the beginning the big picture system level when it was first conceived and started being installed in 2014.
2:45Lex FridmanThe hardware and the cars. What was the vision, the dream?
2:49Elon MuskI wouldn't characterize it as a vision or dream, simply that there are obviously two massive revolutions in the automobile industry. One is the transition to electrification and then the other is autonomy.
3:07Elon MuskAnd it became obvious to Me that in the future, any car that does not have autonomy would be about as useful as a horse. Which is not to say that there's no use.
3:20Elon MuskIt's just rare and somewhat idiosyncratic if somebody has a horse at this point. So it's obvious that cars will drive themselves completely. It's just a question of time. And if we did not participate in the autonomy revolution, then our cars would not be useful to people relative to cars that are autonomous.
3:43Elon MuskI mean, an autonomous car is arguably worth five to 10 times more than a car which is not autonomous in the long term. Depends on what you mean by long term, but let's say at least for the next five years, perhaps 10 years.
4:01Lex FridmanSo there are a lot of very interesting design choices with Autopilot early on. First is showing on the instrument cluster or in the Model 3 on the center stack display what the combined sensor suite sees.
4:15Lex FridmanWhat was the thinking behind that choice? Was there a debate? What was the process?
4:20Elon MuskThe whole point of the display is to provide a health check on the vehicle's perception of reality. So the vehicle's taking information from a bunch of sensors, primarily cameras, but also radar and ultrasonics, GPS and so forth.
4:37Elon MuskAnd then that information is then rendered into vector space and that, you know, with a bunch of objects with properties like lane lines and traffic lights and other cars, and then in vector space that is re rendered onto a display.
4:56Elon MuskSo you can confirm whether the car knows what's going on or not by looking out the window.
5:02Lex FridmanRight. I think that's an extremely powerful thing for people to get an understanding, to become one with the system and understanding what the system is capable of. Now, have you considered showing more?
5:14Lex FridmanSo if we look at the computer vision, you know, like road segmentation, lane detection, vehicle detection, object detection underlying the system, there is at the edges some uncertainty. Have you considered revealing the parts that the.
5:30Lex FridmanThe uncertainty in the system, the sort
5:33Elon Muskof probabilities associated with, with say, image recognition or something like that?
5:36Lex FridmanYeah. So right now it shows like the vehicles in the vicinity, a very clean, crisp image. And people do confirm that there's a car in front of me and the system sees there's a car in front of me.
5:46Lex FridmanBut to help people build an intuition of what computer vision is by showing some of the uncertainty.
5:53Elon MuskWell, I think it's my car. I always look at the sort of the debug view, and there's two debug views. One is augmented vision, which I'm sure you've seen, where basically we draw boxes and labels around objects that are recognized.
6:11Elon MuskAnd then there's what we call the visualizer, which is basically a vector space representation summing up the input from all sensors that does not show any pictures, but it shows all of the.
6:27Elon MuskIt basically shows the car's view of the world in vector space. But I think this is very difficult for people to normal people to understand. They would not know what the heck they're looking at.
6:39Lex FridmanSo it's almost an HMI challenge to the current things that are being displayed is optimized for the general public understanding of what the system is capable of.
6:48Elon MuskIt's like if you have no idea how computer vision works or anything, you can still look at the screen and see if the car knows what's going on. And then if you're a development engineer or if you have the development build like I do, then you can see all the debug information.
7:05Elon MuskBut those would just be like total gibberish to most people.
7:11Lex FridmanWhat's your view on how to best distribute effort? So there's three, I would say technical aspects of Autopilot that are really important. So it's the underlying algorithms like the neural network architecture, there's the data so that it's trained on and then there's the hardware development.
7:26Lex FridmanThere may be others, but so look, algorithm data, hardware. You only have so much money, only have so much time. What do you think is the most important thing to allocate resources to?
7:39Lex FridmanOr do you see it as pretty evenly distributed between those three?
7:44Elon MuskWe automatically get vast amounts of data because all of our cars have eight external facing cameras and radar and usually 12 ultrasonic sensors, GPS obviously and IMU. And so we basically have a fleet that has.
8:09Elon MuskWe've got about 400,000 cars on the road that have that level of data. I think you keep quite close track of it actually.
8:15Lex FridmanYes.
8:16Elon MuskYeah. So we're approaching half a million cars on the road that have the full sensor suite.
8:22Lex FridmanYeah.
