That is all.

"Tesla's self-driving system is quick-and-dirty, in comparison to those used by Waymo"

Nothing could be further from the truth. "quick-and-dirty"?!

Let's see how things go in Austin in June.

Book mark this and we'll revisit in 3 years.

You haven't tried Tesla's FSD lately have you?

"There's a reason robotaxis are only available in urban areas."
Yes, the current generation of robotaxis built on 1990's technology are all geo-fenced and using fussy and error prone lidar. Tesla's can drive anywhere. Even where they're not yet legal.

See:
https://www.youtube.com/shorts/O6d6YMG17UU

And lots of other examples.

Bonus. People, even the experts and market specialists, always underestimate the speed of disruptive change. This blogger mentions Tony Seba, worth looking up on Youtube.

https://alearningaday.blog/2017/10/22/horses-cars-and-the-disruptive-decade/

Two very different markets with different market drivers (pardon the pun).

"On the 10 year prediction of widespread driverless ride share cars, color me skeptical."

Less than 10 years. I'll bet you a beer.

There will be no normal uber, lyft, or taxi drivers, as we know them today, in 10 years as the cars will all be driverless. The only time there will be a driver is if you pay extra for one because you need assistance with luggage, or for security, or just to have some human contact during the ride. You will also be able to pay extra for a humanoid robot in the car instead of a human. Also, Uber will be bankrupt.

The Montreal Metro system has used that style since it first opened in 1966. The Paris Metro as well.

https://www.google.com/search?q=montreal+metro+map&udm=2

"The world's most powerful supercomputer", no it's not.

Tesla's Dojo AI supercomputer is 100 exaflops ("100 quintillion calculations per second"). And they have recently expanded it to 400 exaflops. xAI has another 100 exaflops. And they are doubling the size of each of them (at least) over the next two years.

Yes and... let's say Tesla's claim of 400 exaflops is the theoretical max and divide it by two to bring it closer to reality, and then divide it by 8 because we're going to use 8 CFloat8 numbers to do the equivalent of 1 64-bit flop, and then divide it by 2 again, just for good measure, what do we get? 12.5 exaflops or 7 times the capability of El Capitan.

Their next iteration will be 3 times that. And they won't stop there.

The first rule of making a decision or writing requirements is, "You are wrong."

Gather more information, gather information that goes against your bias, be less wrong.

See also: bike shed syndrome.

p.p.s

Look at Texas! :-O

https://www.seia.org/state-solar-policy/texas-solar

p.s.

Here in the United States there is also a *ton* of Solar PV being installed over the next few years.

More than 4 Hoover Dams worth in Nevada alone (from now to 2028).

Great interactive map of planned and existing solar PC plants here:
https://www.seia.org/research-resources/major-solar-projects-list

For perspective that is the energy equivalent of more than 10 Hoover Dams running at full capacity (which hasn't happened in years).

"If all of the generators are working at full capacity, the Hoover Power Plant is able to produce 2,080 megawatts of power. (This includes the two station-service units that produce power just to run the machinery and power the lights at the Dam.)"

https://www.eia.gov/kids/for-teachers/field-trips/hoover-dam-hydroelectric-plant.php

Analog CCTV + wires + analog DVR.

FWIW, the USS Hornet is the aircraft carrier that picked up the Apollo 11, and 12 astronauts. It's a living museum and has the Apollo Test Capsule CM-011 and the Airstream mobile quarantine facility (aka trailer) onboard, along with many other great exhibits.

Well worth your time if you're in the Bay Area, or the next time you visit.

They have multiple tours, if you're an engineer, the engine room tour is a favorite.

https://uss-hornet.org/history/splashdown