title
Jim Keller: Elon Musk and Tesla Autopilot | AI Podcast Clips

description
Full episode with Jim Keller (Feb 2020): https://www.youtube.com/watch?v=Nb2tebYAaOA Clips channel (Lex Clips): https://www.youtube.com/lexclips Main channel (Lex Fridman): https://www.youtube.com/lexfridman (more links below) Podcast full episodes playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOdP_8GztsuKi9nrraNbKKp4 Podcasts clips playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOeciFP3CBCIEElOJeitOr41 Podcast website: https://lexfridman.com/ai Podcast on Apple Podcasts (iTunes): https://apple.co/2lwqZIr Podcast on Spotify: https://spoti.fi/2nEwCF8 Podcast RSS: https://lexfridman.com/category/ai/feed/ Jim Keller is a legendary microprocessor engineer, having worked at AMD, Apple, Tesla, and now Intel. He's known for his work on the AMD K7, K8, K12 and Zen microarchitectures, Apple A4, A5 processors, and co-author of the specifications for the x86-64 instruction set and HyperTransport interconnect. Subscribe to this YouTube channel or connect on: - Twitter: https://twitter.com/lexfridman - LinkedIn: https://www.linkedin.com/in/lexfridman - Facebook: https://www.facebook.com/lexfridman - Instagram: https://www.instagram.com/lexfridman - Medium: https://medium.com/@lexfridman - Support on Patreon: https://www.patreon.com/lexfridman

detail
{'title': 'Jim Keller: Elon Musk and Tesla Autopilot | AI Podcast Clips', 'heatmap': [{'end': 634.198, 'start': 613.204, 'weight': 0.89}, {'end': 1028.521, 'start': 993.077, 'weight': 1}], 'summary': "Elon musk's vision for exponential improvement in vehicle autonomy, challenges in object detection and autonomous driving, potential of machines improving safety in autonomous vehicles, regulatory challenges, satisfaction and complexity of engineering, and advancements in technology particularly in deep learning and speech recognition are discussed in the video.", 'chapters': [{'end': 114.65, 'segs': [{'end': 114.65, 'src': 'embed', 'start': 49.064, 'weight': 0, 'content': [{'end': 56.568, 'text': 'Elon Musk believes that autopilot and vehicle autonomy, if you just look at that problem, can follow this kind of exponential improvement.', 'start': 49.064, 'duration': 7.504}, {'end': 61.851, 'text': "In terms of the how question that we're talking about, there's no reason why you can't.", 'start': 57.229, 'duration': 4.622}, {'end': 73.04, 'text': "What are your thoughts on this particular space of vehicle autonomy and you're a part of it and Elon Musk's and Tesla's vision for vehicle autonomy.", 'start': 62.431, 'duration': 10.609}, {'end': 80.967, 'text': 'Well, the computer you need to build was straightforward and you could argue well, does it need to be two times faster, or five times or 10 times?', 'start': 73.06, 'duration': 7.907}, {'end': 86.132, 'text': "But that's just a matter of time or price in the short run.", 'start': 82.268, 'duration': 3.864}, {'end': 87.433, 'text': "So that's not a big deal.", 'start': 86.172, 'duration': 1.261}, {'end': 90.055, 'text': "You don't have to be especially smart to drive a car.", 'start': 87.973, 'duration': 2.082}, {'end': 93.458, 'text': "so it's not like a super hard problem.", 'start': 91.016, 'duration': 2.442}, {'end': 100.622, 'text': 'i mean the big problem with safety is attention, which computers are really good at, not skills.', 'start': 93.458, 'duration': 7.164}, {'end': 103.023, 'text': 'well, let me push back on one.', 'start': 100.622, 'duration': 2.401}, {'end': 114.65, 'text': 'you see, everything you said is correct, but we as humans tend to, uh, tend to take for granted how how incredible our vision system is.', 'start': 103.023, 'duration': 11.627}], 'summary': 'Elon musk envisions exponential improvement in vehicle autonomy, citing the straightforward computer build and the attention-related safety concerns.', 'duration': 65.586, 'max_score': 49.064, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg849064.jpg'}], 'start': 1.804, 'title': "Elon musk's vision for autonomy", 'summary': "Highlights elon musk's vision for exponential improvement in vehicle autonomy, discussing the trend of cost in equipment, the development of autopilot, and the role of computers in ensuring safety and attention.", 'chapters': [{'end': 114.65, 'start': 1.804, 'title': "Elon musk's vision for autonomy", 'summary': "Highlights elon musk's vision for exponential improvement in vehicle autonomy, discussing the trend of cost in equipment, the development of autopilot, and the role of computers in ensuring safety and attention.", 'duration': 112.846, 'highlights': ['Elon Musk believes in the exponential improvement of vehicle autonomy and the trend of cost in equipment, emphasizing that little tweaks to the configuration generate something as opposed to figuring out how to build it.', 'The development of autopilot and the straightforward nature of building the required computer for vehicle autonomy, with the focus on time or price as a matter of concern.', 'The significance of attention in safety and the capability of computers in handling attention, and the discussion on the incredible nature of the human vision system.']