Karpathy on AI Compute / CEO's Admit Tesla Is Way Ahead / Elon's Lawyer Goes Off ⚡️

Welcome to Electrified, it's your host, Dileunumis, quick shout out to Gerard V. Thank you for using my Tesla Refer link and congratulations on your new Tesla. It's been a while, so here's some video footage as a reminder that there's a statistical category accidents avoided that will never be reported on when it comes to FSD, but that doesn't mean it's not important. If you scrub through this slowly, you can see the van does put its brakes on right here, so it does realize what it's about to do and stops, but had there been a driver not paying attention that just carried on perfectly straight, there may have been a collision. So just remember when the FSD safety data comes out to bring up the talking point accidents avoided. And here's a pretty impressive clip of 12.6.1. So FSD slowly peeks around the corner right here, sees the car coming and waits, and then proceeds with the opening.

欢迎来到《Electrified》节目，我是你的主持人Dileunumis。先感谢Gerard V.使用我的Tesla推荐链接，并恭喜你购买了新特斯拉。好久不见，所以这里有些视频片段提醒大家，其实在自动驾驶（FSD）技术中有一个叫做“避免事故”的统计类别，这种统计数据从不会被报道，但这并不意味着它不重要。如果你慢慢看这个视频，可以看到面包车在这里刹车，所以它意识到了自己即将做什么并停下了，但如果司机没注意继续直行的话，可能就会发生碰撞。因此，当FSD安全数据发布时，请记得谈论“避免事故”这个话题。这里有一个很不错的12.6.1版本片段，FSD在这里慢慢探出拐角，看到有车驶来，然后耐心等待，待道路开放后再继续行驶。

And then this clip from Tesla Chan of FSD13.2.2 on the Cybertruck navigating out of the Disneyland parking lot, which is rather complex at certain points. So, let's get started. Ray Furtesla shared a video interview of Lejoon, who is Xiaomi CEO, saying that Xiaomi is still 5 to 10 years behind Tesla, who is still the benchmark for EVs. He also went on to say how Xiaomi was using the first principle thinking method as they learned from Tesla, what I like to call the Tesla ripple effect. Ray shared another video of a former Huawei autonomous driving exec, Shuxing, talking about FSD. He said, If you insist on having me choose a benchmark today, frankly, I would still choose FSD because I truly believe the overall quality of the American team and their spirit of innovation are genuinely admirable.

然后，这段来自特斯拉频道的视频展示了FSD13.2.2如何在Cybertruck上导航驶出迪士尼乐园停车场，在某些地方相当复杂。让我们开始吧。Ray Furtesla分享了一段对小米CEO雷军的采访视频，雷军表示，小米在某些方面仍落后于特斯拉5到10年，因为特斯拉仍然是电动汽车的标杆。他还表示，小米正在借鉴特斯拉使用的“第一性原理”思维方法，这就是我称之为“特斯拉波及效应”的原因。Ray还分享了另一段视频，视频中前华为自动驾驶高管舒星谈到了FSD。他说：“如果你坚持让我今天选择一个标杆，坦白说，我仍然会选择FSD，因为我真的认为美国团队的整体素质和他们的创新精神值得钦佩。”

After trying it out myself, I can see that they have many unique strengths in terms of tech, and in some areas they're even ahead by a significant margin. It's okay to acknowledge it. Maybe it's just me, but in 2025, I highly value honesty and candor, especially in the business environment. And you know, for a while it was unspoken, how much the Chinese auto and AI industries really admired Tesla and took them seriously, learning as much as they could from what Tesla was doing. To have a foreign branded higher price vehicle as the number one seller in 2024 in China confirms that in the most competitive EV market in the world, consumers love what Tesla has to offer.

在亲自尝试后，我发现他们在技术方面有很多独特优势，在某些领域甚至遥遥领先。承认这一点没什么不好。也许只是我个人的看法，但在2025年，我非常重视诚实和坦率，尤其是在商业环境中。而且你知道，有一段时间，大家都默契地知道中国的汽车和人工智能行业非常欣赏特斯拉，认真研究特斯拉的做法。从2024年在中国市场，一款高价的外国品牌车辆成为销量第一来看，说明在这个全球竞争最激烈的电动车市场，消费者确实喜欢特斯拉所提供的产品。

And all of these Chinese auto companies talking about moving to end to end neural nets has only happened after Tesla went public with version 12. Say what you will about the Chinese, but they have tried to copy Tesla closely and are now reaping those rewards, but it's great to hear executive share how far ahead Tesla really is. Unlike competing CEOs here in America that ducked the question or tried to hype up their own A-DAS programs, rather than just tipping the cap to how Tesla has pushed so many industries forward, nearly single-handedly. Ray for Tesla also shared that former Huawei exec said Chinese competitors can learn from Tesla but are limited by training compute. His current company takes a hybrid approach to tackle autonomy, yes, end to end, but still using a rule-based approach as well.

