The thing about adopting evidence-based policy is that you also need to review and if necessary change policy when more evidence becomes available. The kind of situation you're describing would surely qualify.

I can explain it very easily. I don't want to talk to a machine. I don't want my car to listen to my conversation with the people riding with me. I don't want smart home assistants listening to my TV program. I don't want my tools telling me what to do. I don't want YouTube to automatically translate video titles.

Just because something is impressive does not mean I want it around me. That we can build a nuclear fusion device is impressive. But I don't want a hydrogen bomb exploding in my backyard.

Your whole argument hinges on the idea that long distance trucking happens within an European country.

And that's plain wrong. Long distance trucking in Europe mainly means transporting goods from the large harbors in the Mediterranean (Genoa, Piraeus) and at the Northern Sea (Rotterdam, Hamburg) to the large industrial centers and back. Additionally, trucks are transporting raw materials, furniture and similar goods from Eastern Europe to the West and machines and machine parts to the East. This means crossing borders all the time.

Let's put it like this: The Han canal was completed in 489 BCE, more than 2500 years ago, and the complete Grand Canal of China, which extends the Han canal from Bejing to Hangzhou to over about 1100 miles, was completed 609 AD.

If China is anything like Germany when it comes to electric truck adoption, then it's long distance trucking which moves to electric. In Germany, there are truck operators which have moved completely to electric - trucks which barely ever touch a town center. Electric trucks typically are rated for about 250 miles of range, which is sufficient for about 4 to 5 hours of driving. And after 4:30 hours, a trucker has to rest for 45 mins mandated by law anyway, while the truck can recharge.

You know, where the second child died of Measles? Lubbock, TX is at about half the distance between Dallas, TX, and Albuquerque, NW, and nowhere near the border to Mexico. All of the infected were not vaccinated, and most of them are under 18 years old - children, whose parents were not very keen on having their children protected. For some reason, none of the measles cases reported were illegal immigrants.

Petrol cars burn 30 times as often as BEVs, according to the statistics of insurance companies. Yes, for each BEV fire, you have thirty conventional car torches.

And that's a problem exactly why? I don't claim it to be 100%, I just point out the error of assuming 100% uptime for any type of energy source. Any power source based on heat and mechanical components has a lot of wear and tear, and components have to be serviced and replaced all the time, be it coal, gas or nuclear.

You have to pay the tariff to register the car with the DMV. I don't know how long non-U.S. license plates are accepted for a car driving in the U.S., but I would expect it to be limited to 12 month.

The most famous fire on a BEV carrying ship, the MV Fremantle Highway accident, was caused by a patrol car catching fire. But apparently, the BEVs get the blame, despite most of them remained undamaged.

You can get Chinese (and some European EVs) for half the $40,000. And in China, even Korean EVs sell for $10,000.

Anthropic's entire pitch has always been safety. Innovation like this tends to favor a very few companies, and it leaves behind a whole pile of losers that also had to spend ridiculous amounts of capital in the hopes of catching the next wave. If you bet on the winning company you make a pile of money, if you pick one of the losers then the capital you invested evaporates. Anthropic has positioned itself as OpenAI, except with safeguards, and that could very well be the formula that wins the jackpot. Historically, litigation and government sponsorship have been instrumental in picking winners.

However, as things currently stand, Anthropic is unlikely to win on technical merits over its competition. So Dario's entire job as a CEO is basically to get the government involved. If he can create enough doubt about the people that are currently making decisions in AI circles that the government gets involved, either directly through government investment, or indirectly through legislation, then his firm has a chance at grabbing the brass ring. That's not to say that he is wrong, he might even be sincere. It is just that it isn't surprising that his pitch is that AI has the potential to be wildly dangerous and we need to think about safety. That's essentially the only path that makes his firm a viable long term player.

It's a lot more complicated. Remember the solar panels to power the data center? They catch Sun light, so you need to add them to the surface of your data center. And to keep it at 300 K, you need twice the area to the other side to radiate off the heat. And those areas should not face each other, because they would then heat each other. It's a lot easier with convection, because then, the moving gas molecules transport away the heat.

You can calculate the amount of heat you can transfer via radiation. It's called Stefan-Boltzmann law. At a temperature of 300 K, you can radiate 460 Watts per square meter as a maximum. But from the Sun, you get 1370 Watts per square meter. That means that you have to have at least twice the area away from the Sun to keep temperatures at 300 K. A spherical body like the Earth would be at equilibrium at 279 K if it gets no other energy except direct Sun radiation.

The average uptime for a coal power plant is between 70% and 90%. Your example should be closer to 8 TWh/year. That means that you would need about 80 of them to provide the same power as Wind and Solar.