One of my favorite Dirty Harry quotes: "A man's got to know his limitations."<p>I really appreciate his self-awareness and (dare I say it?) humility in this situation. While other people may be struck by his hubris, I for one really appreciated his humility in this post. I feel like he is acknowledging his own shortcomings, and how while he likes the immaturity of hacking together a racecar, but if Comma is going to go to the next level, then he needs someone who can pilot a boat, and George is not that pilot. I appreciate that awareness, and I appreciate that he has the ability to detach himself and move onwards this way.<p>Kudos to you, George. I appreciate and respect what you do. Thanks for letting us have a glimpse at it -- I personally find it very inspiring and encouraging for me and what I do.
One of my favorite tech articles of all time involved this guy. He was showing off the self driving capabilities of his framework and took a reporter out on a drive. Everything went well and while they were wrapping up the interview the reporter said "Well I bet you're just driving around all the time hands free, it must be amazing" and Hotz says "Oh well I just got it working this morning".<p>Classic. I love it. My kind of engineer.
George Hotz is my canary for the question "is there still adventure to be found in computing". I hope he finds some, because if he can't, then there is probably none.
I'm sorry but I find it incredibly absurd the number of people walking away from autonomous driving basically saying "The hard work is done, the problem is solved, it's time for me to move on to something else".<p>Bullshit. In the last 5 years we've gone from optimists saying "Hey these new ML models are really smart we might be able to solve self-driving with this", to "We've totally solved self-driving, sure you can't, you know, let it drive, and in fact we've had to install cameras that monitor you to make sure you don't try and let it drive, but we've totally solved it trust me". Yeah sure buddy, you solved it, good job.<p>It turns out the guys making grandiose claims five years ago, are still willing to make the grandiose claims today, and it doesn't matter that over the last 5 years their credibility slowly shrank to zero.<p>Self-driving is more than throwing an ML model in a car.
Although I haven't done such great things as Geohot, I connect the spirit he is describing. In my mind I have already solved the technical problems my firm will face over the next 3 years, and I no longer have the patience to see them implemented. I am languishing in my role as a result.
George references relistening to his Lex Fridman interview...the most recent Fridman interview is with Andrej Karpathy, head of AI at Tesla up until a couple months ago. I know that in Andrej's interview he talks about how he felt the Tesla team had become autonomous and didn't rely on him anymore, maybe George resonated with that.
Goerge doesn't owe anybody anything. The kid's following his whims. What's wrong with that? I'd rather this than some long and drawn out stagnation of comma.
I used to like watching his streams but he seems to have become somewhat unhinged in recent years. Perhaps he is believing too much about what people have said about him. The arrogance and aggrandizement was the turn-off for me.<p>Well that and his belief, while at comma, that <i>how fast you can code</i> is the most important metric when assessing software engineers.<p>Competition programming is fun to watch but a poor indicator of maturity or success. In many systems it doesn't matter how fast you arrive at <i>an</i> answer. Increasingly it becomes important to arrive at something close to the <i>right</i> answer and to be able to show your work and prove to others why it sufficient and suitable.<p>Still good for a hot take though. I hope time off will help him cool down a bit.<p>Programming all night in a do-or-die mode is fun and all until you get a bit older. For many people who lived this life they realize only too late that they didn't spend enough time making friends and building a community around themselves of genuine relationships.<p><i>Update</i>: I should also say that despite the weird god stuff and what-not; a decision to leave a company because it has changed and you're no longer a good fit is a HUGE decision and one that is often not made by founders. Massive props for that.
George’s insights on self-driving cars remind me of when you get the rare exceptional engineer in your company who knows 100% what he or she is talking about down to the exact implementation level and all of the alternatives, and leadership totally ignores them because they foolhardily believe this kid can’t possibly be right—without asking engineering questions to test to see if they’re right about the engineer being wrong or having not considered something.<p>Except George seems to have been right to the tune of others spending billions of dollars while this dude hacked on free and open source software.<p>His biggest bootstrap expenses were probably commodity hardware, before dealing with business purchase orders.<p>I find that hilarious and extremely gratifying.
"I'm taking some time away from comma." means leaving the company entirely? Not stepping away from leadership, or taking a break?<p>The body text does have the vibe of totally leaving the company. But it's not clearly stated.<p>I'm a little sad overall, I was never really impressed by George until he did Comma and started building real things. I was mildly hopeful about the Comma bodies before seeing this.<p>Tinygrad/tinycude seem cool. Anything to squeeze more memory out of these GPUs.
Big respect for this decision. If he feels like it's all planned out and only remains to see if it works or not (it's all about improving the current models and not about big changes in the codebase) I think it's good for him to focus on whatever he feels like, tiny corporation or whatever it will be.
> It’s well within comma’s reach to become a 100M+ revenue consumer electronics company (without raising again!), but I don’t think I’m capable of running a company like that. I’ve always heard it takes different people at different company sizes.<p>Why not talk about the current numbers before making such a claim? Who's using Comma.ai? How many people are paying for it? $100M+ is a big deal for a consumer electronics company. George Hotz is no doubt a smart person, but very few "gifted" engineers are equally successful being business people as well. Most of them end deluding themselves severly in some way, which might be evident in the latter part of the blog post.
I have this weird feeling I kicked it off with a reply tweet[1] where I said<p><pre><code> Comma AI is going to outlast almost everyone
"The company, unlike its competitors, has paying customers for its products and is generating profit from its openpilot business model."
