Tesla aims to launch Full Self-Driving Computer 2 with Cybertruck next year
electrek.co
electrek.co
I'm fully aware of what FSD actually is today vs. Tesla's claims, so I didn't give Tesla an extra dime (nor do I use autopilot on my fully capable car since I believe it's a dangerous driver). Not everyone who buys a Tesla and pays for FSD has the benefit of knowing how close to scam it is.
One thing that would help would be making it illegal to sell features largely software driven before they actually exist (or at least make 100% of money spent on such features 100% refundable no questions asked by law). Tesla is raking in cash from people who don't know better plunking down $10,000 for a feature that doesn't exist (and might be 10+ years away) with the threat of it costing more in the future if they don't buy now.
The only way I've been able to reconcile the cognitive dissonance of giving money to a company that nearly defrauded people on some of their upcoming claims is to think of tesla as 2 distinct companies. One company manufactures and sells a very usable everyday electric vehicle and maintains an excellent network of chargers that let me roam the continent at will. The other company is a hype-machine designed to boost the stock price by announcing more and more features and models far beyond the capabilities of the manufacturing company. Right now Tesla has produced 5 different model of car over its lifespan (roadster and S, 3, X, Y for a childish joke). They have 4 announced and un-produced products (roadster, semi, plaid+ [now cancelled], and cybertruck), some of which have been announced for many years now. Does that seem like a sustainable company? Then add in the FSD debacle and accompanying hardware promises and it feels awfully pyramid-schemy. Luckily for the manufacturing company, the value of their factory, charging network and produced engineering is so high if the hype company fails the manufacturing can keep going. What I tell people is the tangible stuff you can buy from tesla is pretty good and there are lots of other excellent EVs from other companies already out and coming soon. If you want a tesla, buy one they currently manufacture and don't give them a penny or a thought for anything 'coming soon'.
That's ignoring all the basic issues with the cybertruck design. Tesla doesn't have EU pedestrian regulations to contend with (something cybertruck's basic shape likely cannot ever meet), but they still have a bunch of NHTSA stuff that isn't going to be easy with the weird choices they made. I suspect the production cybertruck won't be quite as extreme as what they showed.
I really like my model 3, but I treat any statement by Tesla as a complete fabrication until it's in the metal off a production line.
On the other hand, I'd be shocked if Ford doesn't deliver exactly what they announced for the prices they predicted. If I was looking to buy an EV truck right now I'd put my money on F-150 Lightning.
Looks like Elon Musk had invented perpetual moving target. It is always next year, and obviously as such they never miss the target.
To get there they need the data from their current customers. But once they get an approved version of FSD it will be self sustaining for data and Tesla's interest in their customer's FSD will quickly fade away.
If they are training significantly larger neural nets with dojo, and dojo is not really in production use until next year, it is likely FSD2 will be finalized only after they run some models from dojo.
I think in Elon Speak, they can make a driver safer than humans with the current computer, but FSD2 is needed for "robotaxis".
Maybe it’s reckless, maybe it’s dangerous, maybe it’ll never work in $scenario. But it’s more than a PR image.
This is PR.
This adulation certainly contributes. Just because other companies aren't constantly talking about or hawking their half-baked things doesn't mean they're not doing research. There are some companies spending more on research than Tesla's entire budget, but some of you seem to think Tesla is the only one trying new things because their CEO has to organize an event and riff about anything that comes to mind while pretending it's almost done. It honestly makes me sad that so many people fall for this.
What’s your take on it? Do you remember? I’d like to know more and gain some more context here. I mean, maybe Tesla bot and Asimo are part of something very cyclical, a technology pattern that’s been occurring for decades.
I think the issue with Tesla for me has been too much over-promising. Their products are stellar by all measures in my opinion, but I have great admiration for being intentional, transparent, and communicating clearly in business while also delivering. It’s hard to do but it’s important and I believe it signifies honesty, reliability, and company cohesion - sort of soft metrics I feel I can trust in the long term.
Tesla clearly embodies some of this, but the lack of intention aligning with reality, plans changing, lack of transparency - that’s a company for don’t want to depend on for products or to invest my money in.
If you ever saw the animated film “Appleseed Ex Machina” you might be more sympathetic to my position. Or I, Robot for that matter. I just don’t need a million cars doing whatever the next software update tells them, it’s not worth it for a little line following bot.
You carry in your pocket every day a device that does whatever the next software update tells it, yet you trust it will all your personal information and feel lost if it leaves your side.
Also my phone doesn’t do 0-60 in 2 seconds so I have a little less fear of it being automated
Meanwhile, it's 2021, and the CEO is saying "it's harder than I thought", while again promising that you can put your safety in the hands of a beta software "soon".
"Haha, FSD 9 beta is shipping soon, I swear! Generalized self-driving is a hard problem, as it requires solving a large part of real-world AI. I didn’t expect it to be so hard, but the difficulty is obvious in retrospect. Nothing has more degrees of freedom than reality." [0]
[0] https://electrek.co/2021/07/05/elon-musk-admits-self-driving...
It would have been better he raised funding behind the scenes and continued build good electric cars and keeping FSD as a skunkworks until it was ready.
Like Waymo - show of the tech but say it is not ready yet.