There’s still the first X unaccounted for.
Also which one is a genius?
https://twitter.com/Grimezsz/status/1257836061520101377
https://twitter.com/elonmusk/status/1257925244733452290
… so Elons favourite plane then.
1 Like
My son’s name is Bitching Old Timey Fighter Biplane Cocktasticus. The second.
Worse pickup line, “My dad owns an emerald mine?” or, “My buddy and I are writing a pilot for NBC?”
3 Likes
This is the thing that’s gonna break me. I can feel it.
Why? There’s no better way for celebs to announce their arseholery to the entire world than by giving their offspring one-off freak names.
I meant Elon’s existence.
If he stole the secret of eternal life, would it be known as Elongate?
7 Likes
Ok, you bought yourself another two weeks of me keeping it together.
4 Likes
I certainly agree that if we assume that Google automatically has adequate data for training their models, then Tesla could not possibly have an advantage in this area. However, I (obviously) think they don’t. Assuming that it’s “trivial” to put a bunch more cars on the road is wrong, since bespoke modified cars are very expensive, especially with a bunch of LIDAR and shit in them. Plus they need to be driven by paid drivers. It would probably cost $10-20 billion a year for Google to replicate Tesla’s testing fleet.
The concept that all of the data is useless because most of the data is useless doesn’t seem very convincing at all. That’s the reason to have lots of data in the first place, to make sure that you capture all the weird corner cases.
1 Like
You’re assuming that Tesla needs as much training data as its getting and that Google has to produce the same amount of data to keep up. They’ve gotten the cost of LiDAR down to 7500 per car already (from 75,000).
Waymo is already valued at 175b. I’m pretty confident that they aren’t raising money at that valuation (30b more than Tesla insanely) because they are behind or can’t get the data they need to get to level 4-5 first.
Self driving isn’t about the cars. It’s about the software. Full stop. Waymo has cars that operate fully without human intervention already. They are significantly ahead and their lead is probably expanding not shrinking. The bottleneck isn’t data, it’s machine learning expertise and coding talent. To think that Tesla could compete with Google on either of those things is ridiculous.
Tesla is a poorly disguised battery company. They make a lot of batteries and battery manufacturing capability is the bottleneck that is preventing the other car manufacturers from entering EV’s at full speed. Toyota is on the record as saying that they can make 1.5m hybrids for the same battery cost as 28,000 full EV’s. That right there is what Tesla does better than the competition. Their actual car manufacture is cool and all, but there’s nothing particularly difficult to beat about it. It’s a luxury car with mediocre finishes that has a very cool all electric powertrain which costs an astronomical amount of battery production to produce.
This is the thing about manufacturing. The thing that really matters is always the narrowest bottleneck in the supply chain. For EV’s that’s batteries and it’s not close.
The real question is whether it makes more sense to have 1-2 decade period where all cars are hybrids that get 3-4x as good of fuel economy as traditional all gas power trains while we ramp up EV production. I’d argue it should be shorter… but that argument rests on the energy that recharges the EV’s being 90%+ solar/wind/hydro/nuclear. If you’re charging your Tesla with coal it’s probably not much cleaner than a hybrid that gets 50 mpg. Obviously this calculation isn’t measured 1:1 but rather in how much fossil fuel burning you eliminate per unit of batteries. I wouldn’t be even a little surprised if the hybrids came out way ahead with current battery technology… which is why I think Tesla the battery company would have been a way bigger value add for the planet as well as a much more compelling investment. I think there’s going to be room to sell as many advanced batteries as can physically be produced for a long while to come. That business model has served companies like Intel extremely well over the last few decades.
And the battery case is far from just about cars. The Tesla battery utility projects and powerwall stuff are at least as interesting as the car batteries. Batteries as a business is going to be growing at a triple digit % annually for the rest of our lives probably without any fear of generating more supply than demand. It’s going to be a fantastic business, and if Tesla actually makes it into the future it’ll be because of its battery business not because of Elon’s car building genius. TBH I think the company could replace him tomorrow and be substantially better off. His ego is stopping them from focusing on what matters.