8:23Elon MuskSo this is. I'm not sure how many other cars on the road have this sensor suite, but I'd be surprised if it's more than 5,000, which means that we have 99% of all the data.
8:36Lex FridmanSo there's this huge inflow of data.
8:38Elon MuskAbsolutely massive inflow of data. And then it's taken us about three years, but now we've finally developed our full self driving computer which can process an order of magnitude as much as the Nvidia system that we currently have in the cars.
8:56Elon MuskAnd it's really just a to use it, you unplug the Nvidia computer and plug the Tesla computer in and that's it. And it's, it's. In fact, we're not even.
9:06Elon MuskWe're still exploring the boundaries of its capabilities. We were able to run the cameras at full frame rate, full resolution, not even crop the images. And it's still got headroom even on one of the systems.
9:19Elon MuskThe full self driving computer is really two computers, two systems on a chip that are fully redundant. So you could put a bolt through basically any part of that system and it still works.
9:30Lex FridmanThe redundancy, are they perfect copies of each other or. So it's purely for redundancy, as opposed to an arguing machine kind of architecture where they're both making decisions. This is purely for redundancy, I think
9:42Elon Muskmore like if you have a twin engine aircraft, commercial aircraft, the system will operate best if both systems are operating, but it's capable of operating safely on one. But as it is right now, we can just run.
10:00Elon MuskWe haven't even hit the edge of performance, so there's no need to actually distribute functionality across both SOCs. We can actually just run a full duplicate on each one.
10:17Lex FridmanSo you haven't really explored or hit the limit of this.
10:20Elon MuskYou have not yet hit the limit.
10:22Lex FridmanSo the magic of deep learning is that it gets better with data. You said there's a huge inflow of data. But the thing about driving the really valuable data to learn from is the edge cases.
10:36Lex FridmanSo how do you. I've heard you talk somewhere about autopilot disengagements being an important moment of time to use. Is there other edge cases or perhaps can you speak to those edge cases, what aspects of them might be valuable or if you have other ideas, how to discover more and more and more edge cases in driving?
11:00Elon MuskWell, there's a lot of things that are learned. There are certainly edge cases where say somebody's on autopilot and they take over and then, okay, that's a trigger that goes to our system that says, okay, did they take over for convenience or do they take over because the autopilot wasn't working properly?
11:19Elon MuskThere's also like, let's say we're trying to figure out what is the optimal spline for traversing an intersection. Then the ones where there are no interventions are the right ones.
11:33Elon MuskSo you then say, okay, when it looks like this, do the following. And then you get the optimal spline for a complex navigating a complex intersection.
11:45Lex FridmanSo that's for this. So there's kind of the common case. You're trying to capture a huge amount of samples of a particular intersection. How when things went right. And then there's the edge case where as you said, not for convenience but something didn't go exactly right.
12:02Elon MuskSomebody took over somebody sort of manual control from Autopilot. And really like the way to look at this is view all input as error. If the user had to do input, it does something.
12:12Elon MuskAll input is error.
12:13Lex FridmanThat's a powerful line to think of it that way because it may very well be error. But if you want to exit the highway or if you want to it's a navigation decision that all Autopilot is not currently designed to do.
12:25Lex FridmanThen the driver takes over.
12:27Elon MuskHow do you know the difference that's going to change with Navigate and Autopilot which we've just released and without still confirmed. So the navigation like lane change based like asserting control in order to do a lane change or exit a freeway or do a highway interchange.
12:44Elon MuskThe vast majority of that will go away with the release that just went out.
12:48Lex FridmanYeah, so that, that I don't think people quite understand how big of a step that is.
12:54Elon MuskYeah, they don't. So if you drive the car then you do.
12:58Lex FridmanSo you still have to keep your hands on the steering wheel currently when it does the automatic lane change. What are. So there's there's these big leaps through the development of Autopilot, through its history.
13:09Lex FridmanAnd what stands out to you as the big leaps? I would say this one, navigate an autopilot without confirm, without having to confirm is a huge leap.
13:21Elon MuskIt is a huge leap. It also automatically overtake slow cars. So it's both navigation and seeking the fastest lane. So it'll overtake a slow cars and exit the freeway and take highway interchanges.
13:40Elon MuskAnd then we have traffic light recognition which introduced initially as a warning, I mean on the development version that I'm driving the car fully stops and goes at traffic lights.
13:57Lex FridmanSo those are the steps. Right. You just mentioned something sort of inkling of a step towards full autonomy. What would you say are the biggest technological roadblocks to full self driving?