}], 'duration': 112.846, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81804.jpg', 'highlights': ['Elon Musk emphasizes exponential improvement in vehicle autonomy and cost trends in equipment.', 'Development of autopilot and building required computer for vehicle autonomy are straightforward.', "Computers' capability in handling attention and the significance of attention in safety are discussed."]}, {'end': 394.69, 'segs': [{'end': 161.503, 'src': 'embed', 'start': 114.65, 'weight': 0, 'content': [{'end': 125.454, 'text': 'so you can drive a car with 20 50 vision And you can train a neural network to extract the distance of any object and the shape of any surface from a video and data.', 'start': 114.65, 'duration': 10.804}, {'end': 127.574, 'text': "It's really simple.", 'start': 126.394, 'duration': 1.18}, {'end': 128.914, 'text': "No, it's not simple.", 'start': 127.974, 'duration': 0.94}, {'end': 132.315, 'text': "That's a simple data problem.", 'start': 129.815, 'duration': 2.5}, {'end': 133.536, 'text': "It's not simple.", 'start': 132.675, 'duration': 0.861}, {'end': 143.278, 'text': "Because it's not just detecting objects, it's understanding the scene, and it's being able to do it in a way that doesn't make errors.", 'start': 134.756, 'duration': 8.522}, {'end': 152.06, 'text': "So the beautiful thing about the human vision system and our entire brain around the whole thing is we're able to fill in the gaps.", 'start': 144.358, 'duration': 7.702}, {'end': 155.401, 'text': "It's not just about perfectly detecting cars.", 'start': 153.281, 'duration': 2.12}, {'end': 157.622, 'text': "It's inferring the occluded cars.", 'start': 156.001, 'duration': 1.621}, {'end': 160.183, 'text': "It's trying to, it's understanding the physics.", 'start': 157.682, 'duration': 2.501}, {'end': 161.503, 'text': "I think that's mostly a data problem.", 'start': 160.203, 'duration': 1.3}], 'summary': 'Training neural network to understand scenes is a complex data problem.', 'duration': 46.853, 'max_score': 114.65, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8114650.jpg'}, {'end': 203.321, 'src': 'embed', 'start': 179.557, 'weight': 4, 'content': [{'end': 189.219, 'text': "Right So, you know, if you think that narrative is important to be able to successfully drive a car, then current autopilot systems can't do it.", 'start': 179.557, 'duration': 9.662}, {'end': 197.94, 'text': 'But if cars are ballistic things with tracks and probabilistic changes of speed and direction and roads are fixed and given by the way,', 'start': 189.359, 'duration': 8.581}, {'end': 199.401, 'text': "they don't change dynamically.", 'start': 197.94, 'duration': 1.461}, {'end': 203.321, 'text': 'Right You can map the world really thoroughly.', 'start': 199.421, 'duration': 3.9}], 'summary': "Current autopilot systems can't handle dynamic road changes, but thorough world mapping can.", 'duration': 23.764, 'max_score': 179.557, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8179557.jpg'}, {'end': 257.392, 'src': 'embed', 'start': 230.165, 'weight': 6, 'content': [{'end': 234.386, 'text': "They'll always remember there was a pothole in the road that humans keep forgetting about.", 'start': 230.165, 'duration': 4.221}, {'end': 240.628, 'text': "They'll remember that this set of roads has these weirdo lines on it that the computers figured out once.", 'start': 234.406, 'duration': 6.222}, {'end': 244.709, 'text': 'Especially if they get updates, so if somebody changes a given.', 'start': 240.648, 'duration': 4.061}, {'end': 253.131, 'text': 'The key to robots and stuff, somebody said, is to maximize the givens.', 'start': 246.669, 'duration': 6.462}, {'end': 257.392, 'text': 'Having a robot pick up this bottle cap is way easier if you put a red dot on the top.', 'start': 253.851, 'duration': 3.541}], 'summary': 'Robots use visual cues like red dots to maximize efficiency and ease tasks.', 'duration': 27.227, 'max_score': 230.165, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8230165.jpg'}, {'end': 316.089, 'src': 'embed', 'start': 292.479, 'weight': 3, 'content': [{'end': 300.13, 'text': 'Right So I tend to believe you do have to have theories, mental models of other people, especially with pedestrian cyclists, but also with other cars.', 'start': 292.479, 'duration': 7.651}, {'end': 306.564, 'text': 'So everything you said, is actually essential to driving.', 'start': 300.15, 'duration': 6.414}, {'end': 309.766, 'text': 'Driving is a lot more complicated than people realize, I think.', 'start': 306.604, 'duration': 3.162}, {'end': 316.089, 'text': "So sort of to push back slightly, but- So to cut into traffic, right? You can't just wait for a gap.", 'start': 310.266, 'duration': 5.823}], 'summary': 'Driving requires mental models of others; cutting into traffic requires assertiveness.', 'duration': 23.61, 'max_score': 292.479, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8292479.jpg'}], 'start': 114.65, 'title': 'Challenges in object detection and autonomous driving', 'summary': 'Discusses the complexity of training neural networks for object detection, emphasizing