埃隆·马斯克：特斯拉自动驾驶系统

以下是与埃隆·马斯克的对话。他是特斯拉、SpaceX、Neuralink的首席执行官，也是多家其他公司的联合创始人。本次对话是人工智能播客系列的一部分，该系列涵盖了学术界和工业界的顶尖研究者，包括汽车、机器人、人工智能及科技公司的CEO和CTO们。这次对话发生在麻省理工学院我们团队关于特斯拉自动驾驶模式下驾驶员功能警觉性的论文发表之后。

The following is a conversation with Elon Musk. He's the CEO of Tesla, SpaceX, Neuralink, and a cofounder of several other companies. This conversation is part of the artificial intelligence podcast. The 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.

特斯拉团队联系我，提议与马斯克先生进行播客对话。我接受了邀请，并保留了对提问内容和公开内容的选择权。最终我未对实质内容做任何删减。在此次对话之前，我从未与埃隆有过任何公开或私下的交流。他及其公司对我的观点、以及我在麻省理工学院职位上所秉持的科学方法的严谨性和完整性均无任何影响。

The Tesla team reached out to me offering a podcast conversation with mister Musk. I accepted with full control of questions I could ask and the choice of what is released publicly. I 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.

特斯拉从未资助过我的研究，我也从未拥有过特斯拉汽车或股票。这期播客并非科学论文，而是一次对话。我尊重埃隆，就如同尊重所有与我交谈过的领导者和工程师一样。

Tesla 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. I respect Elon as I do all other leaders and engineers I've spoken with.

我们在某些事上意见一致，在另一些事上则存在分歧。我始终希望通过这些对话理解嘉宾的世界观。本次对话中一个具体分歧在于：基于摄像头的驾驶员监控能在多大程度上改善驾驶结果，以及这项技术对AI辅助驾驶的适用期有多长。作为一名研究并痴迷于人本人工智能的学者，我认为若实施得当，摄像头监控在短期和长期都可能带来益处。而埃隆和特斯拉则更关注提升自动驾驶系统本身，使其统计安全性优势完全覆盖对人类行为和心理因素的考量。

We agree on some things and disagree on others. My goal is always with these conversations is to understand the way the guest sees the world. One 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. As 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 the long term. In 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.

尽管埃隆与我不尽相同，但我对他所主导的工程与创新深表敬意。我的目标是推动工业界和学术界就AI辅助驾驶展开严谨、细致且客观的讨论，最终让世界变得更安全、更美好。现在请听我与埃隆·马斯克的对话：在最初构想并于2014年开始部署时，Autopilot系统在硬件和车辆层面的整体愿景和梦想是什么？

Elon and I may not agree on everything, but I deeply respect the engineering and innovation behind the efforts that he leads. My 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. And 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, the hardware and the cars? What was the vision, the dream?

我不愿将其称为愿景或梦想。汽车行业显然正在经历两大革命：一是电动化转型，二是自动驾驶。对我而言，未来没有自动驾驶功能的汽车就像现在的马匹——并非完全无用，但已属罕见且特立独行。

I wouldn't characterize it as a vision or dream. Simply that there are obviously two massive revolutions in in the automobile industry. One is the transition to elect electrification, and then the other is autonomy. And it became obvious to me that in the future, any 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. It's just rare and somewhat idiosyncratic if somebody has a horse at this point.

汽车终将实现完全自动驾驶，这只是时间问题。如果我们不参与这场自动驾驶革命，我们的汽车相对于自动驾驶汽车将失去实用价值。可以说，一辆自动驾驶汽车的价值是非自动驾驶汽车的5到10倍。

It's just 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. I mean, an autonomous car is arguably worth five to 10 times more than a car which is not autonomous.

从长远来看。

In the long term.

这取决于你所说的长期是指多久，但至少在未来五年，或许十年内。

Depends what you mean by long term, but let's say at least for the next five years, perhaps ten years.

