OT. Well I woke up early in this morning, and while expecting another boring day watching the time value of my option calls slowly eat away at their (current) face value (but also expecting a surprise climb, any day, sometime this year), an idea came to me*.
Forget radar, too slow and complicated to integrate with NN / Vision AI.
Something way easier and useful IMHO would be to incorporate a version of convoy assist driving: assuming there will be more and more Teslas on the roads, Tesla HQ would know which cars could share a segment of road driving, and sync their positions so one could follow the other - the advantage would be that the second (third etc) would benefit from the lead car driver's attention.
They would "see" ahead better than the lead car would as that they would also register that lead car's vision info.
More daring even would be to have following cars follow very closely, thus reducing air drag and improving fuel uh EWH economy - since they see / communicate w/ea other they could do so with extreme precision and confidence.
*and I hope some Tesla engineers read this, I'd be thrilled if it were implemented, tho if it makes sense, it probably is in the pipeline already.
PS. IF our political leaders had any sense they would incorporate V2X tech into our infrastructure and accelerate standards for X2X communications. That would really make sense, also actually remove a big chunk of Tesla's advantage, so rescue legacy car makers. Not going to happen of course, Jimmy Carter was probably the last elected political leader who had enough tech common sense and leadership to understand any of this.