the need for improvement in data collection and computation. it also highlights the challenges of autonomous driving, emphasizing the potential for autonomous systems to excel in certain aspects and the intricate nature of human behavior in driving.', 'chapters': [{'end': 178.918, 'start': 114.65, 'title': 'Challenges in object detection', 'summary': 'Discusses the complexity of training neural networks for object detection, highlighting the importance of understanding scenes and the need for improvement in data collection and computation for better inference, including the challenges of filling in gaps and inferring occluded objects.', 'duration': 64.268, 'highlights': ['Training neural networks for object detection is not simple, as it involves understanding scenes and minimizing errors.', 'The human vision system has the ability to fill in gaps and infer occluded objects, showcasing the complexity of the task.', 'Improvement in data collection and computation is essential for better inference in object detection tasks.']}, {'end': 394.69, 'start': 179.557, 'title': 'Challenges of autonomous driving', 'summary': 'Discusses the challenges of autonomous driving, emphasizing the complexity of driving and the potential for autonomous systems to excel in certain aspects, while also highlighting the intricate nature of human behavior in driving.', 'duration': 215.133, 'highlights': ['Autonomous systems can excel in attention to detail and consistent memory, potentially surpassing humans in certain driving aspects.', "Maximizing the 'givens' is crucial for autonomous systems, simplifying tasks and enhancing efficiency.", 'The complexity of driving involves the need for theories and mental models of other entities on the road, indicating the intricate nature of human behavior in driving.', 'The potential for progress in autonomous driving is expected to disappoint in the short run but surprise in the long run, similar to the evolution of GPS technology.']}], 'duration': 280.04, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8114650.jpg', 'highlights': ['Improvement in data collection and computation is essential for better inference in object detection tasks.', 'The human vision system has the ability to fill in gaps and infer occluded objects, showcasing the complexity of the task.', 'Training neural networks for object detection is not simple, as it involves understanding scenes and minimizing errors.', 'Autonomous systems can excel in attention to detail and consistent memory, potentially surpassing humans in certain driving aspects.', 'The complexity of driving involves the need for theories and mental models of other entities on the road, indicating the intricate nature of human behavior in driving.', "Maximizing the 'givens' is crucial for autonomous systems, simplifying tasks and enhancing efficiency.", 'The potential for progress in autonomous driving is expected to disappoint in the short run but surprise in the long run, similar to the evolution of GPS technology.']}, {'end': 691.849, 'segs': [{'end': 435.733, 'src': 'embed', 'start': 396.531, 'weight': 0, 'content': [{'end': 403.153, 'text': "I'm hoping that there's a lot of things that humans aren't good at that machines are definitely good at like you said, attention and things like that.", 'start': 396.531, 'duration': 6.622}, {'end': 410.616, 'text': "Well, they'll be so much better that the overall picture of safety and autonomy will be obvious, that cars will be safer,", 'start': 403.333, 'duration': 7.283}, {'end': 412.336, 'text': "even if they're not as good at understanding.", 'start': 410.616, 'duration': 1.72}, {'end': 413.517, 'text': "I'm a big believer in safety.", 'start': 412.396, 'duration': 1.121}, {'end': 420.399, 'text': "I mean, there are already the current safety systems like cruise control that doesn't let you run into people and lane keeping.", 'start': 413.537, 'duration': 6.862}, {'end': 429.207, 'text': 'Right There are so many features that you just look at the Pareto of accidents and knocking off like 80% of them is super doable.', 'start': 420.419, 'duration': 8.788}, {'end': 435.733, 'text': 'Just to linger on the autopilot team and the efforts there.', 'start': 430.148, 'duration': 5.585}], 'summary': 'Machines excel in attention and safety, making cars significantly safer with features like autopilot, aiming to reduce 80% of accidents.', 'duration': 39.202, 'max_score': 396.531, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8396531.jpg'}, {'end': 634.198, 'src': 'heatmap', 'start': 613.204, 'weight': 0.89, 'content': [{'end': 622.954, 'text': 'because you say i know the math is always 8-bit integers into 32-bit accumulators and the operations are the subset of mathematical possibilities.', 'start': 613.204, 'duration': 9.75}, {'end': 634.198, 'text': "so AI accelerators have a claimed performance benefit over GPUs because in the narrow mass space you're nailing the algorithm.", 'start': 622.954, 'duration': 11.244}], 'summary': 'Ai accelerators claim 8-bit integers to 32-bit accumulators for performance benefit over gpus in narrow mass space.', 'duration': 20.994, 'max_score': 