早期自动驾驶系统有许多非常有趣的设计选择。首先是仪表盘上的显示——对于Model 3则是中控屏——展示整套传感器所感知的内容。

So there are a lot of very interesting design choices with Autopilot early on. First is showing on the instrument cluster, or in the Model three on the center stack display, what the combined sensor suite sees. What

这个设计背后的考量是什么？当时有争议吗？具体决策流程是怎样的？这个显示的核心目的是让车主能核验车辆对现实环境的感知状态。车辆通过各类传感器（主要是摄像头，也包括雷达、超声波、GPS等）采集信息，

was the thinking behind that choice? Was there a debate? What was the process? The whole point of the display is to provide a health check on the the vehicle's perception of reality. So, the vehicle's taking in information from a bunch of sensors, primarily cameras, but also radar and ultrasonics, GPS and so forth.

这些信息随后被转化为矢量空间数据，包含车道线、交通灯、其他车辆等带有属性的对象。接着矢量空间数据会重新渲染到屏幕上，这样你通过观察窗外就能确认车辆是否准确识别了环境。

And then that information is then rendered into vector space and that, you know, with a bunch of objects with with properties like lane lines and traffic lights and other cars. And then in vector space, that is re rendered onto a display so you can confirm whether the car knows what's going on or not by looking out the window.

没错。我认为这对用户理解系统、与系统建立默契认知其能力边界是极其有效的方式。那么是否考虑过展示更多信息？比如计算机视觉底层的道路分割、车道检测、车辆检测、物体检测等算法在边缘地带存在不确定性，是否考虑过向用户揭示系统的不确定性部分？例如图像

Right. 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? So, 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, the uncertainty in the system, sort of Probabilities associated with, say, image

识别概率之类的参数？

recognition or something like that?

是的，现在系统显示周围车辆的情况，图像非常清晰干净，人们确实确认我前方有辆车，系统也检测到了前方车辆，但为了帮助人们建立对计算机视觉的直观理解，我们会展示一些不确定性。

Yeah, so right now it shows like the vehicles in the vicinity, 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, but to help people build an intuition of what computer vision is by showing some of the uncertainty.

嗯，我认为是在我的车里，我总是查看调试视图，实际上有两种调试视图。一种是增强视觉，你们可能见过，基本上就是给识别出的物体画框加标签。另一种我们称之为可视化器，本质上是将所有传感器输入汇总成向量空间表示。它不显示任何图片，而是以向量空间形式展现车辆对世界的认知。但我觉得这对普通人来说非常难理解。

Well, I think it's in my car, always look at the sort of the debug view, and there's there's two debug views. One is augmented vision, where which I'm sure you've seen where it's basically we we draw boxes and labels around objects that are recognized. And 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 it 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.

他们根本看不懂自己看到的是什么技术。

They would not know what tech they're looking at.

所以这几乎是人机交互的挑战，目前显示的内容是为了让公众理解系统能力而优化的。

So 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.

就像如果你完全不懂计算机视觉原理，你也能看着屏幕判断车辆是否感知到周围环境。但如果你是开发工程师，或者像我一样有开发版本，就能看到所有调试信息。不过这些对大多数人来说就是天书。

It's like if you have no idea what how computer vision works or anything, you can still look at the screen and and see if the card knows what's going on. And then if you're, you know, if you're a development engineer or if you're, you know, if you're you have the development build like I do, then you can see, you know, all the debug information. But those would just be, like, total gibberish to most people.

没错。你如何看待最优资源分配？我认为自动驾驶有三个非常重要的技术方面：底层算法（比如神经网络架构）、训练数据、以及硬件开发。可能还有其他方面，但主要就是算法、数据、硬件，而资金和时间都是有限的。

Right. What'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 that's trained on, and then there's the hardware development. There may be others, but, so, look, algorithm, data, hardware, you you only have so much money, you only have so much time.

你认为应该优先把资源投在哪个方面，还是觉得这三者应该均衡分配？

What do you think is the most important thing to to allocate resources to, or do you see it as pretty evenly distributed between those three?

我们能够自动获取大量数据，因为所有车辆都配备了八个外置摄像头、雷达、通常还有12个超声波传感器，当然还有GPS和IMU。因此我们实际上拥有一个车队，路上大约有40万辆汽车具备这种数据采集能力。说实话，你们对数据的追踪相当严密。

We automatically get fast 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 we basically have a fleet that has and we've got about 400,000 cars on the road that have that level of data. Actually, think you keep quite close track of it, actually.

是的。

Yes.