所有这些中国汽车公司讨论转向端到端神经网络，都是在特斯拉推出第12版本后才发生的。不管你怎样评价中国，他们确实在密切模仿特斯拉，现在正在享受这种模仿带来的成果。不过，听到高管们分享特斯拉实际领先的程度还是令人振奋。在美国，竞争对手的CEO们要么避开这个问题，要么只是夸大自己公司的驾驶辅助系统，而不是承认特斯拉在推动多个行业发展上几乎是独力的。Ray for Tesla还分享说，一位前华为高管表示，中国的竞争者可以向特斯拉学习，但在计算能力上受到限制。他目前的公司采用了一种混合的方法来解决自动驾驶问题，是的，是端到端，但也仍在使用基于规则的方法。

Why are some of these Chinese companies limited when it comes to compute? Well, at least in part because of the US and their implementation of those stringent export controls for the most advanced AI chips. Chinese companies are restricted from buying leading-edge chips like Nvidia's A and H100s. I know sometimes the topics I share on the channel may be a bit more boring like AI regulations, but if you follow the breadcrumbs and how these things actually play out in the long term, you begin to see that well, they do actually have somewhat profound implications.

为什么一些中国公司在计算能力方面受限？这至少部分是因为美国对最先进的AI芯片实行了严格的出口管制。中国公司无法购买像Nvidia的A和H100这样的尖端芯片。我知道有时候我在频道上分享的话题可能像AI法规这样有点无聊，但如果你顺藤摸瓜，看看这些事情在长期如何发展，你会开始看到它们确实有一些深远的影响。

So not only are most of these Chinese automakers admitting their years behind Tesla when it comes to FSD and other technology, but also their chances of catching Tesla are hindered because their compute constrained really with no end in sight. That's why when people like Gary Black say that when Tesla solves for unsupervised, then pretty much everybody else will have it within two to three years. I'm just like, how are you coming to this conclusion? One other point I want to make here requires a bit of reading between the lines, but this deep seek company is a Chinese AI company that's producing some of the frontier or leading edge LLMs and they're largely known for open sourcing their work. Andre Carpathi highlighted how deep seeks version 3 LLM only used 2048 GPUs for two months and had the performance of other clusters using between 16,000 and 100,000 GPUs.

因此，这些中国汽车制造商不仅承认在全自动驾驶（FSD）和其他技术上落后于特斯拉多年，而且因为计算能力有限且看不到尽头，他们赶超特斯拉的机会也受到了阻碍。这就是为什么当像Gary Black这样的人说特斯拉解决了无人监督驾驶后，其他厂商在两到三年内也能实现时，我感到很困惑，真不明白他怎么得出这个结论的。我要提到的另一个点需要从字里行间分析，但这个深势科技公司是一家中国人工智能公司，他们在前沿或领先的大型语言模型（LLM）方面有贡献，并且以开源其作品而闻名。Andre Carpathi指出，深势科技的第三代LLM仅用了2048个GPU工作两个月，其性能相当于使用16000到100000个GPU的其他集群。

Andre compared it to llama 3 in its 405 billion parameter model using 30.8 million GPU hours, while deep seek version 3 only used 2.8 million GPU hours or around 11 times less compute. So my point here as I've been trying to make is that this AI game is far more than just raw compute and raw data. I've seen plenty of people out there saying, well if these other companies have a lot more data, whether it's synthetic or real, and then they just buy a bunch of compute, they're going to be able to catch up to Tesla no problem. But as deep seek has proved, there's a ton of optimization that can happen just from how you decide to handle the data, the quality of that data, the algorithms, the weight of the neural nets and how you adjust all of those, there's an entire world happening here that most people just aren't aware of.