613.204, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8613204.jpg'}], 'start': 396.531, 'title': 'Autonomous vehicle safety and ai in automotive', 'summary': "Discusses the potential of machines improving safety in autonomous vehicles, regulatory challenges, and technological solutions. it also addresses the interest of regulators in autonomous driving safety, scenarios, injury data, and the challenges of building specialized computing systems for automotive applications, including elon musk's goal of achieving 10x safer autonomous systems.", 'chapters': [{'end': 484.872, 'start': 396.531, 'title': 'Autonomous vehicle safety', 'summary': 'Discusses the potential of machines improving safety in autonomous vehicles, highlighting the impact of safety systems and the public scrutiny on ensuring vehicle safety, with a focus on regulatory challenges and technological solutions.', 'duration': 88.341, 'highlights': ['The potential of machines improving safety in autonomous vehicles is emphasized, with the belief that cars will be safer due to the capabilities of machines, even if they are not as proficient in understanding, potentially reducing 80% of accidents.', 'The impact of public scrutiny and media attention on safety for autonomous vehicles is highlighted, reflecting the intense pressure faced by the industry to ensure vehicle safety, with a sense of fairness and reasonableness about the expectations.', 'The regulatory challenges and concerns about technology solutions for autonomous vehicles, such as modern brake systems, are discussed, emphasizing the need for regulations to accommodate technological advancements and not be restrictive to specific solutions like hydraulic brakes.']}, {'end': 691.849, 'start': 485.613, 'title': 'Autonomous driving and ai in automotive', 'summary': "Discusses the interest of regulators in autonomous driving safety, with a focus on scenarios and injury data, elon musk's goal of achieving 10x safer autonomous systems, and the challenges of building specialized computing systems for automotive applications.", 'duration': 206.236, 'highlights': ['Investigation of use cases by regulators and utilization of injury data to focus on safety scenarios', "Elon Musk's goal of achieving 10X safer autonomous systems compared to human drivers", 'Challenges of building specialized computing systems for automotive applications, including concerns about algorithm changes and the balance between acceleration and over-specialization']}], 'duration': 295.318, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8396531.jpg', 'highlights': ['Cars will be safer due to machines, potentially reducing 80% of accidents.', 'Intense pressure faced by the industry to ensure vehicle safety from public scrutiny and media attention.', 'Regulations need to accommodate technological advancements for autonomous vehicles.']}, {'end': 1239.974, 'segs': [{'end': 754.324, 'src': 'embed', 'start': 692.949, 'weight': 1, 'content': [{'end': 696.03, 'text': 'And to hit that, you had to think about the system design.', 'start': 692.949, 'duration': 3.081}, {'end': 696.89, 'text': "That's complicated.", 'start': 696.05, 'duration': 0.84}, {'end': 697.39, 'text': "It's fun.", 'start': 697.05, 'duration': 0.34}, {'end': 699.811, 'text': "It's like, it's Craftsman's work.", 'start': 697.83, 'duration': 1.981}, {'end': 704.394, 'text': 'Like, you know, a violin maker, right? You can say, Stradivarius is this incredible thing.', 'start': 700.271, 'duration': 4.123}, {'end': 705.675, 'text': 'The musicians are incredible.', 'start': 704.434, 'duration': 1.241}, {'end': 713.62, 'text': 'But the guy making the violin, you know, picked wood and sanded it, and then he cut it, you know, and he glued it, you know,', 'start': 706.175, 'duration': 7.445}, {'end': 719.084, 'text': "and he waited for the right day so that when he put the finish on it, it didn't, you know, do something dumb.", 'start': 713.62, 'duration': 5.464}, {'end': 721.065, 'text': "That's craftsman's work, right?", 'start': 719.324, 'duration': 1.741}, {'end': 729.031, 'text': "You may be a genius craftsman because you have the best techniques and you discover a new one, but most engineering is craftsman's work.", 'start': 721.566, 'duration': 7.465}, {'end': 731.112, 'text': 'And humans really like to do that.', 'start': 729.631, 'duration': 1.481}, {'end': 733.053, 'text': 'You know the expression? Smart humans.', 'start': 731.972, 'duration': 1.081}, {'end': 734.233, 'text': 'No, everybody.', 'start': 733.713, 'duration': 0.52}, {'end': 734.933, 'text': 'All humans.', 'start': 734.513, 'duration': 0.42}, {'end': 735.954, 'text': "I don't know.", 'start': 735.033, 'duration': 0.921}, {'end': 737.995, 'text': 'I dug ditches when I was in college.', 'start': 736.094, 'duration': 1.901}, {'end': 738.855, 'text': 'I got really good at it.', 'start': 738.035, 'duration': 0.82}, {'end': 741.456, 'text': 'Satisfying Yeah.', 'start': 739.155, 'duration': 2.301}, {'end': 743.077, 'text': 'Digging ditches is also craftsman-worthy.', 'start': 741.496, 'duration': 1.581}, {'end': 744.078, 'text': 'Yeah, of course.', 'start': 