I have often recommended AMD for price/performance reasons. THAT is something that gets buyer's attention.

IIRC those reactors already had cracks and were in the extended life stage, where they had to be carefully inspected on a regular basis to monitor the development of those flaws. They can't be repaired because they are highly radioactive, they can only be removed, the site cleared and decontaminated, and a new one built on the same spot.

Because Denver has high levels of radon gas, above EPA levels for about half the homes there, according to the City.

It does in fact lead to elevated levels of cancer. The local government put laws in place requiring monitoring and installation of removal equipment above a certain level.

You are also not comparing like-for-like. Radon exposure is different to the kind of exposures that nuclear plant workers get. Different substances, different ways of accumulating in the body, different risks. mSv/year is used because it is easy to measure, but the actual limit is set based on the type of radiation and radioactive material involved.

Therefore you cannot compare two mSv/year limits directly, unless they are often the same situation, such as two identical nuclear plants.

Renewables were 62.7% of Germany's electricity generation last year. At the rate they are increasing, by the time a new nuclear plant is ready it will be completely redundant.

Indeed. For example, a lot of cost overruns are because the construction companies don't stick to the plans without understanding the safety and security implications of changes they make. It's the norm for most building projects, they can substitute things and it doesn't matter... Until someone realizes that the substitute is flammable, as is happening with cladding on many buildings in Europe after the Grenfell Tower fire.

The mistakes have to be rectified, or an exception granted. At best it's a delay.

SMRs won't help. Each site is unique and has different requirements. Maybe if we had a Europe wide single developer building up experience with these projects, but everyone wants these massive, expensive projects to go to local firms. And even then, it probably wouldn't make a huge difference for decades.

It is a VM running on Hyper-V, but the Linux kernel is customized so the performance is near native. In benchmarks it's usually between 0 and 3% slower than native Linux. It is also less isolated than a typical VM, sharing memory with Windows, having access to local drives and other resources.

I've seen people use it for role playing games. The upload the entire rulebook and then ask it questions, and it responds with answers and references.

I'm going to give it a try with some notes I have made, just to see how well it works. I use Joplin for note taking and I'd love to see something like this in it. One of the few actually useful uses of an LLM.

Have you looked at what connectivity options your inverter has? You may be able to program it to not feed the grid when the price is negative. I don't know what options you have in Denmark, but in the UK data about energy prices is published a day in advance and there is plenty of open source software you can run on something like a Raspberry Pi to read it and control your inverter.

As well as avoiding negative pricing for people on that kind of tariff, you can do things like look at weather predictions to estimate how much energy you will generate the next day, and then decide if you need to charge your home battery/EV on cheap overnight energy or just wait for the sun.

That has been debunked: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.ch.cam.ac.uk%2Fnews%2F...

They note that the road itself releases more than the tyres. It's still an issue we should look at, but it's not as bad as is being claimed.

EVs tend to be better for these kinds of emissions too, because they use regen instead of just friction brakes. Friction brakes release particles as they wear down. People find they get to 100k miles on the original brake pads with an EV.

Problem is that nuclear is the most expensive option, and prone to sudden and unpredictable failures. If you want stability, add spinning mass and distribute everything as much as possible, don't centralize it.

Ireland has done it, they have flywheels installed at old fossil plants.

We don't have a full analysis for the blackout in Spain yet, but it appears to have been completely unrelated to renewables.

When blackouts happen, renewables are key to getting the grid up and running again. Fossil and nuclear plants need power to start up, which comes from other plants that didn't go down. Hydro is a popular choice because it only needs some generators to work the gates, so is ideal for bootstrapping fossil/nuclear.

NHK has some good documentaries about Fukushima on YouTube. I haven't seen the blame for the engineers, only upset at the management and the way they handled basically everything. It's on-going too, as they fail to meet their timetable for decommissioning.

Japan's problem is that they want to use domestic technology for everything, and their manufacturers like Mitsubishi are well behind on renewable technology. They have good wind resources, for example, but don't have the turbines to exploit it.

Nuclear has always been unpopular. Companies owning nuclear plants that were forced to shut down after the Fukushima meltdowns want to start them up because those are big investments that cost them money to maintain but don't produce any profit.

As for these magical new reactors that are safer that Fukushima Daiichi was, they don't exist, they would cost too much, and they do nothing to prevent corporate culture and chasing profit from neglecting them. Fukushima Daiichi was supposed to be safe, but TEPCO skimped on defences, skimped on upgrades when problems were identified, and didn't have a proper emergency plan in place. Now they are screwing up the decommissioning, behind schedule and struggling to make their robots work.

It's always the same story with nuclear. Just another few billion, the next generation will be better, trust me bro.

Taiwan is aiming for about 20% renewable energy this year, up from about 1.5% in 2020.

Their emissions have probably peaked, and have been dropping of late as renewables replaces coal and gas: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fourworldindata.org%2Fco2...

Of course they are still at a relatively early stage in terms of renewable deployment, and in the short term there might be some increased fossil fuel usage due to the nuclear shut-down, but the trend is pretty clear and renewables are going to continue to increase.

Of course nobody is suggesting that Taiwan become entirely solar powered. They have wind and hydro already too. If you check Open Sea Map (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fmap.openseamap.org%2F) you can see that they have shallow areas that are already being exploited, but again it is early days.

Nuclear has proven to be expensive, and they are reliant on importing the materials they need to operate it. As you point out, they are a relatively small island, so not much in the way of natural nuclear fuel deposits or good places to store the waste. Plus having so many eggs in one basket is dangerous if things ever kick off with China, so distributed generation is preferable.

The idea that AI has to be like a human to be general or powerful is probably flawed anyway.