没错。我们即将达到50万辆配备全套传感器的上路车辆。我不确定路上还有其他多少车辆装有这种传感器套件，但如果有超过5000辆我会很惊讶——这意味着我们掌握了99%的相关数据。

Yeah. So we're approaching half a million cars on the road that have the full sensor suite. Yeah. So 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.

所以这形成了巨大的数据流入

So there's this huge Absolutely, inflow of

海量数据涌入。我们花了大约三年时间，现在终于开发出了全自动驾驶计算机，其处理能力是目前车载NVIDIA系统的十倍。使用方式很简单：拔掉NVIDIA电脑，插入特斯拉电脑即可。事实上我们仍在探索它的性能边界——它能以全帧率、全分辨率运行所有摄像头，甚至无需裁剪图像，单系统仍有性能余量。

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. And it's really just 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 we'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 the systems.

这套全自动驾驶硬件实际上是两台计算机，两个完全冗余的片上系统。即使该系统任何部分被击穿，仍能正常工作。

The hard full self driving computer is really two computers, two systems on a chip that are fully redundant. So you could put a build through basically any part of that system, and it still works.

这种冗余设计，它们是彼此的完美复制品吗？还是说...？另外，这是纯粹为了冗余，而不是像争议机器架构那样让两者都参与决策？这就是纯粹的冗余设计？

The redundancy, are they perfect copies of each other? Or Yeah. Also, it's purely for redundancy as opposed to an arguing machine kinda architecture where they're both making decisions. This is purely for redundancy.

我更倾向于这样理解：就像一架双引擎商用飞机，当两套系统都正常运行时效果最佳，但单系统运行也能保障安全。就目前而言，我们甚至还没触及性能极限，所以实际上没必要将功能分散到两个SOC上。我们完全可以在每个SOC上运行完整的重复系统。

I think of it more like it's if you have, say, a twin engine aircraft, commercial aircraft, the system will operate best if both systems are operating, but it's it's capable of operating safely on one. So but as is right now, we can just run we're we haven't even hit the the the edge of performance, so there's no need to actually distribute functionality across both SOCs. We we can actually just run a full duplicate on on each one.

所以你们其实还没有真正探索或触及这个系统的极限？

So you haven't really explored or hit the limit of this

确实还没有达到极限，没有。

They have not yet hit the limit, no.

深度学习的魔力在于它会随着数据积累而不断优化。你提到有大量数据涌入，但驾驶领域真正有价值的学习数据其实是那些边缘案例。我听说你曾提到自动驾驶系统解除的瞬间是关键时间点。除此之外还有哪些边缘案例？能否详细谈谈这些特殊情况？

So 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. So how do you, I mean, I've heard you talk somewhere about autopilot disengagements being an important moment of time Yes. To use. Is there other edge cases, or perhaps can you speak to those edge cases?

这些案例中有哪些方面特别有价值？或者你有没有其他方法来发现越来越多驾驶中的边缘案例？

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?

确实有很多值得学习的方面。比如当用户在使用自动驾驶时突然接管，系统就会触发警报：这次接管是出于便利，还是因为自动驾驶出了问题？再比如，我们在研究如何规划最优的过弯曲线时，那些没有人工干预且系统处理正确的案例就是最佳样本。

Well, 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 take over for convenience, or did they take over because the autopilot wasn't working properly? There's also, like, let's say we're trying to figure out what is the optimal spline for traversing an intersection. Right. Then then the ones where there are no interventions and and we are the right ones.

于是你就能确定：当出现类似情况时执行以下操作。这样就能为复杂路况导航得出最优的过弯曲线。

So you then say, okay. When it looks like this, do the following. And then and then and then you get the optimal spline for a complex navigating a complex

交叉路口。所以这是针对常见情况的。嗯。你试图收集特定交叉路口的大量样本，研究事情顺利时的情况。然后还有边缘情况，如你所说，不是为了方便，而是某些环节出了问题。

intersection. So that's for this so there's kinda the common case. Mhmm. So 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.

有人接管了，有人从自动驾驶模式切换到了手动控制。实际上，看待这个问题的方式是把所有人工输入都视为错误。如果用户不得不进行输入，那就说明所有输入都是错误。

Somebody took over somebody asserted 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 all input is error.