743.117, 'duration': 0.961}, {'end': 747.919, 'text': "So there's an expression called complex mastery behavior.", 'start': 744.658, 'duration': 3.261}, {'end': 751.041, 'text': "So when you're learning something, that's fun because you're learning something.", 'start': 748.6, 'duration': 2.441}, {'end': 754.324, 'text': "When you do something and throw it in simple, it's not that satisfying.", 'start': 751.781, 'duration': 2.543}], 'summary': "Engineering is like craftsman's work, requiring complex mastery behavior and satisfaction in the process.", 'duration': 61.375, 'max_score': 692.949, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8692949.jpg'}, {'end': 843.793, 'src': 'embed', 'start': 814.891, 'weight': 0, 'content': [{'end': 816.372, 'text': "There's more invention than that.", 'start': 814.891, 'duration': 1.481}, {'end': 821.835, 'text': 'But the rejection to practice.', 'start': 820.094, 'duration': 1.741}, {'end': 824.377, 'text': 'once you pick the architecture, you look inside and what do you see?', 'start': 821.835, 'duration': 2.542}, {'end': 828.561, 'text': 'Adders and multipliers and memories, and you know the basics.', 'start': 824.858, 'duration': 3.703}, {'end': 838.829, 'text': 'So computers is always this weird set of abstraction layers of ideas and thinking that reduction to practice is transistors and wires,', 'start': 828.861, 'duration': 9.968}, {'end': 841.131, 'text': 'and you know pretty basic stuff.', 'start': 838.829, 'duration': 2.302}, {'end': 843.793, 'text': "And that's an interesting phenomenon.", 'start': 841.591, 'duration': 2.202}], 'summary': 'Computers involve abstraction layers, such as transistors and wires, leading to invention and reduction to practice.', 'duration': 28.902, 'max_score': 814.891, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8814891.jpg'}, {'end': 940.445, 'src': 'embed', 'start': 912.989, 'weight': 3, 'content': [{'end': 915.63, 'text': 'Putting together, working on a factory.', 'start': 912.989, 'duration': 2.641}, {'end': 918.771, 'text': 'Two smart people can disagree.', 'start': 916.33, 'duration': 2.441}, {'end': 921.532, 'text': 'Yay I think driving a car.', 'start': 919.151, 'duration': 2.381}, {'end': 924.994, 'text': "We'll get you in the factory someday and then we'll see how you do.", 'start': 922.193, 'duration': 2.801}, {'end': 927.195, 'text': 'No, not for us humans driving a car is easy.', 'start': 925.014, 'duration': 2.181}, {'end': 931.657, 'text': "I'm saying building a machine that drives a car is not easy.", 'start': 927.235, 'duration': 4.422}, {'end': 938.483, 'text': "Okay, driving a car is easy for humans because we've been evolving for billions of years.", 'start': 932.257, 'duration': 6.226}, {'end': 939.184, 'text': 'Drive cars.', 'start': 938.503, 'duration': 0.681}, {'end': 940.445, 'text': 'Yeah, I noticed that.', 'start': 939.444, 'duration': 1.001}], 'summary': 'Debating ease of driving a car and building a machine to drive, with 2 smart people.', 'duration': 27.456, 'max_score': 912.989, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8912989.jpg'}, {'end': 1031.643, 'src': 'heatmap', 'start': 980.544, 'weight': 6, 'content': [{'end': 989.192, 'text': 'have taken away from your time working with Elon Musk, working at Tesla, which is known to be a place of chaos, innovation,', 'start': 980.544, 'duration': 8.648}, {'end': 991.274, 'text': 'craftsmanship and all those things?', 'start': 990.353, 'duration': 0.921}, {'end': 992.396, 'text': 'I really liked the way he thought.', 'start': 991.294, 'duration': 1.102}, {'end': 997.743, 'text': 'Like you think you have an understanding about what first principles of something is,', 'start': 993.077, 'duration': 4.666}, {'end': 1001.288, 'text': "and then you talk to Elon about it and you didn't scratch the surface.", 'start': 997.743, 'duration': 3.545}, {'end': 1006.936, 'text': 'You know, he has a deep belief that no matter what you do is a local maximum.', 'start': 1002.49, 'duration': 4.446}, {'end': 1011.145, 'text': 'And I had a friend, he invented a better electric motor.', 'start': 1008.883, 'duration': 2.262}, {'end': 1014.308, 'text': 'And it was like a lot better than what we were using.', 'start': 1012.086, 'duration': 2.222}, {'end': 1019.554, 'text': "And one day he came by, he said, you know, I'm a little disappointed because, you know, this is really great.", 'start': 1014.709, 'duration': 4.845}, {'end': 1020.915, 'text': "And you didn't seem that impressed.", 'start': 1019.674, 'duration': 1.241}, {'end': 1028.521, 'text': 'And I said, you know, when the super intelligent aliens come, are they going to be looking for you? Like, where is he? The guy who built the motor.', 'start': 1020.995, 'duration': 7.526}, {'end': 1031.643, 'text': 'Probably not.', 'start': 1030.061, 'duration': 1.582}], 'summary': "Elon musk's perspective on innovation and first principles challenges conventional thinking, emphasizing the pursuit of global maximums.", 'duration': 51.099, 'max_score': 980.544, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8980544.jpg'}, {'end': 1126.131, 'src': 'embed', 'start': 1098.483, 'weight': 