Elon Musk is pushing Tesla to go all-in on Robotaxi. From everything that happened in the last few weeks, speaking to a few new sources, if it wasn't clear before, Elon has pushed Tesla to go all-in on self-driving. - Cheaper EV program, NV9, postponed. - Robotaxi timeline accelerated - Rush to build new data center at GFTX - billions spent on compute power If Tesla can't make self-driving work, it will have a minimum of three years without meaningful growth in deliveries because of this focus on self-driving. I get it, to a degree. AI is moving at incredible speeds right now, and there’s this feeling that technologies will move a lot faster because of it. Many tech leaders are rushing to attach their companies to its growth. I’ve had FSD Beta in my car for 2 years now, and it didn’t feel to me like FSD was improving at AI speeds, if at all, but I’ve had v12 for a week now, and I have to admit that this is a meaningful improvement. But Tesla FSD needs a lot more than a meaningful improvement to truly become self-driving. The idea is that the AI rate of improvement will help it get there fast, this is a huge bet. If it doesn't work, Tesla will simply not be the dominant player in the post-electrification auto industry that it was on the path to become. Elon is basically saying: Tesla wins self-driving or nothing. To be fair, he warned us all about it. He said: "Tesla is worth nothing without self-driving". Now, we can confirm it with the recent moves at Tesla. He set up the company to be like that whether you like it or not.
Full story on @ElectrekCo: electrek.co/2024/04/16/elo…
@FredericLambert @aargreen v12 on HW3 works almost perfectly under good conditions HW4 emulates H3 at the moment so we don’t know its performance, but should be at least 100x better So a Robotaxi with HW5 is certainly possible - it will be many orders of magnitude better through better HW alone.
@FredericLambert They’ll solve this, alongside Google, Uber and others. Uncertainty over regulatory approval is a big one and then the long road of building our transportation infrastructure optimized for autonomous vehicles.
@FredericLambert With the rest of legacy making moves to studio back EV investment there will be a lag period.
Autonomous driving technology saves lives. So it's for the best too IMO. Millions die on the road yearly. Many more are seriously injured and live a life as a handicap from then onwards. Not nice indeed for them. So I'll always push for importance of achieving autonomous driving technology.
@FredericLambert Very similar to the $NVDA story. Jensen missed the mobile revolution on purpose with 👀 on 1 price: “AI”.
@FredericLambert Elon does VERY well knowing when to go all in. He has proven this time and time again. It's all hands on deck with urgency, which is when he thrives and makes miracles happen. It's as good as done
@FredericLambert And if he didn’t go all in, someone else like NVDA beats catches them, and poof, it all goes away. Gotta do it.
While the anticipation for FSD v12.4.x is high, it's clear that substantial improvements are necessary to bridge the gap to Robotaxi's ambitious goals. Given the reported pivot to all-in on self-driving, every iteration from now on, including the upcoming v12.4.x, must address persistent issues like lane wavering, erratic acceleration, and the notorious UPL challenges. Optimizations are crucial as we approach the Robotaxi reveal on 8/8. Tesla's strategic focus intensifies the pressure for these versions to perform not just incrementally better, but transformationally so. @Tesla_AI @tesla_na @aelluswamy @chazman