Big loss – Tesla Director of AI announces departure (Ep. 623)

Andrej Karpathy, Director of AI, leaves Tesla. What are the implications:?

Apple Podcast:
Dave Lee on Parenting,.

NOTE: Please do not succumb to fraudsters who might impersonate me or others in the remark section. I do NOT give out my number or ask people to contact me.

♂ Disclaimer: All material on this channel is for conversation and illustrative functions only and must not be construed as professional monetary advice or recommendation to purchase or offer any securities. Ought to you require such advice, seek advice from a licensed monetary or tax consultant. All views revealed are personal viewpoint since date of recording and are subject to alter without duty to update views. No guarantee is given relating to the accuracy of details on this channel. Neither host or visitors can be held responsible for any direct or incidental loss sustained by applying any of the details provided. Author is long TSLA and other stocks at time of original video release date. When you buy through links in this video description, author might earn an affiliate commission.

Open an brokerage account at Interactive Brokers:.

Take a look at my archived articles/posts on Tesla and investing:.

#Tesla #TSLA #Stocks.

Big loss – Tesla Director of AI announces departure (Ep. 623)

Wealth Builders Club
Wealth Builders Club Secrets Revealed – Click Here to Discover the #1 Investment Resource!

You May Also Like

About the Author: Richard Money


  1. I’ve believed for some time now that Bot could enter limited production before true driverless robotaxis. And Andrej will not be needed to get this to completion.
    1. Both Bot and robotaxis have an AI aspect to them (deciding “what” to do based on what the camera sees and neural net outputs). Bot is obviously going to leverage all of this.
    2. Then there is a smaller piece of regular programming (not AI) which has to do with “how” to do what the AI decides needs to be done. For car, it is steering wheel, brake/accelerator, turn signals. For Bot it is a bit harder with many degrees of freedom of humanoid form – but this is not the AI piece.
    3. And finally robotaxi is fast, can kill, can generate lawsuits and has a virtually infinite number of edge cases. Bot when designated for a specific task like something on the factory floor, help at a fast food restaurant has a vastly smaller number of edge cases. Put the Bot on some taask and the number of environmental stimuli that are different/unexpected will go down to ZERO quite quickly.

    1. @Himansu Desai I’m not talking about the AI part of it, just think about manufacturing the BOT itself and the use cases. There is no point of manufacturing a humanoid robot to complete repetitive and boring tasks, for that we already have many different type of robots. It’s just another GIMMICK by musk , a mere deviation for the public and investors. I bet the bot won’t be ready this year, even if they make the bot I don’t think anyone would buy the robot for any general tasks. Humanoids are a great field for research, they are not great use case for public or regular use. I bet my roomba cleans faster than a humanoid 😂

    2. I agree that Bot is a much easier problem, and the early adopters will buy a ton of them. The average household might not have enough appropriate tasks for a Bot, so it might not be a long term product success. Retail and restaurant applications could keep it afloat.
      If it could be trained to drive a car, now then you’d have something!!

    3. @Kevin Dickson when you have a FSD feature, why do you need a bot to drive a car? 🤦‍♂️

  2. 5 years is a long time to stay in any high-pressure role, good on him for following his heart into the future.

    1. I hope he follows his heart and goes into academia. The more valuable contribution could be forming the next gen of AI contributors. He will continue to contribute, Imo.

    2. Exactly and in those 5 years he likely made 3 lifetimes worth of money. Can’t blame the guy for moving on to other things in life.

  3. Dave, first – thanks for jumping on this. I was looking forward to hearing your thoughts.
    Question: isn’t it possible that AK would be entering academia to spearhead AI with a University maybe an Elon-type TITS institution? I know the whole thing was a big giggle-fest, but the most ironic outcome is the most likely. I so hope we see him resurface in a role where he helps form a whole new gen of Elon-Drew-Andrej-JB types. We need them AND lithium.

    1. Yes.This needn’t be a negative if he uses his knowledge and communication skills to widen and deepen the expertise in AI either in Acedemia or the private sector or both.Didn’t Musk recently say in respect of corporate culture ”build Technology Trees not Moats”.To extend this metaphor one could see Karparthy’s departure as extending the roots of Tesla’s Technology Tree.

  4. There was a situation in one of the recent presentations where in my opinion Elons body language showed some impatience with Carpathy. I think Tesla will find a replacement, there are so many talents at Tesla and Tesla is a talent magnet.

    1. yep will have another talent replace him, it’s just AI. Not like design where it’s a bit harder…

  5. Since you mentioned the Tesla Ambassador Role Karpathy had, I think going back to Stanford (which is where he taught before going to Tesla) is a greater opportunity for being a ambassador for Tesla, answering questions to young and bright minds about opportunities for applied computer vision and ML inside Tesla. I think he’ll be a much closer ambassador than he was up until this point, even if he will not be employed by Tesla. Also, by the way, he worked at OpenAI before Tesla and he was giving consulting to the AI team at Tesla at the time before 2017, so I don’t think they have cut ties completely.