这种思考方式很有力量，因为它很可能确实是错误。但如果你想驶离高速公路，或者做出自动驾驶目前无法处理的导航决策时，驾驶员就会接管。你怎么区分

That's a powerful line to think of it that way because it may very well be error. But if you wanna exit the highway or if you want to it's a navigation decision that all Autopilot is not currently designed to do, then the the driver takes over. How do you know

其中的差异？随着我们刚发布的导航版自动驾驶功能及停车确认功能，这种情况将改变。基于车道变换的导航控制，比如变道、驶离高速或进行高速互通，大部分这类操作在新版本中都将消失。

the difference? That's gonna change with Navigator and Autopilot, which we we just released, and and with that stall confirm. So the navigation like, lane change based, a certain control in order to do a lane change or exit a freeway or doing a highway interchange, the vast majority of that will go away with the release that just went out.

是的。我觉得人们还没有完全意识到这一进步有多重大。

Yeah. So that I don't think people quite understand how big of a step that is.

没错，他们确实没意识到。只有亲自驾驶过这辆车的人才会明白。

Yeah, they don't. If you drive the car, then you do.

所以目前在进行自动变道时，你仍需将手放在方向盘上。在自动驾驶的发展历程中，有哪些重大突破让你印象深刻？我认为这次无需确认的导航自动驾驶就是一个巨大飞跃。

So you still have to keep your hands on the steering wheel currently when it does the automatic lane change. What are, so there's these big leaps through the development of autopilot, through its history, and what stands out to you as the big leaps? I would say this one, navigating autopilot without confirm without having to confirm is a huge leap.

这是一个巨大的飞跃。

It is a huge leap.

它会怎样

What it'll

该系统会自动超车慢行车辆。它兼具导航与寻找最快车道的功能。它会自动超车、驶离高速并选择立交桥路线。此外我们还配备了交通信号灯识别功能，初期会以警示形式呈现。在我试驾的开发版本中，车辆已能完全自主识别红绿灯并启停。

It'll is automatically overtake slow cars. So it's it's both navigation and seeking the fastest lane. So it'll it'll it'll, you know, overtake a slow cause and exit the freeway and take highway interchanges. And and then we have traffic light traffic light recognition, which is introduced initially as a warning. I mean, on the development version that I'm driving, the car fully fully stops and goes at traffic lights.

所以这些都是阶段性进展对吧？你刚才提到了实现完全自动驾驶的初步构想。嗯。你认为实现完全自动驾驶最大的技术障碍是什么？

So those are the steps, right? You've just mentioned something, of an inkling of a step towards full autonomy. Mhmm. What would you say are the biggest technological roadblocks to full self driving?

实际上我认为——我们刚推出的全自动驾驶计算机（特斯拉称之为FSD芯片）现已投产。订购任何搭载全自动驾驶套件的Model SRX或Model 3车型都将配备该芯片。拥有足够的基础算力至关重要，之后只需持续优化神经网络与控制软件，这些都能通过远程升级实现。真正深远的意义在于——这也是我将在自动驾驶主题投资者日重点强调的——目前量产的车辆

Actually, I don't think I think we just the full self driving computer that we just that the the Tesla, what we call the FSD computer, that's now in production. So if you order any Model SRX or any Model three that has the full self driving package, you'll get the FSD computer. That'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 what, that Investor Day that we're having focused on autonomy, is that the car's currently being produced,

或者说当前量产的硬件，已具备完全自动驾驶的潜力。但'潜力'这个词很有意思，因为硬件本身是...

or the hardware currently being produced, is capable of full self driving. But capable is an interesting word because The hardware is.

没错，硬件基础已经具备。随着软件不断优化，车辆性能将显著提升，可靠性也会大幅增强，最终获得监管批准。本质上，现在购买特斯拉就是投资未来——你购买的我认为最颠覆性的是：如今购买特斯拉，你买到的是一项会增值的资产，而非贬值资产。

Yeah, the hardware. 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, a car today is an investment in the future. You're essentially buying car you're you're 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.