安德烈将其与Llama 3进行比较，Llama 3的模型有4050亿个参数，花费了3080万小时的GPU运行时间，而Deep Seek 版本3仅使用了280万小时的GPU时间，计算量减少了约11倍。因此，我想表达的观点是，人工智能领域远不止是简单的计算能力和数据量。我看到很多人说，如果其他公司拥有更多的数据，无论是合成的还是实际的，再加上大量的计算能力，就能轻松赶上特斯拉。但正如Deep Seek所证明的，通过决定如何处理数据、数据的质量、算法、神经网络的权重以及如何调整这些方面进行优化，还有很多工作可以完成，这是大多数人未曾意识到的一个庞大的世界。

So not only does Tesla have the best data set, one of the largest compute clusters in cortex, and some of the best AI talent in the world, but they've been optimizing their neural nets and algos for years, continually removing dead weight as they make advances which is what Eun Tassai was referring to when he compared FSD13 to version 3 of the SpaceX Raptor engine. That's just my long-winded way of trying to make the point that it is not just about how much compute and how much data, yes that's a big part of the story, but there's far more nuance and a lot more to the story than just those two factors. So even a company like Waymo who has access to deep pockets isn't going to be able to just throw a bunch of money at this problem and effectively have it solve itself.

所以，特斯拉不仅拥有最好的数据集、全球最大之一的计算集群和一些世界顶尖的AI人才，而且他们多年来一直在优化神经网络和算法，不断去除多余的部分，以取得进展。这就是Eun Tassai在把FSD13与SpaceX的猛禽发动机3型相比时所提到的。我的意思是，这不仅仅是关于有多少计算资源和数据，虽然这是很重要的一部分，但这其中还有更多的微妙之处和故事。即使像Waymo这样拥有深厚资金的公司，也不能仅仅通过大量投入资金来解决这个问题并让其自动迎刃而解。

Doing that would likely lead to inefficiencies and these companies need to get in the mindset of deep-seek and really constrain their compute to push the engineers to be a bit more creative with what they have. Part of the reason I got the Model 3 performance is to spread the word about how special these cars really are. I want everyone to have an experience when they ride with me. To optimize that experience, I always want the interior of my car smelling great because that's a huge part of somebody having a pleasant experience. I've tried car fresheners but honestly they smell cheap to me and oftentimes it's way too much scent. I also don't want my passengers or myself constantly smelling toxic chemicals even if it smells good.

这样做很可能导致效率低下，这些公司需要转变思维，深入探索，并真正限制他们的计算资源，以激励工程师在现有条件下更加创新。我买Model 3性能版的部分原因是为了传播这些车有多特别的信息。我希望每个和我一起乘车的人都有一个美好的体验。为了优化这种体验，我总是希望车内闻起来很舒服，因为这对创造愉快的体验非常重要。我尝试过车用清新剂，但说实话，我觉得它们闻起来很便宜，而且味道常常过于浓烈。我也不希望乘客或我自己一直闻到有毒化学物质的味道，即便它是好闻的。

So after a long search I found drift, the sponsor of today's video. I chose them in part because they create air care products for your car and home using only premium essential oil blends with no harmful chemicals. They're vegan, cruelty-free and pet friendly. They do offer a subscription so every 30 days you can have a new scent and they also have a limited edition scent of the month inspired by the season. I have the winter Zen one right now and I have to be honest I literally kept the bag at my desk to sniff every now and then because I'm telling you it's that good and calming. They offer products as wood, stone or metal depending on your preference.

所以，在经过一番长时间的寻找后，我找到了今天视频的赞助商——drift。我选择他们部分原因是因为他们只使用优质精油混合物制作适用于车内和家中的空气护理产品，而且不含有害化学物质。他们的产品是纯素的，不进行动物试验，并且对宠物友好。他们提供订阅服务，因此每30天你可以收到一种新香味，他们还有根据季节灵感制作的限量版月度香味。我现在用的是冬季禅香味，说实话，我真的把装产品的袋子放在桌子上不时闻一闻，因为这味道实在是太好闻、太让人平静了。他们的产品可以选择木质、石质或金属材质，完全看你的喜好。

I'm trying the wood version this month in the model 3 but I'll probably go back to stone as it just blends with the interior a bit better. You just attach the scent block to the included clip and it magnetically snaps in. Pro tip, keep it off your surfaces when you first open it because it will be oily but once it dries out a bit you can really place it anywhere. Don't forget they have home scent diffusers as well and their products range from 9 to 15 dollars so very reasonable. If you'd like to please your passengers or guests in your home you can use my code DL55 to get 55% off your first month at drift.co or you can use the QR code on the screen. Enjoy.

