Our strategy is working. In the third quarter of 2024, we achieved another GM record for EV sales, delivering more than 32,000 all-electric vehicles in the U.S. - an increase of 46% quarter-over-quarter and a 60% improvement year-over-year.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fnews.gm.com%2Fhome.detai...

What's at risk of dying off is not EV's as a whole, but Tesla's first-mover advantage.

Musk placed all their chips on Robotaxi. I am very skeptical since "full self driving" is so short of solving that "last 5%" and they are so far behind Waymo in demonstrated capability and deployment. But, we shall see.

Synthetic fuels? They have two primary use cases: 1. When an oil-poor nation is at war. 2. When the fossil fuel industry wants to keep ass-fucking the world.

How about airplanes?

I would be happy to have pure EVs for my family's around-town vehicles, then a PHEV for the SUV which would make it affordable around-town (currently we avoid it) but still useful for longer trips, often with towing, to very out-of-the-way places.

How is that conservatism? It makes perfect sense.

The name of the ship was Bayesian, surely there are many nerd jokes to be had about a ship named Bayesian that sunk in a freak accident - or was it? Time to update our priors...

It's a really big ship, I would have thought a person-sized hatch would take a good while to let on enough water to prevent it from righting.

I didn't say that. If this is what NY wants then that is their business.

They jacked up the price and people's access to public roads decreased! Success!!!

As an argument against banning "FSD" as it currently exists - that is, with supervision - Tesla's stats do make sense. There is some argument that people cannot reasonably be expected to keep monitoring when they don't have to do much, so assistance requiring supervision actually reduces safety - Tesla's stats refute that.

But certainly, Tesla could not use stats from supervised driving to justify letting Teslas drive without supervision, i.e. the Robotaxi.

It's not just 14 stations, but also 150 battery-swapping electric vehicles. This could work in certain fleet applications and it sounds like that is their goal.

But it doesn't. For example, the lawyers that keep submitting court filings with fake references evidently have wrong assumptions about what AI is and is not (for now)

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.washingtonpost.com...

250 is less than 1/3 of 500?

A more typical use case for now would be using AI to generate some code, and then testing/fixing the code. Not running the AI every time to solve an instance of the problem.

Side note, I wonder if this paper compared AI performance to human performance. You think people can do towers of hanoi consistently?

In 1994 the msrp of the cheapest Geo Metro was $7800 and median household income was $32,260, or 4.1 x the cost of the metro. If a Versa manual is $17,000, 4.1 of that would be $67,700. Instead the median household income in 2023 was $81,000. So, the Versa now is cheaper than the Metro was then.

On the other hand people have long paid good money for pets, racing horses, or cattle with a prestigious pedigree. You want a pure-bred Samoyed pup? That will set you back $14,000. Nobody ever thought the offspring with be born with the life experience of their parents, they wanted the heritable traits of the parents.

Especially when the first example given is beef cattle. That application makes a lot of sense, but it certainly doesn't involve the cattle living forever.