What if Elon knows he might not need more factories - TBD?
Of course, he did say one factory per contintent. I know... it's a strong counter argument speaking in terms of the ideal today. But what if part of the equation to predict future demand is widely unknown, therefor staulling the decision on where to build next greenfield factory, if at all... ever. If the final volume or peak number of cars on the road is sufficiently low, there might not need to be another factory if it is faster to ship than wait for the next factory to be built. And if one factory were to create all the vehicles necessary for one continent, would you literally move the factory or simply start shipping and looking for more efficient ways to move inventory over, above, or even under the seas? Which is faster?
Although unlikely they don't build another factory, do we have a final count for # of robo-vehicles needed in the world over time? Tesla vehicles could last 30 yrs or more with new batteries, motors, and chips. The driving demand is also dropping from working at home, so are we using the correct math models today on future vehicle needs? How near is the Singularity really?
OK, I'm having trouble with the mere 1% BEVs sold comment which is a long way from 100%. But hasn't Tesla been just adding another line to existing factories with lots of land around the existing spaces?
Then throw in the robots and his dream of a blurred factory whizzing away... I'm not convinced 100% he needs more factories, and I don't think Elon's convinced either.
I think Elon is highly convinced they need fewer factories than previously estimated.
For what is the limiting factor for factory productivity?
If some of the key limiting factors are a combination of high yield, high quality, low (unit) cost factory equipment, then a bot could be the new piece that makes all the other pieces than n percent more effective. By its nature, a bot is a combination of equipment and human capabillities.
A bot can be viewed as either a highly adaptable piece of equipment, producing a predictable amount of predictable precision units of work. Or as a highly specialized human, working in concert with a set of specialized tools, also producing predictable units of work.
Or both.
What if the Tesla Bot is not only an option play, based on the strength of the AI stack,
but also and truly the new building block of the future Tesla factories?
Anywhere you own space, you can roof it, put in some standard production equipment, (maybe) enclose it and drop in 5-50 robots, and some unknown number of real, human workers to manage them and help out when they get stuck - now you have a micro-factory.
Want a bigger factory? OK -
scaling up is seamless.
Elon may very well have a strong hunch that Tesla can crank up production yield of existing giga-factories *with* bots and also utilize ordinary factories whereever they may happen to be optimally placed, or a bargin - or both.
The Tesla Bot is a force-multiplier: It makes all future Tesla factories more efficient.
By how much is yet unknown. But as Elon says:
The rate of improvement is really important. And I think the rate will be high.
The Tesla Bot may be the crucial building block for the alien dreadnought.
So, instead of only building more buildings and equipment, Tesla is starting to build the most adaptable and flexible factory equipment they can think of: Bots.
Imagine Elon thinking:
OK, so building new factories is time-consuming and boring. And gives the competitors and opponents ample chance for playing obstruction games on their own turf (Berlin).
So just invent a kind of factory with extreme scalabillity, which can be put into operation quickly: A collection of bots.
Done. Next.
Some play checkers, some play chess, some play Go.
Some invent new pieces, and play bigger and better games. And win.