14:09Elon MuskActually I think we just the full self driving computer that we just that the Tesla, what we call the FSD computer that's now in production. So if you order any model SRX or any Model 3 that has the full self driving package, you'll get the FSD computer.
14:32Elon MuskThat's important to have enough base computation then refining the neural net and the control software. But all of that can just be provided as an over there update. The thing that's really profound and what I'll be emphasizing at the sort of investor day that we're having focused on autonomy is that the cars currently being produced with the hardware currently being produced is capable of full self driving.
15:00Lex FridmanBut capable is an interesting word because
15:04Elon Musklike the hardware is. And as we refine the software, the capabilities will increase dramatically and then the reliability will increase dramatically and then it will receive regulatory approval. So essentially buying a car today is an investment in the future.
15:19Elon MuskYou're essentially buying. I think the most profound thing is that if you buy a Tesla today, I believe you are buying an appreciating asset, not a depreciating asset.
15:33Lex FridmanSo that's a really important statement there because if hardware is capable enough, that's the hard thing to upgrade.
15:39Elon MuskYes, usually. Exactly.
15:40Lex FridmanSo then the rest is a software problem.
15:43Elon MuskYes, software has no marginal cost really.
15:47Lex FridmanBut what's your intuition on the software side? How hard are the remaining steps to get it to where, you know, the experience, not just the safety, but the full experience, is something that people would enjoy.
16:09Elon MuskWell, I think people enjoy it very much. So on the highways, it's a total game changer for quality of life for using, you know, Tesla Autopilot on the highways.
16:21Elon MuskSo it's really just extending that functionality to city streets, adding in the traffic light recognition, navigating complex intersections and then being able to navigate complicated parking lots so the car can exit a parking space and come and find you, even if it's in a complete maze of a parking lot.
16:46Elon MuskAnd then if, and then you can just, it can just drop you off and find a parking spot by itself.
16:52Lex FridmanYeah. In terms of enjoyability and something that people would actually find a lot of use from the parking lot is a really, you know, it's rich of annoyance when you have to do it manually.
17:04Lex FridmanSo there's a lot of benefit to be gained from automation there. So let me start injecting the human into this discussion a little bit. So let's talk about full autonomy.
17:15Lex FridmanIf you look at the current Level 4 vehicles being tested on road, like Waymo and so on, they're only technically autonomous. They're really level two systems with just a different design philosophy.
17:28Lex FridmanBecause there's always a safety driver in almost all cases and they're monitoring the system. Do you see Tesla's full self driving as still for a time to come, requiring supervision of the human being?
17:44Lex FridmanSo its capabilities are powerful enough to drive, but nevertheless requires a human to still be supervising, just like a safety driver is in a other fully autonomous vehicles?
17:57Elon MuskI think it will require detecting hands on wheel for at least six months or something like that from here,
18:09Lex Fridmanreally,
18:09Elon Muskit's a question of, like, from a regulatory standpoint, how much safer than a person does autopilot need to be for it to be okay to not monitor the car?
18:25Elon MuskAnd this is a debate that one can have. And then, but, you know, a large sample, a large amount of data, so you can prove with high confidence, statistically speaking, that the car is dramatically safer than a person and that adding in the person monitoring does not materially affect the safety.
18:45Elon MuskSo it might need to be like 200 or 300% safer than a person.
18:50Lex FridmanAnd how do you prove that?
18:51Elon MuskIncidents per mile.
18:52Lex FridmanIncidents per mile. So crashes and fatalities, yeah, fatalities would
18:58Elon Muskbe a factor, but there are just not enough fatalities to be statistically significant at scale. But there are enough crashes. There are far more crashes than there are fatalities.
19:10Elon MuskSo you can assess what is the probability of a crash. Then there's another step, which is probability of injury, then probability of permanent injury, then probability of death. And all of those need to be much better than a person by at least perhaps 200%.
19:32Lex FridmanAnd you think there's the ability to have a healthy discourse with the regulatory bodies on this topic.
19:40Elon MuskI mean, there's no question that regulators pay disproportionate amount of attention to that which generates press. This is just an objective fact. And Tesla generates a lot of press.
19:53Elon MuskSo the, you know, in the United States, there's, I think, almost 40,000 automotive deaths per year, but if there are four in Tesla, they'll probably receive a thousand times more press than anyone else.
20:08Lex FridmanSo the psychology of that is actually fascinating. I don't think we'll have enough time to talk about that. But I have to talk to you about the human side of things.