我这个月在我的Model 3车里试用了木质香氛，但我可能会换回石质的，因为它和车内装饰搭配得更好。只需将香氛块固定在随附的夹子上，它就会磁力吸附。小提示，初次打开时不要放在表面，因为那会有些油腻，但等它稍微干了之后，就可以随意放置了。别忘了，他们还有家用香氛扩散器，产品价格从9美元到15美元不等，非常合理。如果你想让乘客或家中的客人感到愉悦，可以使用我的代码DL55在drift.co首月享受55%的折扣，也可以扫描屏幕上的二维码。享受吧。

On Tesla's official NACS blog page they now have BMW, Genesis, Hyundai, JLR or Jaguar Land Rover, Kia, Lucid and Mercedes all under the coming soon headline. Barring any further delays most of these companies should have access to the supercharger network sometime in the first half of this year. I saw a drive Tesla Canada point this out first but the new Model Y apparently has a two-piece trunk instead of the one-piece on the outgoing model. It looks like this line at the bottom of the glass separates this top portion from the bottom portion and Tesla may have chosen to do that because one of the most common areas for damage is this rear bumper whereas previously if this part was damaged you also had to replace the top part because it was one large piece. But now from a repairability standpoint think insurance prices if there's damage to this bottom part you don't also have to replace the top part.

在特斯拉官方的NACS博客页面上，现在显示宝马、Genesis、现代、捷豹路虎（JLR）、起亚、Lucid和奔驰都被列在了即将推出的标题下。如果不再有延迟，这些公司中的大多数应该会在今年上半年某个时候获得超级充电网络的访问权。我首先在Drive Tesla Canada上看到这一点，但新的Model Y似乎有一个两部分组成的后备箱，而不是旧型号上的一体式后备箱。看起来玻璃底部的这条线将上部和下部分隔开，特斯拉可能这样设计是因为后保险杠是最常见的受损区域之一。如果以前这一部分受损，你还必须更换整个大部件，因为它是一整片的。但现在从维修的角度来看，比如保险费用，如果这个下部分受损，你不必也替换上部分。

If you're not familiar a big reason why Tesla insurance rates are still sometimes higher than people would like is because the cost of repairs. Here's a look at the outgoing model and you can see all the way up to here this does appear to be all one piece. Not only that but the outgoing model has this rounded glass whereas the new model has a traditional corner or a flat squared off piece of glass which may lower manufacturing costs and complexity. Thus as Tesla continues to innovate and push for more manufacturing efficiencies they also may be adding some tweaks to focus a bit more on the repairability side so over time insurance prices can continue to come down thanks to repair prices ultimately coming down. Cox Automotive put out its quarter four 2024 EV sales report for the United States. Focusing on year to date for 2024 Tesla came in at over 633,700 whereas the second place brand was Ford at 97,800. Tesla sales were down 5.6% for the year however their market share is still at 48.7% as we know with only five models. And yes at 48.7% market share I think it's fair to say that roughly for every two EV sold in the US one of them was a Tesla in 2024. Another note I wanted to highlight Mercedes EV sales were down 41% year over year. Volkswagen's EV sales were down 52% and Volvos were down 62%. Just remember most of these companies are ramping new models or discontinuing old ones. So yeah Tesla was down 5.6% for the year but they also had the new model 3 introduction and that ramp at Fremont that took a bit longer than expected throughout quarter two. And finally I know everybody's been waiting on Cybertruck sales for quarter four the number was 12,991 and for all of 2024 the number was 38,965 which means yes the Cybertruck was the best selling EV truck in the US for 2024 the F-150 lightening came in second at 33,500. And then just rounding to the nearest 1000 the Rivian R1T did 11,000 for all of 2024 and the R1S did 27,000. The Hummer truck and SUV did 14,000 and the Chevy Silverado did 7,000. So for last year it really was the Cybertruck the F-150 lightening and then everybody else. Cox Automotive also shared this chart ranking them in order in terms of sales for 2024 top to bottom. Of course the Model Y and Model 3 are in a league of their own. From there rounding out the top five the Mach-E was in third, the IONIQ 5 came in fourth and the Cybertruck was the fifth best selling EV of any kind for last year. And as James Stevenson articulated in a different way all 20 of Tesla's competitors on the top 25 list combined did not sell as many EVs last year in the US as Tesla did alone. Scrolling down a bit the Model S had the lowest sales of any Tesla model at 12,400. However that was good for the 24th spot on this list so that means that all five Tesla vehicles cracked the top 24 in US EV sales for 2024. I did want to shout out the Honda prologue coming in the seventh spot because remember they did not begin deliveries until March of 2024. Looking ahead Cox Automotive predicts that one out of every four vehicles sold this year will be electrified with fully electric vehicles making up about 10% of the market total which would be up from 7.5% last year. And then of course hybrids and plugins would account for another 15% of the market leaving the ice market at 75% of total volume. As I've said before I do not want to become the gatekeeper here so I'm going to pass this along but don't shoot the messenger.