5, 'content': [{'end': 1109.052, 'text': "Do people look back on it and say, boy, I'm really happy I had that experience to go take apart that many layers of assumptions.", 'start': 1098.483, 'duration': 10.569}, {'end': 1112.435, 'text': 'Sometimes super fun, sometimes painful.', 'start': 1110.633, 'duration': 1.802}, {'end': 1117.861, 'text': 'So it could be emotionally and intellectually painful, that whole process of just stripping away assumptions.', 'start': 1113.056, 'duration': 4.805}, {'end': 1122.786, 'text': 'Yeah, imagine 99% of your thought process is protecting your self-conception.', 'start': 1118.581, 'duration': 4.205}, {'end': 1126.131, 'text': "And 98% of that's wrong.", 'start': 1124.309, 'duration': 1.822}], 'summary': 'Challenging assumptions can be emotionally and intellectually painful, with 98% of self-conceptions being wrong.', 'duration': 27.648, 'max_score': 1098.483, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81098483.jpg'}], 'start': 692.949, 'title': 'Engineering and innovation', 'summary': "Discusses the satisfaction and complexity of engineering as craftsman's work, highlighting challenges in mastering intricate skills, such as in autonomous driving technology, and the innovative mindset of elon musk in factory work. it also delves into reducing complex ideas to practical elements in computer engineering and the impact of reading and applying knowledge from books in improving one's abilities.", 'chapters': [{'end': 841.131, 'start': 692.949, 'title': "The joy of craftsman's work in engineering", 'summary': "Discusses the satisfaction and complexity of engineering as craftsman's work, highlighting the fun and challenges of mastering intricate skills, such as in autonomous driving technology, and the reduction of complex ideas to practical elements in computer engineering.", 'duration': 148.182, 'highlights': ["Autonomous driving technology requires building a resource-constrained computer that needs to be cheap enough to put in every car, which essentially boils down to craftsman's work. Craftsman's work in engineering is exemplified in the challenges of building a resource-constrained computer for autonomous driving technology, where the computer must be affordable for every car.", 'The satisfaction in engineering comes from mastering complicated skills and making thoughtful decisions and trade-offs, such as determining the number of camera inputs or solving safety concerns. The satisfaction in engineering stems from mastering complicated skills and making thoughtful decisions and trade-offs, such as determining the number of camera inputs and addressing safety concerns.', 'The reduction to practice in computer engineering involves working with basic elements like transistors, wires, adders, multipliers, and memories, despite the complex layers of abstraction in computer design. In computer engineering, the reduction to practice involves working with basic elements like transistors, wires, adders, multipliers, and memories, despite the complex layers of abstraction in computer design.']}, {'end': 1239.974, 'start': 841.591, 'title': 'Factory work and innovation', 'summary': "Discusses the complexity and satisfaction of factory work, the innovative mindset of elon musk, the process of stripping away assumptions, and the impact of reading and applying knowledge from books in improving one's abilities.", 'duration': 398.383, 'highlights': ["Elon Musk's deep belief in striving beyond local maximum and his innovative mindset are emphasized. Elon Musk's approach to always seeking beyond the local maximum and his ability to innovate are highlighted, showcasing the depth of his thinking and his commitment to continuous improvement.", 'The complex and satisfying nature of factory work is described, emphasizing the skill and satisfaction derived from assembling cars on the moving assembly line. The complexity and satisfaction of factory work, particularly in assembling cars on a moving assembly line, are emphasized, showcasing the intricate skill and gratification involved in such work.', "The impact of reading and applying knowledge from books in improving one's abilities is discussed, highlighting the effectiveness of learning from books and applying the knowledge in practical scenarios. The impact of reading and applying knowledge from books in improving one's abilities is discussed, emphasizing the effectiveness of learning from books and applying the knowledge in practical scenarios, as exemplified by the speaker's experience.", 'The process of stripping away assumptions and its emotional and intellectual challenges are explored, showcasing the difficulty and significance of removing assumptions to enhance thought processes. The process of stripping away assumptions and its emotional and intellectual challenges are explored, highlighting the difficulty and significance of removing assumptions to enhance thought processes, with an emphasis on the emotional and intellectual challenges involved.']