所以，这是一个非常重要的观点，因为如果硬件性能足够强大，升级硬件反而是最困难的部分。没错。通常如此。正是这样。那么剩下的就是软件问题了。

So, that's a really important statement there, because if hardware is capable enough, that's the hard thing to upgrade Yes. Usually. Exactly. So, then the rest is a software problem.

是的，软件确实没有边际成本。但是

Yes, software has no marginal cost, really. But

你对软件方面有什么直觉？要实现不仅安全，而且整体体验令人愉悦的程度，剩下的步骤有多困难？

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.

我认为人们在高速公路上已经非常享受自动驾驶了。特斯拉的自动驾驶功能彻底改变了高速公路上的生活质量体验。现在只需要将这种功能扩展到城市街道，增加交通信号灯识别能力，处理复杂交叉路口，以及能在迷宫般的停车场里自主驶出停车位并找到车主。然后它就能自动放下你并自己寻找停车位。

Well, I think people enjoy it very much on the highways. It's a total game changer for quality of life for using, you know, Tesla autopilot on the highways. So it's really just extending that functionality to city streets, adding in the traffic light 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. And and then if and then you can just it could just drop you off and find a parking spot by itself.

确实。就用户体验和实用性而言，停车场确实是个痛点——手动停车时总是让人烦躁不已，所以自动化在这方面能带来巨大好处。现在让我们引入一些人类视角来讨论。谈谈完全自动驾驶，如果看看目前路测的L4级车辆，比如Waymo等，它们只是技术上的自动驾驶。

Yeah. In terms of enjoyability and something that people would would actually find a lot of use from, the parking lot is a is a really, you know, it's rich of annoyance when you have to do it manually, so 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. If you look at the current level four vehicles being tested on a road, like Waymo and so on, they're only technically autonomous.

它们本质上还是L2级系统，只是设计理念不同，因为几乎所有情况下都有安全驾驶员在监控系统。你认为特斯拉的全自动驾驶在可预见的未来是否仍需要人类监督？即虽然它的驾驶能力足够强大，但仍需要像其他自动驾驶车辆那样由人类担任安全员的角色？

They're really level two systems with just a different design philosophy, because there's always a safety driver in almost all cases, and they're monitoring the system. Right. Do you see Tesla's full self driving as still, for a time to come, requiring supervision of the the human being. So 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?

我认为至少在未来六个月内仍需检测驾驶员手扶方向盘。从监管角度来说，核心问题是自动驾驶需要比人类安全多少才能取消监控？这确实存在讨论空间。你需要大量数据样本，才能以高统计置信度证明车辆比人类驾驶安全得多，而且人类监督不会实质影响安全性。可能需要达到比人类安全200%到300%的水平。

I think it'll it'll require detecting hands on wheel for at least six months or something like that from here. Really, it's a question of, like, from a regulatory standpoint, what how much safer than a person does Autopilot need to be for it to to be okay to not monitor the car? You know, and and this is a a debate that one can have. And then if you but you need you need a a large sample, a large amount of data so that 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. So it might need to be like two or 300% safer than a person.

那你如何证明这一点呢？

And how do you prove that?

每英里事故率。

Incidents per mile.

每英里事故率。对。所以就是碰撞和死亡事故。

Incidents per mile. Yeah. So crashes and fatalities.

死亡事故确实是个考量因素，但死亡案例数量太少，在统计上难以形成规模效应。不过碰撞事故就多得多——你知道的，碰撞事故远比死亡事故常见得多。所以你可以评估碰撞概率，然后是受伤概率，接着是永久性伤害概率，最后才是死亡概率。所有这些指标都必须比人类驾驶至少优秀200%才行。

So Fatalities would be a factor, but there there are just not enough fatalities to be statistically significant at scale. But there are enough crashes, You know, there are much far more crashes than there are fatalities. So you can assess what is the probability of of a crash. Then 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%.

你认为有能力与监管机构就这个话题

And you think there's the ability to have

进行建设性对话吗？毫无疑问，监管者会过度关注那些引发媒体报道的事件。这是个客观事实，而特斯拉总能引发大量报道。比如在美国，每年约有四万起汽车死亡事故，但特斯拉只要发生四起，获得的媒体报道可能就是其他车企的千倍。

a healthy discourse with the regulatory bodies on this topic? I mean, there's no question that regulators pay a disproportionate amount of attention to that which generates press. This is just an objective fact, and Tesla generates a lot of press. So the you know, in The United States, there's, I think, almost forty thousand automotive deaths per year. But if there are four in Tesla, they'll probably receive a thousand times more press than anyone else.

