My best bear argument is that full self driving is a long way off, not "a solved problem" as Elon has claimed, and not something we're going to see in any form at the end of 2017 as Elon has also claimed. Now, why do I think this might be the case? Why in this case might Elon simply be wrong rather than just a little over-optimistic?
Well, the argument goes like this. Elon is famous for eschewing reasoning by analogy and doing everything from first principles, starting with the physics. This is all well and good, and has worked well for him in seeing opportunities that others have missed. But, so far as software goes, there is no physics, there are no first principles. It's all fantasy. So Elon, rather than being something special, is just a smarter version of the rest of us. And that's just not good enough. This argument leads to not believing that Elon's path to FSD is going to be especially better than anybody else's.
So what if FSD doesn't work for another few years? How bad will that be for Tesla? for TSLA? I don't know. I'm way long TSLA, but this is one bear argument that worries me. It's really easy to be stupid about software.