如果你不太了解，特斯拉保险费率有时仍然高于人们的预期，主要是因为维修成本。看看旧款车型，你会发现这里似乎是一整块结构。此外，旧款车型采用了圆形玻璃，而新款则使用传统直角或平面玻璃，这可能降低制造成本和复杂性。因此，随着特斯拉不断创新并提高制造效率，他们可能会在可维修性方面进行一些调整，以便随着时间推移，保险价格可以因维修成本的下降而继续降低。 Cox Automotive发布了2024年美国第4季度电动车销售报告。2024年迄今，特斯拉的销量超过633,700辆，而排名第二的品牌福特销量为97,800辆。尽管特斯拉的销量同比下降了5.6%，其市场份额仍为48.7%。在美国，每卖出两辆电动车就有一辆是特斯拉。另一个值得注意的是，奔驰的电动车销量同比下降了41%，大众下降了52%，而沃尔沃则下降了62%。这些公司大多正在推新车型或停产旧车型。 尽管特斯拉的销量年降5.6%，但他们也推出了新款Model 3，并在弗里蒙特的产能提升上花费了比预期更长的时间。大家一直在关注的Cybertruck在第4季度的销量为12,991辆，而2024全年为38,965辆，使其成为2024年美国最畅销的电动卡车。F-150 Lightning以33,500辆排名第二。Rivian R1T全年售出11,000辆，R1S售出27,000辆。悍马卡车和SUV售出14,000辆，而雪佛兰Silverado售出7,000辆。因此，去年实际上是Cybertruck、F-150 Lightning引领市场，然后是其他车型。 Cox Automotive还分享了一张2024年销量排名表。毫无疑问，Model Y和Model 3领先于其他车型。Mach-E位列第三，IONIQ 5第四，Cybertruck为第五畅销电动车。正如詹姆斯·史蒂文森所指出的，榜单前25名中的其他20个品牌加起来的销量都没有特斯拉多。 Model S是特斯拉车型中销量最低的，为12,400辆，排名第24，这意味着特斯拉的五款车型都进入了2024年美国电动车销量榜前24名。我想特别提到本田Prologue排在第七，因为他们直到2024年3月才开始交付。 展望未来，Cox Automotive预测今年售出的每四辆车中将有一辆是电动车，其中纯电动车将占市场总量的约10%，高于去年的7.5%。混合动力和插电式车型将占市场的另外15%，内燃机汽车将占总量的75%。正如我之前所说，我不想成为这里的把关者，所以我传达此信息，但请不要责怪传递消息的人。

This one is from Business Insider so I am highly skeptical. With that disclaimer the word is that Tesla is moving some workers off the Cybertruck production line to the Model Y line. The same employees noted that during meetings with supervisors they learned Tesla is scaling down its Cybertruck team and plan to produce fewer parts for the truck during the first quarter of 2025. If you've been following along the past few weeks you will have seen there have been some hints that there may be a slight Cybertruck slowdown so we'll continue to monitor through quarter one.