这种心理现象其实非常有趣。虽然时间有限无法深入探讨，但我必须和你谈谈人性层面。我和MIT团队最近发表了关于使用Autopilot时驾驶员功能警觉性的论文，这项研究从三年前Autopilot首次公开时就开始了，我们收集了驾驶员面部和身体的视频数据。看到你转发了摘要里的引文，我猜你至少浏览过这篇论文。

So 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. So 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. So I saw that you tweeted a quote from the abstract, so I can at least guess that you've glanced at it.

是啊，没错。

Yeah, right.

我能向你说明我们的发现吗？

Can I talk you through what we found?

当然。

Sure.

好的。根据我们收集的数据显示，驾驶员保持着功能性警觉状态——我们观察了18,000次自动驾驶系统脱离案例和18,900次标注记录，发现他们都能及时接管控制权。也就是说，他们始终关注路况准备接管。这确实与自动化警觉性研究文献中的普遍预测相悖。

Okay. 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. So 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.

现在的问题是，你认为这些结论是否适用于更广泛的人群？我们的样本只是小范围数据。有种批评观点认为，可能只有少数高度负责的驾驶员会因使用自动驾驶而导致警觉性下降加剧。

Now 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.

我...我觉得这些讨论很快就会被淘汰。系统正在飞速进步，警觉性问题很快就会变得无关紧要——当系统安全性远超人类时，人为干预对安全性的影响其实很有限，甚至可能是负面的。

I I think this is all really gonna be swept. I mean, the system's improving so much so fast that this is gonna be a moot point very soon, where vigilance is, like, if something's many times safer than a person, then adding a person does, the effect on safety is limited. And in fact, it could be negative.

这很有意思。所以即便部分人群会出现警觉性下降，也不会影响整体安全统计数据？

That's really interesting. So the the fact that a human may, some percent of the population, may exhibit a vigilance decrement will not affect overall statistics numbers of safety?

不。事实上，我认为它会非常、非常快地实现，可能今年年底前就能看到，但我敢说最迟明年一定会实现——人为干预反而会降低安全性。就像电梯的例子：过去需要电梯操作员，你无法独自操作电梯，更别说拉动操纵杆在楼层间移动。

No. In fact, I think it it will become very, 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 the purse having a human intervene will decrease safety. Decrease. It's it's like, imagine if you're in an elevator. Now it used to be that there were elevator operators, and and you you couldn't go in an elevator by yourself and and work the the lever to move between floors.

如今没人需要电梯操作员，因为自动停靠楼层的电梯比人工操作安全得多。事实上，让某人拿着操纵杆控制电梯在楼层间移动反而相当危险。

And 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.

这是个非常有力且有趣的观点。但从用户体验和安全角度，我特别关注算法层面的摄像头检测——不仅要感知人类，还要检测驾驶员视线、认知负荷和身体姿态。计算机视觉确实是个迷人的领域，但行业里很多人认为必须配备基于摄像头的驾驶员监控。你认为这种监控能带来实际益处吗？

So that's a 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's looking at, cognitive load, body pose. On the computer vision side, that's a fascinating problem, but do you and there's many in the industry who believe you have to have camera based driver monitoring. Do you think there could be benefit gained from driver monitoring?

如果系统可靠性处于或低于人类水平，驾驶员监控就有意义。但如果系统远优于人类，监控就没什么帮助。就像我说的——你会愿意让陌生人拿着操纵杆在电梯里控制楼层吗？我宁愿相信按钮。

If you have a system that's that's at or below human level reliability, then driver monitoring makes sense. But if if your system is dramatically better, more reliable than than a human, then driver monitoring is not does not help much. And like I said, you you just like as an you wouldn't want someone into like, you wouldn't want someone in the elevator. If you're you're in an elevator, do you really want someone with a big lever, some random person operating an elevator between floors that they could I wouldn't trust that. I'd rather have the buttons.