}], 'duration': 547.025, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg8692949.jpg', 'highlights': ["Autonomous driving technology requires building a resource-constrained computer that needs to be cheap enough to put in every car, essentially craftsman's work.", 'The satisfaction in engineering comes from mastering complicated skills and making thoughtful decisions and trade-offs.', 'The reduction to practice in computer engineering involves working with basic elements like transistors, wires, adders, multipliers, and memories, despite the complex layers of abstraction in computer design.', "Elon Musk's deep belief in striving beyond local maximum and his innovative mindset are emphasized.", 'The complex and satisfying nature of factory work is described, emphasizing the skill and satisfaction derived from assembling cars on the moving assembly line.', "The impact of reading and applying knowledge from books in improving one's abilities is discussed, highlighting the effectiveness of learning from books and applying the knowledge in practical scenarios.", 'The process of stripping away assumptions and its emotional and intellectual challenges are explored, showcasing the difficulty and significance of removing assumptions to enhance thought processes.']}, {'end': 1440.009, 'segs': [{'end': 1298.808, 'src': 'embed', 'start': 1270.219, 'weight': 3, 'content': [{'end': 1272.72, 'text': 'Because it is emotionally challenging.', 'start': 1270.219, 'duration': 2.501}, {'end': 1274.92, 'text': "Well, it's hard to just keep it front and center,", 'start': 1272.86, 'duration': 2.06}, {'end': 1281.342, 'text': 'because you know you operate on so many levels all the time and you know getting this done takes priority.', 'start': 1274.92, 'duration': 6.422}, {'end': 1287.284, 'text': 'or you know being happy takes priority, or you know screwing around takes priority.', 'start': 1281.342, 'duration': 5.942}, {'end': 1289.885, 'text': 'Like how you go through life is complicated.', 'start': 1288.125, 'duration': 1.76}, {'end': 1294.046, 'text': 'Yeah And then you remember, oh, yeah, I could really think first principles.', 'start': 1289.905, 'duration': 4.141}, {'end': 1296.807, 'text': "Oh, shit, that's tiring, you know.", 'start': 1294.247, 'duration': 2.56}, {'end': 1298.808, 'text': "But you do for a while and that's kind of cool.", 'start': 1297.287, 'duration': 1.521}], 'summary': 'Balancing priorities is emotionally challenging. thinking from first principles is tiring, but can be rewarding.', 'duration': 28.589, 'max_score': 1270.219, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81270219.jpg'}, {'end': 1343.29, 'src': 'embed', 'start': 1316.415, 'weight': 2, 'content': [{'end': 1321.558, 'text': 'so one, two, three, five, ten years as opposed to a century?', 'start': 1316.415, 'duration': 5.143}, {'end': 1325.22, 'text': 'yeah, definitely, Just to linger on it a little longer.', 'start': 1321.558, 'duration': 3.662}, {'end': 1327.742, 'text': "where's the confidence coming from??", 'start': 1325.22, 'duration': 2.522}, {'end': 1333.445, 'text': 'Is it the fundamentals of the problem, the fundamentals of building the hardware and the software?', 'start': 1327.822, 'duration': 5.623}, {'end': 1343.29, 'text': 'As a computational problem, understanding ballistics, roles, topography, it seems pretty solvable.', 'start': 1334.145, 'duration': 9.145}], 'summary': 'The confidence in solving computational problems within 10 years is based on the fundamentals of hardware and software, making it seem pretty solvable.', 'duration': 26.875, 'max_score': 1316.415, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81316415.jpg'}, {'end': 1430.438, 'src': 'embed', 'start': 1383.429, 'weight': 0, 'content': [{'end': 1393.173, 'text': "I'll tell Elon once again, as I did before, that when you add human beings into the picture, it's no longer a ballistics problem.", 'start': 1383.429, 'duration': 9.744}, {'end': 1397.915, 'text': "It's something more complicated, but I could be very well proven wrong.", 'start': 1393.493, 'duration': 4.422}, {'end': 1400.396, 'text': 'And cars are highly damped in terms of rate of change.', 'start': 1397.935, 'duration': 2.461}, {'end': 1405.319, 'text': 'like the steering, and the steering system is really slow compared to a computer.', 'start': 1401.596, 'duration': 3.723}, {'end': 1408.781, 'text': 'the acceleration of the acceleration is really slow.', 'start': 1405.319, 'duration': 3.462}, {'end': 1414.366, 'text': "yeah, on a certain time scale, on a ballistics time scale, but human behavior, i don't know it.", 'start': 1408.781, 'duration': 5.585}, {'end': 1420.67, 'text': "i i shouldn't say the beans are really slow too weirdly, we operate, you know, half a second behind reality.", 'start': 1414.366, 'duration': 6.304}, {'end': 1423.032, 'text': "i'll be really understands that one either.", 'start': 1420.67, 'duration': 2.362}, {'end': 1424.313, 'text': "it's pretty funny.", 'start': 1423.032, 'duration': 1.281}, {'end': 1428.576, 'text': "yeah, yeah, so yeah, we'll be.", 'start': 1424.313, 'duration': 4.263}, {'end': 1430.438, 'text': 'we very well could be surprised.', 'start': 1428.576, 'duration': 1.862}], 'summary': 'Human behavior adds complexity to ballistics problems, but