The point I'm making is that his "pfft, obviously Ukraine should have given up its nukes" stance is a Russio centric view that is biased towards treating one state more fairly than another. Not that it wasn't obvious that Russia treated the other soviet socialist republics that way before.

Not true. The radiation levels on the surface of Mars are around 230mSv/yr. The radiation levels in Ramsar, Iran, are 260mSv/yr. That's not because Iran has been doing all kinds of weird dirty things with nuclear material, that's just the natural background radiation level there. Want to know what issues they have living in such a high radiation area? Absolutely nothing. It actually has a *lower* cancer rate than the average. Our preferred model for determining how much radiation is safe for nuclear workers assumes that there's a linear relationship between how much radiation you receive, and how likely you are to have negative effects. It assumes that that's true no matter how spread out the dose is. The model says

- If one person receives a dose of 1kSv, 50 people will get cancer.
- If 1,000 people receive a dose of 1 Sv each in one second, 50 people will get cancer.
- If 1,000 people receive a dose of 1 Sv each spread out evenly over the course of 100 years, 50 people will get cancer.
- If 1,000,000 people receive a dose of 1mSv, 50 people will get cancer.
- If 1,000,000,000 people receive a dose of 1Sv, 50 people will get cancer.

It's a clearly, ridiculous model that doesn't in any way represent reality. In reality, you need to tie an individual's dose rate, and total dose to their risk of health issues. When you do that, you discover that below about 60mSv of radiation exposure health outcomes are actually better than baseline. The amount of radiation that we have actual solid evidence causes problems is 100mSv in a single exposure. That's. far far higher dose rate than anyone on Mars would experience

All that is to say, the radiation levels on Mars certainly are elevated, but they're not elevated to the point where our body can't handle it. The dose rates are low enough that it's likely people will have absolutely no ill effects, and may even have some small health benefits. If there are issues, there's a really easy way to deal with it - build your shelter underground.

There are some problems potentially with the trip *to* mars where radiation levels are around 1Sv/yr. Of course, we do have solutions for that. The steel sheet on the outside of the ship alone will attenuate radiation levels by about 10%. Wrapping the water storage around the outside of the ship, assuming a 20cm thick layer of water, around the ship, you're going to get attenuation of around 87.5%. With just those two, we've reduced the dose rate to around 100mSv per year, which... guess what... Perfectly fine.

Your estimatations are honestly way off. It takes NASA aproximately 3-4 years to build a single SLS, which is the only ship they have that can reach the moon with humans. They'd need 2 to send a rescue mission, due to the limitations of Orion's reentry, and that's assuming the Starship only has 3 people on board, which it has capacity for many many more.

So... chance of NASA sending a rescue mission to the moon? 0%. Literally 0%. There's no way they can do it. The only ship able to rescue a stranded Starship crew on short notice would be another Starship.

As far as xplodeship... Yeh, they've had a couple of failed launches recently. They've also solved the problems with them. They've got even more fixes coming in Block 3. Specifically, the problem was caused by a leak in the plumbing in the "attic" - area above the engines, but below the tanks, that you can think of essentially as a giant manifold. They've figured out the vibration issues that caused the leak, and changed the flight profile so that those vibrations don't occur. They've increased the capacity of the purge system that forces any leaked gasses out of the area. They've redesigned the plumbing that had the issue to not resonate in the same way. Version 3 eliminates the attic entirely, and mounts the engines directly to the bottom of the tanks. They've now had two consecutive successful launches, and reentries. I see no reason to think that next year they won't be able to get a pair of ships fully into orbit, and test docking, and fuel transfer between ships. Once they've done that, the only thing left to get to the moon is to nail the landing, which, frankly, if SpaceX can't nail landings... I have no idea what's going on.

NASA really hasn't paid for much of it. The HLS contract was for $4.5bn, of which, NASA has paid out $2.6bn so far. Starship development has so far cost somewhere around $10bn, so NASA's funded maybe a quarter of it. That said, you're right that SpaceX will do it for PR alone. They're trying to get to Mars. The problems you need to solve for that are a superset of the problems you need to solve for the moon, so they're going to be solving all the problems for the moon anyway. They'd be mad not to go to the moon.

Because you need to measure what happens when it goes off. You donâ(TM)t just want to find out if big boom happens, you want to find out what the properties of that boom were and if they match your modelling. Just finding out if one nuke goes boom tells you if one nuke is in working condition. Measuring everything you can about that detonation tells you if the entire arsenal works.

The problem is that Ukraine couldnâ(TM)t maintain nuclear weapons anyway. Giving them up was partly done because having a bunch of aging nuclear weapons hanging around in unguarded, unmaintained facilities is a terrible plan, and because guarding and maintaining facilities is expensive.

Why would âoeSoviet nukesâ get returned to Russia? Russia was only one Soviet republic, just like Ukraine was, or are you saying g that everyone was equal except some were more equal than others?

âoeTaken overâ - meaning actually got a democratically elected government, instead of a Russian stooge government.

Because thatâ(TM)s literally what nuclear weapons testing means.

Russia has had multiple failed nuclear tests recently, either with the bomb failing to detonate, or the launch system failing. Neither has convinced anyone that itâ(TM)s worth the risk that a few of them might work.

Given that the US has had other methods of making sure that its nukes are functional for decades, and doesnâ(TM)t have the massive problem with maintainence money getting funnelled into the wallets of oligarchs that Russia has, Iâ(TM)d be very surprised if we saw any failed tests.

Desirable to an extent, untilâ¦

1. You realise that thereâ(TM)sa bunch of people out there with legitimate expertise but no formal qualifications. Think, people like everyday astronaut, who has no formal engineering or rocketry education, but whose videos are always well researched and cited.
2. Politics can oh so easily become a âoeseriousâ topic. No degree from the places we approve of? No talking about how shot our policies are!

This doesn't sound like the result Big Oil would want in the report - only in reality. The report saying this will drive politicians to make efforts to increase renewables deployment, and lower data center electricity usage. What Big Oil would want is for the report to say "don't worry, you don't need to do anything, fossil fuels will be dead in 10 years anyway."

Yeh, and the Mac is far faster.

Wait... Apple doesn't come to mind for AI? You clearly no absolutely nothing about buying hardware for running AI.

In order of speed, your options are:

A server with $5000 nVidia GPUs with lots of VRAM on board.
A Mac, preferably with their highest end chip in it
A desktop PC with a high end desktop graphics card

Thanks to their unified memory architecture, Apple's neural engines are the fastest thing that doesn't cost $5000 for the GPU alone that doesn't instantly run out of memory running any vaguely serious model.

Apple ones - yes. The unified memory architecture makes them the fastest things you can buy for running AI outside of $5000 nVidia AI GPUs. Every desktop GPU is going to run out of memory on any serious model, and nothing else is as fast as these.

I mean, sure, you're likely wanting the M5 Ultra or M5 Max for doing that, rather than the base M5, but given that they're the same architecture, just with bits chopped off or doubled up, yeh... they're gonna need that.