[legolas2412]

> My guess is that Tesla will have robotaxis driving all around the world somewhere after 2025, but before 2030.

Just like how they had a cross country summon in 2017, coast to coast drive in 2018, and tesla network "robotaxis" by 2019 already?

[Isinlor]

Yes, Musk is overly optimistic. On average you can double his predictions. And his missed predictions on self-driving are extreme even for him.

But Tesla is making progress. Maybe even more importantly, the world is making progress on fundamentals of machine learning. If you want more reliable estimate than Musk, then you can check forecast aggregators with established track record [0]. I've asked "When will Tesla self-driving taxis be available to Metaculus users?" [1] Metaculus being the forecast aggregator. Community predicts 25% before 2024, 50% before 2026, 75% before 2029, 95% before 2035.

I'm personally more aggressive and predict 95% before 2030.

[0] https://www.metaculus.com/questions/track-record/

[1] https://www.metaculus.com/questions/5304/widely-available-te...

[nickik]

I love how if Musk gets something wrong for years people who hate him relentlessly shit on him. However he has been right about so many other things, but those things are simply not talked about and all the experts who wrote op-eds telling him how wrong he was are long since forgotten.

And as soon as a prediction doesn't come true, they rejoice and celibate assuming because it hasn't come true in the original timeline it would never happen.

When SpaceX couldn't land the Falcon 9 for a couple years the amount of industry insider and SpaceX haters who celebrated was amazing to see, and now landing and re-flying boosters is beyond boring because it is so common.

I have seen this cycle so many times now. So, I guess sit around and feel superior for now, be sure in the fact that in 2027 nobody is gone remember all your comments.