这条消息来自《商业内幕》，所以我持高度怀疑态度。尽管如此，有消息说特斯拉正在将一些员工从Cybertruck（赛博卡车）的生产线上调到Model Y的生产线。同样的员工还注意到，在与主管的会议中，他们了解到特斯拉正在缩小Cybertruck团队的规模，并计划在2025年第一季度减少生产该车型的零部件。如果你过去几周一直在关注这个话题，你会注意到有一些迹象显示Cybertruck的生产可能会稍有放缓，所以我们会在第一季度持续关注。

The weekly Tesla China number came in at 7,800. Comparing that to the same week in quarter four that number was 8,500. So now quarter over quarter Tesla China is down 22.6%. Then week two of quarter one last year was 7,400 and so now year over year Tesla China is up 25.5%. If you remember Tesla did an excellent job with the Model 3 change over last year at Shanghai so hopefully they can do the same thing with the new Model Y.

本周特斯拉中国的销量为7800辆。与去年第四季度的同一周相比，当时的销量为8500辆。所以，现在特斯拉中国的季度环比下降了22.6%。去年第一季度的第二周销量为7400辆，因此现在特斯拉中国的年同比增长了25.5%。如果你还记得，特斯拉去年在上海非常出色地完成了Model 3的换代，希望他们能够在新的Model Y上取得同样的成功。

I wanted to give everyone some context behind who James Thorne actually is as he's been posting more about Tesla as of late. He said Tesla is positioned to lead the AI revolution emphasizing autonomy and robotics with a target of a $2 trillion market cap this year. With Tesla's market cap at 1.27 trillion at the close today that implies about a 57% upside. James Thorne is the chief market strategist at Wellington Altus.

我想给大家一些关于詹姆斯·索恩（James Thorne）背景的信息，因为他最近在社交媒体上更多地谈论特斯拉。他表示，特斯拉在引领人工智能革命中处于有利位置，特别是强调自动驾驶和机器人技术，并预测今年市场市值将达到2万亿美元。今天收盘时，特斯拉的市值为1.27万亿美元，这意味着还有大约57%的上涨空间。詹姆斯·索恩是Wellington Altus公司的首席市场策略师。

I did some digging and Wellington has been ranked the number one investment firm in Canada and not just once but they earned this title five years in a row. As of last year they surpassed $30 billion with a B in assets under management. The firm has been focused on a new investment advisor acquisition strategy so that certainly helps to boost your AUM numbers and the company was only founded in 2017 so it's relatively young and for last year it was one of Canada's top growing companies. All that to say as far as I can tell this firm seems rather legit so yeah I would say this prediction carries some weight and we'll be tracking it throughout the course of this year.

我查了一下，惠灵顿被评为加拿大第一的投资公司，而且不仅仅是一次，而是连续五年获得这个称号。到去年为止，他们管理的资产超过了300亿（用“B”指出是十亿）。这家公司专注于新的投资顾问收购策略，这肯定有助于提升他们的资产管理总额。而且该公司是在2017年成立的，相对来说很年轻。去年，它还是加拿大增长最快的公司之一。总的来说，据我了解，这家公司相当可靠。所以，我认为这个预测有一定的分量，我们将在今年持续关注这家公司。

Bank of America is telling people to stay away from Tesla stock for now. The analyst downgraded Tesla from buy to neutral but simultaneously upped the price target to $490. But as you know I only highlight these if there's something I want to call out in what they said. Talking about FSD adoption they said it's risen sharply climbing from 22% in quarter one 2023 to nearly 60% among Cybertruck buyers in 2024. Of course the question becomes where in the world is John Murphy getting this data. More importantly though I wanted to contrast his forecast for the number of Tesla Robotaxes on the road. He's expecting 23 million Tesla vehicles will be FSD capable by 2030. Just yesterday we talked about how Adam Jonas from Morgan Stanley was only expecting 7.5 million by 2040. My point here is this most Wall Street analysts are really just guessing nobody actually has any idea how the next 10 to 15 years are going to go.

