Chuck Cook is well known as being a very thorough tester of Tesla FSD software. He has had four drives with the new single software stack Tesla FSD 11.3.1. He conclusion is that with minor adjustments that it will be ready for wide release in the next dot release which would be FSD 11.3.2.
The significance of having FSD driving highway means that FSD highway should be safer and superior to Autopilot for highway. This means FSD highway should have over 6-7 million miles between accidents which is the safety level of Tesla Autopilot. It would also mean a rapid increase in the live miles driven by FSD. Tesla Autopilot is used for billions of miles each year. Tesla has FSD operating in shadow mode but going live with FSD would increase the pace of live data gathered and used by FSD.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
That page is tracking city miles, Brian was talking about highway, which is a lot easier.
What if fsd can navigate to avoid unprotected left turn ?
I like the limited lane change.
How about don’t change lane within one mile of turn and on ramp?
It would seem that Elon is kind of stubborn with respect to the FSD computer. All major breakthroughs with language models have come when they have increased the scale of the models, but the FSD HW stays the same for years. HW 3 – in production now – came 2016(!!). The next iteration – HW 4 – will have only marginally more processing power within the same meager 100 W power envelope. Tesla is trying to solve FSD by inventing better NN architecture, but it does not seem to increase the miles until disengagement noticeably over the last few years.
Furthermore, if you do solve FSD, nobody is going to care if you burn a few miles of range for each hour with FSD. Make the HW 4 at 1 kW with 10 times more calculations per watt for a total of a 100x improvement. Once the FSD is solved with the new HW, you offer the old FSD customers the choice of getting their money back or a discount on the FSD package for their next Tesla to avoid litigation and bad will.
Problem solved.
Correction. HW 3 came 2019, not 2016. It was developed from 2016. Still, four years without improvement is a long time…
No, the next dot iteration will not be able to drive millions of miles until the next accident would occur.
The FSD beta tracker shows that miles until disengagement has remained constant at around 7-8 for a couple of years now. If you just count disengagements that were safety critical, the corresponding number 70 miles between disengagement, also constant for a couple of years. So now we are jumping to millions from just 7? Even if Cook also had to disengage FSD with the singer stack FSD on his trips?
https://www.teslafsdtracker.com/