1 Like
Tesla’s trying to do self-driving without LiDAR, it’s sort of an apples to oranges comparison.
This is true. The question is whether that’s a good idea. Another way of looking at it is whether Tesla is drawing dead to level 4-5 automation without LiDAR.
It would not blow my mind if Musk was totally wrong about this. The guys at Google sure seem to think he is, and honestly I trust them on software engineering way more than I trust him.
What Tesla autopilot does now is super cool. It is not however something that would enable you to run millions of cars in fully autonomous mode on a business model that disrupted how we do car ownership now.
1 Like
jesus, imagine defending musk after this
I’m also fairly confident that Waymo isn’t raising capital at an absurd $175bn valuation because they’re behind Tesla. My reasoning though is that in real life they’re raising capital at a $30bn valuation. And that’s a private-market valuation so, roughly two-thirds of a WeWork.
As I said, I think Musk’s “full self-driving” stuff is vastly overblown and would be surprised if I was willing to let my Tesla drive itself without paying close attention, especially off a highway, at any time in the next five years. However, there is ample, ample, ample reason to be extremely skeptical of the claim that building a self-driving car is a pure coding problem and that the part where you manufacture the car is trivial. (Indeed, given the heavy usage models you would want to fully exploit a L5 car, the maintenance costs of a gas-burner would become very important, and EVs are challenging.) Moreover, Tesla has a built-in fallback where their autonomous driving features just become high-quality collision avoidance safety features they can use to sell luxury SUVs to parents like me. That fallback isn’t available to a company that: (a) doesn’t sell cars, and (b) needs to charge $10k more for the equipment that runs the safety feature.
At a more general level still, most of Musk’s operations work on the start-up concept of building a minimum viable product, then iterating, but by applying that to the physical world. The irony is that tech company Google is planning to show up with a full-fledged “self-driving solution” that will come straight from the labs to production, while Musk’s car company is doing agile development. (Similar thoughts apply to the “Tesla-killers” that are always supposed to be just around the corner whenever a “real” car company decides to recreate the entire EV technology stack and roll it out…)
Your thesis about Tesla as a battery company is interesting, but you seem much too wedded to it. For one, it’s extraordinarily far-fetched to imagine that any battery company would be worth $150bn. Panasonic, which actually makes most of Tesla’s batteries, is a $20bn company. Intel, which dominates the market for the highest-value component of all computers, is worth like $225bn. Batteries for EVs are not going to be as valuable of a market. More importantly, Musk has often said that the reason he started Tesla was in part to speed adoption of EVs. Absent Tesla-as-car-company, the market for EV batteries would be Nissan Leafs and such. Tesla-the-car-company is Tesla-the-battery-company’s biggest potential customer. If you didn’t have the car company, the battery company would be vastly less valuable, even if the car company isn’t worth much on its own.
If Grimes marries Elon will her name be Grimes Musk?
2 Likes
You would be right that Tesla is worth nowhere close to 150bn. The reason it gets so much short interest is that it’s one of the most obviously overvalued stocks trading today, and its valuation isn’t just crazy high as a ratio but in absolute terms. There are plenty of companies valued <10b that I think are similarly overvalued, but they at least can claim that it’s theoretically possible for them to someday justify their valuation… Tesla on the other hand is trading 50b behind the largest automaker on the planet and at 7x the valuation of the largest battery producer on earth (LG Chem at 20.63b). There’s just no possible world where the valuation is even a little rational…
Which leads us to me calling Musk the greatest fundraiser ever. Which he almost indisputably is. In an era of tech startup valuations that make virtually every value investors spleen shrivel absolutely no one (including Travis Kalanick who is very very good at fleecing investors lol) has ever come close to matching Elon Musk. That guy can talk the financial markets into giving him enough money to take a moonshot at a valuation twice as big as the best case scenario if he hits the target dead on. It’s a goddamn mystery how he does it. I couldn’t pull the same trick off with all the blackmail material Epstein ever collected. You can’t help but be impressed.
And fair enough about the 30B. I, probably like you, just googled ‘Waymo valuation’ and took the number from the first article it pulled up. You probably dug one layer deeper and found out what they had gotten recently.