  6. With current AI tech, it is hard to make real breakthrough in short time in terms of car autodriving, as far as I can tell, I guess this was perhaps something wore Andrew out, not to mention the pressure from Tesla & shareholder side. After all, I feel Andrew is more of an academic guy.

  7. Anybody who’s had dealings with academics can tell that Karpathy is an academic at heart.

    The first thing he did during his ‘sabbatical’ was reconstruct an old AI classic paper and see how modern tech could improve the results.

    That isn’t the act of someone with burnout either.

    He’s going back to doing what he loves, with the added luxury of lots of money and a world-class reputation.

    1. Great addition Fred ! It is telling that he highlights his passion for open source and education. In his position he can carve out a really interesting deal with any of the big tech companies, without directly competing with Tesla in meaningful ways.

  8. While disappointing – this will happen to talent after a period of time when self interests, burnout, and financial opportunities are at play. Agree with you Dave – Karpathy is very likable, knowledgeable and has been a great evangelist for Tesla and Tesla AI.

  9. Dave, thanks for your immediate thoughts on this. I guess the true reasons for his departure won’t come up until months or years later, once enough time has passed for Andreij to feel comfortable disclosing them in an interview or some public forum format.

    1. 👏🏻❤️Thanks for viewing and contributing!
      I’ll share some beneficial knowledge
      Let’s converse with the number above..

  10. Thank you so much for a great presentation. It was fair insightful and informative.

    I think part of his departure related to the overwhelming job and possibly extreme pressure from Elon.

  11. To me, it looks very much like he was fired. I floated this idea on some boards at the time he originally went on holiday & was howled down as completely wrong. I also conjectured that Elon was disappointed with FSD advancement. No matter how brilliant his academic abilities are it was obvious to me that Elon’s patience would one day run out. And it has

  12. Dave, is everything OK? you seem low in energy lately. I hope you are good and if there is anything we can do to support just let us know. We all really appreciate your work.

  13. Sounds like he served honourably, productively, and honestly. 5 years of maximum effort may have caused burnout. He came, he saw, he automated.

  14. Five years is a long time. Karpathy may have been struggling with burn out and/or he may have been struggling with something else. What is that something else? Speaking as a former Silicon Valley professional programmer, (in their hayday) you get to a point where you not only have to learn new ways of thinking, but you have to “unlearn” your old ways of thinking — sort of, some of it. You have to decide for yourself what is hype or clever marketing and what is real and useful — “beneath the hood”. In my career I had to re-invent myself five times. This is a hard and painful process — especially if there are deadlines and you are pressed for time. Experience does count, and it does give you valuable insight, but you have to “break up and plough the ground” to uncover those insights. You have to continually “re-invent” yourself to keep up with the times — just to be able to talk to the bright, energetic young kids right out of college. Henry Ford would have a hard time keeping up with and understanding Tesla’s emphasis on innovation even though he was doing something very similar in his day. I think Karpathy took his sabbatical to read some books, and to reflect and think about what it was going to take to reinvent and update himself to get the job done. I think it was something like that. Maybe it was like “If you can’t take the heat get out of the kitchen”, but I don’t think it was like that for him. He has been innovating for five years. Karpathy is a smart guy who is driven by the challenge. I think he has decided to take time off to doodle, and to think, to do some creative problem solving exercises, on his own schedule, in his own way, in the shower, without all the pressure of immediate deadlines. He may fail. He may succeed. But I think he has decided to try. That is my best guess. Don’t be surprised in in a few months of a year he is back telling us the next new way of thinking and everyone else will have to do the adjusting. We live in interesting times.

  15. When it comes to investing, diversification is the key, that’s why I engage in sectors based on projected growth. There’s no shortcut to getting rich but there are smart ways to go about it

    1. Elisa deborah ramos has been doing a great job reviewing all chart, trade and techniques on BTC which has enhance the growth of my portfolio to 4 BTC lately.

    2. Started my investment with $1000 with her about a week now. if you stick to her strategies you win, if you don’t you have yourself to blame

    3. @Edward Albret I came across her name while reading an article on CNBC, But i wasn’t given much information about her I think she’s well recommended on YouTube channels

    4. @Sandra celin No she’s not!…You can just put her name on google and you will be directed to her website and drop her your message

    5. @paul dom just googled about her now and she looks legit, How do I get in touch with Elisa Deborah Ramos.

  16. Thanks Dave I appreciate your perspective on this huge development. Tesla’s 5-7 year lead in the marketplace from a technical point of view gives investors a chance to see how all of this will shake out in real time. In this market patience is key. 1 out of 4 Americans look to buy an EV vehicle due to high gas prices. What other company has the capability of coming close to meeting this demand while remaining profitable?

    1. 👏🏻❤️Thanks for viewing and contributing!
      I’ll share some beneficial knowledge
      Let’s converse with the number above..

  17. It’s frustrating as from what I recall when he started the sabbatical the statement were that he was not leaving Tesla.
    While I agree institutional investors don’t see Tesla as an AI company they will react negatively if they feel FSD is stalled or won’t be finalized as a result of this.

Leave a Reply

Your email address will not be published. Required fields are marked *