好的。你对系统改进速度很乐观，从你观察到的全自动驾驶计算机来看，进步是指数级的。早期有个与此相关的有趣设计选择——Autopilot的运作设计域（即其可激活的范围）。作为对比，我们研究的另一套系统是凯迪拉克的Super Cruise。

Okay. You're optimistic about the pace of improvement of the system, that from what you've seen with a full self driving car, computer. The rate of improvement is exponential. So one one of the other very interesting design choices early on that that connects to this is the operational design domain of Autopilot, so where Autopilot is able to be turned on. The so contrast, another vehicle system that we're studying is the Cadillac Super Cruise system.

就运作设计域而言，Super Cruise严格限定在特定类型的高速公路，经过充分测绘和测试，但范围远比特斯拉车辆窄。这就像注意力缺陷障碍（ADD）的比喻...对，这个类比不错。这种不同设计理念的决策背后有哪些利弊考量？

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. What's, there's It's like ADD. Yeah. That's good, that's a good line. What was the design decision, in that different philosophy of thinking where so there's pros and cons.

宽泛运作设计域的好处是：特斯拉驾驶员能更早探索系统边界，配合仪表盘显示，他们能逐步理解系统能力。弊端则是你允许驾驶员几乎在任何地方使用它。

What we see with a wide ODD is Tesla 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, so that's a benefit. The con is you're letting drivers use it basically anywhere.

总之，我能自信地检测车道。

Well, anyway that I can detect lanes with confidence.

当时是否存在具有挑战性的哲学设计决策？或者从一开始就是有意为之的？

Was 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?

说实话，我觉得让人类手动驾驶两吨重的致命机器简直疯狂。未来人们会难以置信，居然允许任何人随意驾驶这些两吨重的死亡机器。就像电梯一样，你本可以用那个操纵杆随意移动电梯，甚至停在楼层之间。

Well, I mean, I think it's, frankly, it's pretty crazy giving letting people drive a two ton death machine manually. That's crazy. Like, in the future, people will be like, I can't believe anyone was just allowed to drive one of these two ton death machines, and they just drive wherever they wanted. Just like elevators, you can just, like, move the elevator with that lever wherever you want. Can stop it halfway between floors if you want.

这确实很疯狂。所以未来会显得...

It's pretty crazy. So it it's gonna seem like

未来人们会认为开车是件疯狂的事。我有很多关于人类心理和行为的问题，这些问题将变得无关紧要。

a mad thing in the future that people were driving cars. So I have a bunch of questions about the human psychology, about behavior and so on. That would become mute.

它们并非完全无关紧要，因为...

They're not moot, totally moot. Because

你对AI系统有信心——不是盲信，而是相信硬件端和从数据中学习的深度学习方法会使其远比人类更安全。没错。最近有些黑客用对抗样本欺骗自动驾驶系统做出意外行为。嗯。

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. Yeah. Exactly. Recently, there are a few hackers who tricked Autopilot to act in unexpected ways with adversarial examples. Mhmm.

众所周知，神经网络系统对这些输入对抗样本的微小扰动非常敏感。你认为有可能为整个行业防御这类攻击吗？当然。那么能否详细解释一下这个答案背后的信心来源？

So we all know that neural network systems are very sensitive to minor disturbances to these adversarial examples on input. Do you think it's possible to defend against something like this for the broader for the industry? Sure. So Yeah. Can you elaborate on the on the confidence behind that answer?

其实神经网络本质上就是一堆矩阵运算。你需要成为真正理解神经网络的高手，基本上要逆向工程矩阵的构建方式，然后制造一个刚好让矩阵运算出现微小偏差的东西。但通过建立反负面识别机制来阻断这种攻击非常简单——就像系统发现疑似矩阵攻击时直接排除它，这很容易做到。

Well, you know, 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 like 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. But it's very easy to then block it block that by by having, basically anti negative recognition. It's like if the system sees something that looks like a matrix hack, exclude it. It's such an easy thing to do.

也就是说同时在有效数据和无效数据上学习。本质上就是在对抗样本上训练，从而能够识别并排除它们。

So learn both on the the valid data and the invalid data. So basically learn on the adversarial examples to exclude be able to exclude them.