could be proven wrong, as cars are highly damped and operate half a second behind reality.', 'duration': 47.009, 'max_score': 1383.429, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81383429.jpg'}], 'start': 1241.243, 'title': 'Advancements in autonomous driving', 'summary': 'Delves into the emotional challenge of consistently applying first principles thinking and the advancements in technology, particularly deep learning, in solving problems related to speech recognition and autonomous driving, emphasizing increasing data gathering, computation, and algorithm understanding.', 'chapters': [{'end': 1343.29, 'start': 1241.243, 'title': 'Challenging first assumptions', 'summary': 'Discusses the challenge of consistently applying first principles thinking, which can be emotionally challenging, yet it is essential for solving complex problems, such as autonomous driving, with confidence in its solvability.', 'duration': 102.047, 'highlights': ['Consistently applying first principles thinking is emotionally challenging, as it requires circling back to the observation despite competing priorities.', 'The complexity of life makes it hard to maintain a focus on first principles thinking, leading to periods of forgetting its importance.', 'Autonomous driving is seen as solvable within a timeline of one to ten years, driven by confidence in the fundamentals of the computational problem and understanding ballistics, roles, and topography.']}, {'end': 1440.009, 'start': 1344.192, 'title': 'Advancements in technology and autonomous driving', 'summary': 'Discusses the advancements in technology, particularly deep learning, in solving problems related to speech recognition and autonomous driving, with an emphasis on the increasing data gathering, computation, and algorithm understanding, leading to potential pleasant surprises in the future.', 'duration': 95.817, 'highlights': ['Advancements in technology, particularly deep learning, have significantly improved speech recognition and autonomous driving, with increasing data gathering, computation, and algorithm understanding.', 'The rate of improvement in compute and software, as well as hardware, is expected to lead to pleasant surprises in various aspects.', 'Human behavior introduces complexity into autonomous driving, as it is not solely a ballistics problem, and the differences in rate of change between human and computer behavior pose challenges.']}], 'duration': 198.766, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/ymcOLL2qEg8/pics/ymcOLL2qEg81241243.jpg', 'highlights': ['Advancements in technology, particularly deep learning, have significantly improved speech recognition and autonomous driving, with increasing data gathering, computation, and algorithm understanding.', 'Autonomous driving is seen as solvable within a timeline of one to ten years, driven by confidence in the fundamentals of the computational problem and understanding ballistics, roles, and topography.', 'The rate of improvement in compute and software, as well as hardware, is expected to lead to pleasant surprises in various aspects.', 'Human behavior introduces complexity into autonomous driving, as it is not solely a ballistics problem, and the differences in rate of change between human and computer behavior pose challenges.', 'Consistently applying first principles thinking is emotionally challenging, as it requires circling back to the observation despite competing priorities.', 'The complexity of life makes it hard to maintain a focus on first principles thinking, leading to periods of forgetting its importance.']}], 'highlights': ['Cars will be safer due to machines, potentially reducing 80% of accidents.', 'Advancements in technology, particularly deep learning, have significantly improved speech recognition and autonomous driving, with increasing data gathering, computation, and algorithm understanding.', 'Elon Musk emphasizes exponential improvement in vehicle autonomy and cost trends in equipment.', 'Autonomous driving is seen as solvable within a timeline of one to ten years, driven by confidence in the fundamentals of the computational problem and understanding ballistics, roles, and topography.', 'The potential for progress in autonomous driving is expected to disappoint in the short run but surprise in the long run, similar to the evolution of GPS technology.', 'Regulations need to accommodate technological advancements for autonomous vehicles.', 'The complexity of driving involves the need for theories and mental models of other entities on the road, indicating the intricate nature of human behavior in driving.', 'The satisfaction in engineering comes from mastering complicated skills and making thoughtful decisions and trade-offs.', 'The process of stripping away assumptions and its emotional and intellectual challenges are explored, showcasing the difficulty and significance of removing assumptions to enhance thought processes.', "The impact of reading and applying knowledge from books in improving one's abilities is discussed, highlighting the effectiveness of learning from books and applying the knowledge in practical scenarios."]}