I'm not following his reasoning on Tesla being unable to deploy a robotaxi fleet. I understand it requires a large headcount but, a few ideas to workaround that:
1. it could make sense for Tesla to acquire a single renting co in order to have a foot in the market. I doubt the acquisition would cost a lot as technology becomes the main differentiator for robotaxi networks, and the first mover could well be the last survivor (any renting co not finding a tech partner might go bankrupt fast)
2. Tesla could partner with networks of gas and service stations, to turn them into robotaxi centers (where cars are cleaned up while being recharged)
3. AFAIK all recently leased Tesla are to be returned to the company (no purchase option) so when Tesla will be ready to release true FSD, it may have a large enough fleet ready to deploy in the biggest cities
Either competing OEM will have to retrofit vehicles with 3rd party FSD tech faster than Tesla can pump out cars (which I doubt) or they'll manage to design and build FSD vehicles faster. If Tesla can get gigafactories up and running in ~a year, and they have the cash to build them, I doubt other OEMs will be able to move faster than Tesla. I'm surprised that Pierre disagree, as he's been able to recognize Tesla's lead for some time now.