20:17Lex FridmanSo myself and our team at MIT recently released a paper on functional vigilance of drivers while using Autopilot. This is work we've been doing since Autopilot was first released publicly over three years ago, collecting video of driver faces and driver body.
20:34Lex FridmanSo I saw that you tweeted a quote from the abstract, so I can at least guess that you've glanced at it.
20:43Elon MuskYeah.
20:43Lex FridmanRight. Can I talk you through what we found?
20:46Elon MuskSure.
20:46Lex FridmanOkay, so it appears that in the data that we've collected that drivers are maintaining functional vigilance such that we're looking at 18,000 disengagements from autopilot. 18,900. And annotating. Were they able to take over control in a timely manner?
21:05Lex FridmanSo they were there present, looking at the road to take over control? Okay, so this goes against what, what many would predict from the body of literature on vigilance with automation.
21:19Lex FridmanNow the question is, do you think these results hold across the broader population? So ours is just a small subset, do you think? One of the criticism is that, you know, there's a small minority of drivers that may be highly responsible where their vigilance decrement would increase with autopilot use.
21:40Elon MuskI think this is all really going to be swept. I mean, the system's improving so much so fast that this is going to be a moot point very soon.
21:52Elon MuskWhere vigilance is, if something's many times safer than a person, then adding a person, the effect on safety is limited and in fact it could be negative.
22:10Lex FridmanThat's really interesting. So the fact that a human may, some percent of the population may exhibit a vigilance decrement will not affect the overall numbers of safety?
22:22Elon MuskNo, in fact, I think it will become very quickly, maybe even towards the end of this year. But I'd say I'd be shocked if it's not next year at the latest that having a human intervene will decrease safety decrease.
22:40Elon MuskIt's like imagine if you're in an elevator. Now it used to be that there were elevator operators and you couldn't go in an elevator by yourself and work the lever to move between floors.
22:52Elon MuskAnd now nobody wants an elevator operator because the automated elevator that stops the floors is much safer than the elevator operator. And in fact it would be quite dangerous to have someone with a lever that can move the elevator between floors.
23:10Lex FridmanSo that's a really powerful statement and really interesting one. But I also have to ask, from a user experience and from a safety perspective, one of the passions for me algorithmically is camera based detection of just sensing the human, but detecting what the driver is looking at.
23:27Lex FridmanCognitive load, body pose. On the computer vision side, that's a fascinating problem. But do you, and there's many in industry who believe you have to have camera based driver monitoring, do you think there could be benefit gained from driver monitoring?
23:41Elon MuskIf you have a system that's at or below human level reliability, then driver monitoring makes sense. But if your system is dramatically better, more reliable than a human, then driver monitoring is, does not help much.
23:59Elon MuskAnd like I said, just like you wouldn't want someone interview, like you wouldn't want someone in the elevator if you're in an elevator, do you really want someone with a big lever, some random person operating an elevator between floors?
24:12Elon MuskI wouldn't trust that. Or rather have the buttons.
24:17Lex FridmanOkay, you're Optimistic about the pace of improvement of the system. From what you've seen with the full
24:23Elon Muskself driving car computer, the rate of improvement is exponential.
24:28Lex FridmanSo one of the other very interesting design choices early on that connects to this is the operational design domain of autopilot. So where autopilot is able to be turned on.
24:43Lex FridmanSo contrast. Another vehicle system that we're studying is the Cadillac Super Cruise system. That's in terms of odd, very constrained to particular kinds of highways, well mapped, tested, but it's much narrower than the odd of Tesla vehicles.
25:00Lex FridmanThere's add. Yeah, that's good, that's a good line. What was the design decision in that different philosophy of thinking where there's pros and cons, what we see with a wide odd is drive.
25:19Lex FridmanTesla drivers are able to explore more the limitations of the system, at least early on, and they understand together with the instrument cluster display, they start to understand what are the capabilities.
25:30Lex FridmanSo that's a benefit. The con is you're letting drivers use
25:35Elon Muskit basically anywhere anyway that could detect lanes with confidence.
25:41Lex FridmanWas there a philosophy, design decisions that were challenging that were being made there or from the very beginning, was that done on purpose, with intent?
25:54Elon MuskWell, I mean, I think it's frankly it's pretty crazy. Letting people drive a 2 ton death machine manually, that's crazy. In the future people were like, I can't believe anyone was just allowed to drive one of these two ton death machines and they could just drive wherever they wanted, just like elevators.