没错。就像你既要明白什么是汽车，也要明确什么绝对不是汽车。你训练模型识别这是汽车而那不是，这是两种完全不同的概念。人们其实根本不了解神经网络。

Yeah. You, like, you basically wanna both know what is what is a car and what is definitely not a car. And you train for this is a car and this is definitely not a car. Those are two different things. People have no idea neural nets really.

他们可能以为神经网络只涉及网络而已。

They probably think neural nets involves, like, you know, net only.

那么，正如你所知，抛开特斯拉和自动驾驶不谈，当前深度学习方法在某些方面似乎仍与通用智能系统相去甚远？你认为现有方法能带我们实现通用智能，还是需要发明全新的思路？

So, 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?

我认为我们距离通用人工智能还缺少几个关键思路。但它会来得非常快，到时候我们还得考虑是否真有选择权。但令人惊讶的是人们分不清狭义AI（比如让车辆识别车道线并导航）与通用智能的区别——这就像烤面包机和电脑都是机器，但复杂程度天差地别。

I think we're missing a few key ideas for general intelligence, general artificial general intelligence. But it's gonna be upon us very quickly, and then we'll need to figure out what shall we do if we even have that choice. But but it's it's amazing how people can't differentiate between, say, the narrow AI that, you know, allows a car to figure out what a lane line is and and and, you know, and navigate streets versus general intelligence. Like, these are just very different things. Like, your toaster and your computer are both machines, but one's much more sophisticated than another.

你确信特斯拉能造出世界上最好的烤面包机。

You're confident with Tesla you can create the world's best toaster.

世界上最好的烤面包机，没错。世界上最好的自动驾驶系统。我是说真的。就目前而言，这就像比赛结束胜负已定。我不是...我的意思是，我不想显得自满或过度自信，但这确实就是当前局势最真实的写照。

The world's best toaster, yes. The world's best self driving. I'm yes. To to me, right now, this seems game set match. I don't I mean, that's how I don't want to be complacent or overconfident, but that's what it that is just literally what it how it appears right now.

我可能会错，但现状看起来特斯拉确实遥遥领先于所有竞争对手。

I could be wrong, but it appears to be the case that Tesla is vastly ahead of everyone.

你认为我们未来能否创造出像电影《她》里那样，让我们深爱且能真挚回应感情的AI系统？

Do 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?

我认为AI将能够让你深陷其中——

I think AI will be capable of convincing you to fall in

爱上它易如反掌。但这与我们人类——

love with it very well. And that's different than us

不同？要知道，这开始涉及形而上学问题了，比如情感和思想是否存在于物理层面之外的领域？或许存在，或许不存在。我不确定。但从物理学角度，我倾向于...我更习惯用物理思维来思考，毕竟那是我接受的主要学术训练。

humans? You 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. But but from a physics standpoint, I tend to think I tend to think of things, you know, like physics was my main sort of training.

从物理学角度来看，本质上，如果它以你无法辨别真假的方式爱你，那它就是真实的。

And from a physics standpoint, essentially, if it loves you in a way that you can't tell whether it's real or not, it is real.

这是物理学视角下的爱情观。

That's a physics view of love.

没错。如果无法证明其不存在，如果没有任何测试能让你分辨差异，那么实际上就没有差异。

Yeah. If there's no if you if you cannot just 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.

对。这类似于将我们的世界视为模拟。可能不存在能区分现实世界与模拟世界的测试，因此从物理学角度而言，它们可能就是同一回事。

Right. And it's similar to seeing our world as simulation. There may not be a test to tell the difference between what the real world is in simulation, and therefore, from a physics perspective, it might as well be the same thing.

是的。或许存在检测是否处于模拟中的方法——我不是说没有——但完全可以想象，模拟系统能够自我修正：当模拟中的实体发现检测方法时，它可以重启、暂停当前模拟、开启新模拟，或是采取其他多种纠错措施。所以也许当你...

Yes. And 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 once 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. So when maybe you

或者其他人创造出AGI系统时，你只能问她一个问题，你会问什么？

or somebody else creates an AGI system, and you get to ask her one question, what would that question be?

模拟之外是什么？

What's outside the simulation?

埃隆，非常感谢你今天接受访谈。非常愉快。

Elon, thank you so much for talking today. It was a pleasure.

好的，谢谢。

All right, thank you.