So with Elon talking about a major re-write of the vision NN that's "almost complete", and his recent answer on the Q4 earnings to the question about feature complete FSD I no longer expect a feature complete release to be shipped (and associated deferred revenue to be recognized) in Q1 or Q2 this year.
Reading the below transcript it "feels like" they are still building the foundational tools they will need to implement FSD, and they are not able to deliver even feature complete FSD without them including: The vision/decision system re-write which brings more of the problem into the NNs as well as the Dojo video training platform which will increase labeling efficiency by an order of magnitude.
Both of these examples sound like significant advancements, but given that Tesla is still working on the foundational aspects here, and has yet to really iterate with them yet I'm now starting to think the earliest we'll see feature complete FSD is Q3 but Q4 2020 is more likely, and that it could slip into 2021. I'd love to be wrong about this as both an investor and FSD purchaser.
What are other's predictions here with this new info?
From the 2019 Q4 Transcript: Elon Musk: "Well, I mean, to be precise, I said I was hoping it would be feature-complete with both FSD by the end of last year. We got pretty close. It's looking like we might be feature-complete in a few months. The feature-complete just means like it has some chance of going from your home to work let's say with no interventions. So, that's -- it doesn't mean the features are working well, but it means it has above zero chance. So I think that's looking like maybe it's going to be couple of months from now. And what isn't obvious regarding Autopilot and Full Self-Driving is just how much work has been going into improving the foundational elements of autonomy. The -- like the core autopilots in Tesla or Autopilot software and AI team is just is I think very, very strong in making great progress. And we're only beginning to take full advantage of the Autopilot hardware and the FSD hardware.
So I think it's -- the apparent progress, as seen by consumers, will seem to be extremely rapid, but actually what's really going on my head it seems like that is just having the foundational software be very strong and we've got really strong foundation. And then a really fundamental thing is moving to video training. So in terms of labeling, labeling with video in all eight cameras simultaneously. This is a really, I mean in terms of labeling efficiency, arguably like a three order of magnitude improvement in labeling efficiency. For those who know about this, it's extremely fundamental, so that's really great progress on that."