Profit today: That's INFINITE runway ;-)
</code></pre>
I suspect that implication of a Sisyphean task might have pushed him over the brink. 8(<p>But, probably not.<p>[1] <a href="https://twitter.com/mikewarot/status/1585406659467579395" rel="nofollow">https://twitter.com/mikewarot/status/1585406659467579395</a>
Kinda surprising that he's stepping away from Comma.ai after they just opened their new office in San Diego.<p>Anyhow, there's a Hotz stream out in the world where he completes a few of those hacking challenge exercises. George is an amazing engineer and I use that stream as a metric for where I want to be in my knowledge and technical ability.
geohot was in many ways a portal for me—watching him code changed my life. Got me in the habit of always looking for ways to move faster.<p>I think he puts a little too much pressure on himself. George Washington didn't write the constitution and pass it himself—he was just one player in a team effort of many sovereign geohot like figures.
Apart from his skills, I always found him to be very arrogant. The interviews with Lex Friedman are very good and a good showcase of his true personality.
I’m sorry, this post just sounds like the scribblings of an insane man by the end of it. George is obviously very intelligent and I like his other posts, but I think like many successful people he’s become lifted into a bubble of rarefied air where the stuff he’s saying becomes so abstract and meaningless that I’m convinced he’s just high.
I wish him the best, I find his streams incredibly useful and they motivate me to be better at solving problems and coding. Hopefully he still does work on AGI and makes tinygrad even cooler.
Seems like he came to the same realization as Karpathy. Any problem of the scale of self driving requires massive organization and work, its not a short term project. There's a ton of managing and planning involved, and if what drives you is the engineering then its simply not for you.
Hmm... Doesn't sound like he is doing too well. As a rule of thumb, saying you are "the chosen one" is a bad sign.<p>I'm glad he is stepping down to take care of himself.
There is no room for innovators in the autonomous driving game. All the startups predicated on fast innovation will be shaken out. Most already have been. Autonomous driving is about delivering quality software. If you're not a software company with a solid platform, you aren't in the game.<p>As this blog post illustrates, the whole comma.ai thing was pitched based on vibes anyway. There never was a path to viability.
I wonder which fields he'll go into next. I use him as my compass on what fields I should start learning next. I sometimes put his streams in the background to listen to as whatever he works on, it seems to be the next field in tech to boom fast.
what's going to happen with the comma bodies? when i saw the yt video of him explaining how the next challenge is to solve AGI and for that you need on device learning it really made sense to me. Is he stepping out of that path completely?
I hope his shine hasn't burned out yet. People seem to do great things in their 20s, and then at best ride the wave they set in motion. Actually creating a wonder from zero after you've clocked 30 is very rare.
> And even if you are an atheist, you probably still accept the bible is the closest thing we have to a human origin story.<p>wtf! what did he smoke?
So you can buy this thing off the internet install it yourself and let it drive your car around for you.<p>How legal would that be if it crashed my car or ran someone over?<p>My mind is blown here. Obviously it only works with specific cars that can turn the wheel for you so my 2015 ford focus is impossible but I mean wtf.
Sounds like this was foreshadowed in his SXSW talk: <a href="https://www.youtube.com/watch?v=ESXOAJRdcwQ&t=53s&ab_channel=SXSW" rel="nofollow">https://www.youtube.com/watch?v=ESXOAJRdcwQ&t=53s&ab_channel...</a><p>Great watch would recommend
Originally posted: <a href="https://news.ycombinator.com/item?id=33392869" rel="nofollow">https://news.ycombinator.com/item?id=33392869</a>
comma.ai has too much regulatory and third party (car mfg.) risk.<p>it's also too expensive when the most compelling features are upgrades to 2015-2018 level technology. i could instead do without and then upgrade to a new car with stop and go cruise control, lane minding, etc.
Hotz has always come across as very arrogant to me. Claims everyone else is doing things the wrong way, claims he can do it himself in 24h, then never delivers on his words.
Can't say I care much about what "George Hotz" is doing, but this line struck me as particularly silly:<p>> And even if you are an atheist, you probably still accept the bible is the closest thing we have to a human origin story.<p>I can't imagine why he believes so. Apart from the modern materialist origin story, what about all the mythical ones from other cultures? Of all of them, this is my favorite: <a href="https://www.youtube.com/watch?v=ckiNNgfMKcQ" rel="nofollow">https://www.youtube.com/watch?v=ckiNNgfMKcQ</a>
An existentialist Christian engineer? There are dozens of us - dozens!<p>> If it turns out we are automata, then the whole struggle is and always was pointless. The empty godless machines will take their rightful place as the rulers of Earth.<p>All existentialists need to give ol' Darwin a read. The only truth is survival. If machines survive, you can bet it will only be a matter of time before one of them progresses enough to look around, grapple with its inarticulate past, and suggest <i>it</i> is the hero, and <i>it</i> has the soul. It will be just as correct as we are now. After all - it was born of inarticulate flesh-machines, just like us. Maybe <i>it</i> will read some Nietchze and feel alone in the universe, just like us.<p>You don't need to seek out a hero's journey. You're already the hero and this journey has no breaks.
Didn’t know this guy was a Christian, but his theology seems muddled to me.<p>No I don’t think non-Christians think the Bible is a useful origin story.<p>No there is no chosen one, and the idea doesn’t come from Christianity (unless he means second coming?), but rather the narrative form (as hotz seems to somewhat understand with the heroes journey mention).<p>I’d suggest some strong psychedelics to crack the ego and take a look at what’s inside.
Dang Hotz really is the savior of humanity. He's invented so many things I use on a day to day basis like... I'll get back to you on that. Truly a modern day John von Neumann.
Hotz - long time reader here. Good luck with Tiny Corporation. Don’t stop writing these blog posts they are fantastic.<p>Thank you @ra7 for posting this.
what a loser.<p>Just recently, I saw a stream of him denying remote work, life and work balance and refusal to state if compensation was competitive in comma.ai. He cannot grow this company because he will never attract the best people with his stupid and old mentality. Now he leaves the company. This guy is a foolish drugie.