26:15Elon MuskYou just move the elevator with that lever wherever you want. It can stop it halfway between floors if you want. It's pretty crazy. So it's going to seem like a mad thing in the future that people were driving cars.
26:33Lex FridmanSo I have a bunch of questions about the human psychology, about behavior and so on that would become that. Because you have faith in the AI system? Not faith, but both on the hardware side and the deep learning approach of learning from data will make it just far safer than humans.
26:55Elon MuskYeah, exactly.
26:57Lex FridmanRecently there are a few hackers who tricked Autopilot to act in unexpected ways with adversarial examples. So we all know that neural network systems are very sensitive to minor disturbances to these adversarial examples on input.
27:11Lex FridmanDo you think it's possible to defend against something like this for the industry?
27:15Elon MuskSure.
27:19Lex FridmanCan you elaborate on the confidence behind that answer?
27:23Elon MuskWell, the. A neural net is just like a basic bunch of matrix math. You have to be like a very sophisticated somebody who really understands neural nets and basically reverse engineer how the matrix is being built and then Create a little thing that just exactly causes the matrix math to be slightly off.
27:45Elon MuskBut it's very easy to then block that by having basically anti negative recognition. If the system sees something that looks like a matrix hack, exclude it. It's such an easy thing to do.
28:02Lex FridmanSo learn both on the valid data and the invalid data. So basically learn on the adversarial examples to be able to exclude them.
28:09Elon MuskYeah, you basically want to both know what is a car and what is definitely not a car you train for. This is a car and this is definitely not a car.
28:19Elon MuskThose are two different things. People have no idea. Neural nets, really? They probably think neural nets involves fishing nettle only.
28:29Lex FridmanSo as you know. So taking a step beyond just Tesla and Autopilot, current deep learning approaches still seem in some ways to be far from general intelligence systems. Do you think the current approaches will take us to general intelligence, or do totally new ideas need to be invented?
28:55Elon MuskI think we're missing a few key ideas for general intelligence, general artificial general intelligence, but it's going to be upon us very quickly and then we'll need to figure out what shall we do if we even have that choice.
29:16Elon MuskIt's amazing how people can't differentiate between, say, the narrow AI that allows a car to figure out what a lane line is and navigate streets versus general intelligence. These are just very different things.
29:33Elon MuskYour toaster and your computer are both machines, but one's much more sophisticated than another.
29:38Lex FridmanYou're confident with Tesla you can create the world's best toaster.
29:43Elon MuskWorld's best toaster, yes. The world's best self driving. Yes. To me right now, this seems game, set, match. I mean, I don't want to be complacent or overconfident, but that's what it appears.
29:59Elon MuskThat is just literally how it appears right now. I could be wrong, but it appears to be the case that Tesla is vastly ahead of everyone.
30:10Lex FridmanDo you think we will ever create an AI system that we can love and loves us back in a deep, meaningful way, like in the movie Her?
30:20Elon MuskI think AI will be capable of convincing you to fall in love with it very well.
30:25Lex FridmanAnd that's different than us humans.
30:29Elon MuskYou know, we start getting into a metaphysical question of like, do emotions and thoughts exist in a different realm than the physical? And maybe they do, maybe they don't, I don't know.
30:37Elon MuskBut from a physics standpoint, I tend to think of things, you know, like physics was my main sort of training. And from a physics standpoint, essentially, if it loves you in a way that you can't tell whether it's real or not.
30:55Elon MuskIt is real.
30:57Lex FridmanIt's a physics view of love.
30:59Elon MuskYeah. If there's no. If you cannot prove that it does not. If there's no test that you can apply that would make it allow you to tell the difference, then there is no difference.
31:17Lex FridmanRight. And it's similar to seeing our world, a simulation. There may not be a test to tell the difference between what the real world and the simulation. And therefore, from a physics perspective, it might as well be the same thing.
31:29Elon MuskYes, there may be ways to test whether it's a simulation. There might be. I'm not saying there aren't, but you could certainly imagine that a simulation could correct that.
31:39Elon MuskOnce an entity in the simulation found a way to detect the simulation, it could either restart, you know, pause the simulation, start a new simulation, or do one of many other things that then corrects for that error.
31:53Lex FridmanSo when maybe you or somebody else creates an AGI system and you get to ask her one question, what would that question be?
32:16Elon MuskWhat's outside the simulation?
32:21Lex FridmanElon, thank you so much for talking today. It was a pleasure.
32:24Elon MuskAll right. Thank you,