I'd think more nuclear would be better for environment. Since sun does not shine around the clock and other renewables have larger negative environmental impact. Batteries for storage is not good either with today's technology.
Pure nuclear, or even as a majority production method, would be fool's errand, though. Unless someone manages to invent small enough reactors that can be started and stopped at will, to adjust a day's power demands. I doubt that can even be possible, though.
Given they don't complement each other, I predict that equilibrium to be unstable: either nuclear or renewables will grow to mostly replace the other, due to economic forces.
But they do complement each other... Nuclear provides the base generation that's online 24/7 while renewables are unstable and able to provide the peak demand.
No, nuclear wants a dispatchable generation source to provide the power. It only makes sense as a complement when the dispatchable generation is expensive to run somehow, if it's cheaper than the nuclear than you should just not bother with nuclear. The two things which complement it are gas turbine generators (cheap to build, dispatchable, expensive fuel), and storage (very expensive to build ATM, needs to buy power when there's excess, but otherwise cheap). Renewables are not this: the energy they produce is cheap but not at all dispatchable (curtailable, yes, but you can't just get more wind blowing on demand). What this means is that sometimes they fail to provide the peak and sometimes they can provide the whole peak and more, which both doesn't provide a reliable grid and eats into the economic justification for nuclear. So, you want to pair them with dispatchable generation to fill in the gaps, which sounds familiar, no? In fact the only difference is with nuclear your gaps are more periodic and there's not such a large range of the gaps.
That's why they don't complement each other: they actually want the same, different thing to complement them: something which can fill in the gaps in the power that they can economically provide. And renewables are a heck of a lot cheaper than nuclear at the moment.
No, they are not typically complementary. The optimal solutions for powering a grid tend to either be all-nuclear or all-renewable (usually the latter now), depending on cost assumptions. Optimal solutions with a mixture are uncommon.
At planetary scale it seems to be quite important not having to replace the whole fleet every few years don‘t you think? Just from a resource perspective this planet shouldn‘t drown in defunct solar panels.
Luckily, they are made of just a few basic, very recyclable chemical elements. Aluminium is fully recyclable. Glass is fully recyclable. Silicon gets purified as part of its manufacturing process. And that just leaves the plastic backing.
Just lately there was an article about one of the last 50 surviving video stores in Germany that has recently seen an uptick in customers. They have a huge collection of DVDs. They get a lot more requests for DVDs rather than Blu-rays.
The problem with these tools is that despite having worked with computers for 35 years, I don‘t get them. My brain is not made for them.
I only use out of the box vim when I work on consoles (which is still a fair amount of the time), I can exit (hey!), mark/cut/copy/paste (ok, yank!), save, and find/replace if I must. Everything else is just beyond what my brain wants to handle.
A lot of Jupyter lab and some VSCode otherwise. I can‘t say I know all about those either.
The last IDE that I knew pretty well was Eclipse, in about 2004. I even wrote plugins for it for my own use. That wasn‘t too bad for its time, I don‘t quite get why it got out of fashion.
There are those of us that still use it :) productivity gains lie elsewhere. And running various maven and git commands from command line instead of clicking around... something about keeping the skills in better shape
The metal support for pytorch is a real sh*t show at this point - all APIs are either legacy or unfinished. Also, app memory is about 3GB even for the largest phones.
It really still feels that it‘s all in its infancies.
Apple already has their own environment in CoreML and the neural engine. They have ways to train/run models on it in highly optimized formats and unified memory.
PyTorch came from PCs running on GPUs with their own pool of dedicated memory and a slower link to main memory.
I know people have been speeding it up a lot on Apple processors, but I’m not sure it’s a good indicator of how ready the iPhone is for running models for end user programs.
Rest assured 6-digit salaries are quite common in areas like Munich, and if you work for US companies like Google or Microsoft they can be double or triple that, or even more, through stocks. The sad truth as well is that there are not enough successful and large tech companies in Germany that you would see those there as well.
An org's learning curve. First time it happened on this scale. They will probably do a thorough post-mortem to make sure processes are improved next time.
I did this when I was still by myself, but a) having family and b) having more and more micro transactions (like coffee, transit, lime scooters…) and c) more and more accounts (amazon, paypal) means I had to declare time bancruptcy meanwhile
LPT: If you do switch to importing use CSV, not QIF. A lot of (i.e. all but one of my) credit card companies have two "description" fields, a short one and a long one. The OFX usually only has the short one, leaving you with often cryptic entries; things from Stripe and PayPal often get truncated before the name of the vendor upstream of them, making them impossible to decipher.