美国银行建议人们暂时远离特斯拉的股票。分析师将特斯拉的评级从“买入”下调到“中性”，但同时将目标价上调至490美元。不过，我只会在他们的说法中有特别想引起注意的地方时强调这些。在谈到FSD（全自动驾驶）采用率时，他们表示采用率急剧上升，从2023年第一季度的22%增加到2024年Cybertruck买家中的近60%。当然，问题是John Murphy的数据来源于哪里。更重要的是，我想对比他对未来特斯拉Robotaxi数量的预测。他预计到2030年，将有2300万辆特斯拉车辆具备FSD功能。而就在昨天，我们还谈到摩根士丹利的Adam Jonas预计到2040年，这一数字只有750万辆。我的观点是，大多数华尔街分析师其实都只是在猜测，实际上没有人知道未来10到15年会如何发展。

My point here is this most Wall Street analysts really have no idea how the next 5 to 10 years are going to go as you can tell from their varying predictions. What that means is there's still plenty of uncertainty and Wall Street does not like to invest when there is uncertainty. Which yes creates alpha or opportunity for retail or anybody that can see ahead of the curve. Chrysler's electric crossover that was supposed to debut this year but then was delayed until 2026 has now been put on hold until further notice. Any spending on the program is to be paused and the CEO said the program is paused for now as we assess the market and our customers needs and wants.

我的观点是，大多数华尔街分析师实际上并不清楚未来5到10年的走势，这可以从他们不同的预测中看出来。这意味着仍然存在大量的不确定性，而华尔街在面对不确定性时不喜欢投资。是的，这确实为零售投资者或任何能够提前预见趋势的人创造了收益或机会。原定今年亮相的克莱斯勒电动跨界车推迟到了2026年，现在又被无限期搁置。任何与该项目相关的支出都将暂停，首席执行官表示，我们目前暂停该项目，以评估市场以及客户的需求和期望。

In what feels like another beat the buzzer scenario the SEC today has sued Elon alleging failure to properly disclose twitter ownership. They're alleging Elon committed securities fraud in 2022 by failing to disclose his ownership in twitter and buying shares at artificially low prices. In response Elon's lawyer said in an email that this action is an admission by the SEC that they can't bring in actual case. He said Elon has done nothing wrong and called the suit a sham and the result of a multi-year campaign of harassment culminating in a single count ticky tack complaint. In the complaint the SEC is seeking a jury trial and asks that Elon be forced to pay a discouragement of his unjust enrichment as well as a civil penalty.

在一个让人感觉像是 "压哨" 的情境下，美国证券交易委员会（SEC）今天起诉了埃隆·马斯克，指控他未能正确披露他在推特中的持股。他们指称埃隆在2022年犯下证券欺诈罪，因为他没有及时披露他在推特的持股，并以人为低价购买股票。对此，埃隆的律师在一封邮件中表示，这一行动实际上是SEC承认他们无法提出实际控诉。他说埃隆没有做错任何事，并称这起诉讼是一个骗局，是多年来骚扰行动的结果，最终演变为一项琐碎的指控。在投诉中，SEC要求进行陪审团审判，并要求埃隆支付因不当得利而获得的利润，以及民事罚款。

Yet again we might have to take the wait and see approach here as under Trump Gary Gensler is no more the new guy is supposed to be Paul Atkins and under Trump there are supposed to be numerous changes at the SEC so it's somewhat of a question mark how this lawsuit is going to play out post January 20th. But for now the SEC is claiming that Elon underpaid twitter shareholders by over $150 million during the time when he did not disclose how much of twitter he was actually buying.

我们可能又得采取观望的态度，因为在特朗普任期内，盖瑞·根斯勒不再担任官员，而是由保罗·阿特金斯接任。在特朗普执政期间，美国证券交易委员会（SEC）将会进行众多变革。因此，这场诉讼在1月20日以后会如何发展，仍然存在一些不确定因素。但目前，SEC指控埃隆·马斯克在未透露他实际购买了多少推特股份期间，使推特股东损失超过1.5亿美元。

Tesla stock closed the day at $396.36 down 1.72% while the Nasdaq was down 0.23%. It was an average volume day for Tesla.

特斯拉的股价以396.36美元收盘，下跌了1.72%，而纳斯达克指数则下跌了0.23%。这一天对特斯拉来说是一个成交量正常的交易日。

Don't forget if you're interested head to drift.co use my code DL55 to get that 55% off the link is in my description and I do not think you'll be disappointed. Hope you guys have a wonderful day please like the video if you did you can find me on X-linked below and a huge thank you to all of my patreon supporters.

别忘了，如果你感兴趣，可以去 drift.co，使用我的代码 DL55 获得55%的折扣，链接就在我的描述中。我相信你不会失望的。希望大家有个美好的一天，如果你喜欢这段视频，请点赞。你可以在下面的 X 链接中找到我，非常感谢我所有的 Patreon 支持者。