If anyone is interested in what the end result of “move fast and break things” looks like, this 4680 battery cell is a good example.

For those that don’t know one of the most expensive parts of making EV car batteries is, surprisingly: time

Think of each cell like a jelly roll pastry. A long thin flat strip of dough is rolled out, and the contents (jelly or chemical slurry cathode material) are added on top. This then has to be completely dried so there’s no liquid in it anymore. A second strip of dough has the same process with a different filling (chemical slurry anode material). This is dried too. Then the two strips are laid one on top of the other and are rolled up.

Its the liquid process that is expensive. It uses expensive solvents to make the slurry, and it has to have all kinds of time and energy drying them (requires a huge amount of space too!). Tesla was able to develop a process that worked in a lab that used dry materials skipping the liquid. This means no expensive solvents and no long drying time!

The problem is that this dry process isn’t scaling. Yes, they can successfully make these dry process cells, but many MANY of them fail quality checks and have to be rejected as scrap. So the biggest benefit of time (and therefor cost savings of manufacturing) never materialized. As soon as they got the process working at a small scale in the lab, they “moved fast” and simply assumed that they’d be able to scale it without issue. They build the Cybertruck architecture around it even before the 4680 cells could be manufactured at scale (at low cost). So when the cheap cells never came, the only option they had was to continue to produce the 4680 with very low quantities passing QA. This drove up the cost of the Cybertruck and drove down the promised range for the originally promised price.

This is a possible, and proven, result of “move fast and break things”.

There are a few more details to this story such as Telsa got half of the cell dry process (one layer of the jelly roll) to work fairly consistently, but the other half they never could. Another company (Chinese or Korean, I can’t remember) did get the missing half process working. Tesla wanted to buy the tech, the other company said “no”. Tesla then wanted to buy (in quantity) the missing dry half part from that other company. The other company said “no”. The company said “you can buy fully manufactured batteries from us though”, and that didn’t happen.

Fascinating article. If you’re asking yourself why they don’t just use these in another vehicle:

“the cells are still reportedly difficult to manufacture at scale due to the dry electrode process”

I hope they drop down to like $10k then I’m in