The pressure differential for a space station doesn't need to be above one atmosphere. You reach that at 10 m underwater.

> no need to mine salt, it's everywhere

Most salt that we use is mined. Desalinization is expensive. It's cheaper and faster to use heavy machinery to dig it out of the ground. There's a mine in New York that produces 18,000 tons a day, and that's just one of several large mines in the US alone.

Promethium's natural availability is around 1e-15 ppm of the Earth's crust. It's actually cheaper to make it by bombarding uranium and then isolating the products to make the sub-kilogram annual volumes needed to meet commercial demand.

Salt mines are even easier and cheaper, and they're located all over the world, often well away from oceans (Austria, Switzerland, and Kazakhstan all produce several hundred thousand tons or more per year from mines and salt flats). There's a mine in New York that produces 18,000 tons of salt per day. Remove the chlorine, and that's 7,000 tons of sodium per day from just one location.

That's the case with every single one. I was not terribly impressed by AI three or four years ago. It's gotten better, but I really think that it's plateaued in its capabilities. There are still improvements happening, but nothing so far like the shift we saw when ChatGPT first hit most people's radar. The coming changes are likely to be more subtle and focus on accuracy and focused tasks.

For all the worries about massive job displacement, it hasn't happened. At my workplace, people were almost screaming for access to the enterprise AIs. Most use them for a few weeks, and then usage drops to zero, or close to it. If we had let everyone have it a year ago, we would have had to buy thousands of licenses. We have several hundred, but really only about 100 seriously use it. No jobs have been cut in favor of AI. No one has closed a job req because AI could do it. Academic studies are showing that the supposed cost savings aren't there. That's not to say that no one else has fired people in favor of AI, but I think it was used more as an excuse in most cases to cut positions that had built up and were already seen as excess. It was better PR (or at least they thought it was) to say that it was due to AI rather than good old-fashioned cost-cutting.

The fertility rate in the US has been below replacement level since 1972, except for a handful of years. It started to decline with the Great Recession, and it continued a steady decline until 2022. It blipped up in 2024, but it's still not enough.

Tuitions aren't helping, but birth rates are a real factor. My kids' school district closed five elementary schools over the summer, and it may close one of its middle schools in the next couple of years after it escaped this round. The district built up about 20-25 years ago to accommodate over 50,000 students based on population forecasts, but it peaked at 44,000 and has been slowly sliding for several years. Other districts are doing or considering the same thing for the same reason. This is in the Dallas suburbs, an area that is still growing, but the number of kids being born is declining.

The planned investment numbers they announced in late 2024 were the moment that I knew that they were a dead company walking, and that the entire AI industry is unstable. It reminds me of what I saw a domain for a homepage portal company in the dotcom boom sell for some ridiculous number of millions of dollars because people were still convinced that everyone would use the web by starting with a portal page like they did with AOL, ignoring that AOL basically forced you to do that.

Altman said that he plans to make a trillion dollars in investments for OpenAI by 2030. I don't think they have or will have anything remotely like the income or investments needed to cover that, and if they do an IPO to raise the cash, their debt and cash flow are likely going to become problems. They've been in panic mode for a few months as ChatGPT has fallen behind at least Gemini if not also Claude in quality and performance. I admin ChatGPT Enterprise where I work, and the interface and options are slapdash at best. Support is worse, as they don't have a ticket management system (because that's how old business does it, not how new business does it), and I've seen ticket/case numbers in three different formats.

Even Microsoft, one of OpenAI's biggest investors, is adding Claude as a model option, and with their in-house work on AI chips, I will not be surprised if Microsoft takes the path of Google Vertex and start acting as a portal for dozens or hundreds of models, build their own competitor as a primary, or both (Google's total model).

> Meanwhile Amex are advertising cards at 98%.

The highest credit card rates are around 36%, and those are for people who have incredibly bad credit. Amex certainly isn't issuing 98% APR cards.

Credit card companies won't. It's not Visa an MasterCard issuing the debt. The banks will write it down, which will affect their bottom line and the results of their stress tests. They'll rein in credit lines, making it harder for people to do the things they want and need to do.

It would hit most businesses, exploitative and not. The US economy's backbone is effectively credit cards. Take those away, and a lot of people that use them as bridges between paychecks are going to have a hard time.

Most hotels and rental car companies accept debit cards connected to the credit networks, but they will put a lock on a certain amount of funds for incidentals. For cars and hotels, this is typically $300.

In 2016 and early 2017, his brilliant plan for dealing with the US debt was to simply let it default and declare bankruptcy, discharging the debt and starting over. He figured that this would mean the US would have little to no debt left over and wouldn't have to worry about the interest anymore. Fortunately, enough people told him that doing so would crash the national and global economies and turn the US into an economic pariah for generations. He hasn't really brought it up since.

Ukraine just got a €90 billion loan from the EU. That's enough to cover their financial needs for another 12-18 months. They're starting to sell their designs to Western manufacturers, too, who are intensely interested in cooperation. Ukrainian companies are starting joint ventures outside the country, so that's another cash lifeline, albeit a smaller one.

It is up in the air whether Ukraine gets much territory back. Ukraine is not, however, on the verge of collapse. (Neither is Russia, for that matter, but more later.) Pokrovsk is likely to fall, true, and Ukraine holds only some outskirts, but when I look back to January 1, 2025, Russia was 5km from Pokrovsk, and a year before that, Russia was 35km from the city. Looking at the wider map, the battle lines have not changed that much in two years. Russia has made some advances south of Kupiansk and is contesting Kupiansk itself. If these small cities with pre-war populations of 60,000 and 30,000, respectively, give Russia so much trouble, what is going to happen when they try to take Kramatorsk and Sloviansk, with their pre-war populations of 150,000 and 110,000, if they can even get there?

Russia is not on the verge of collapse, either, but it is on the verge of desperate measures, and may already be in them. It's selling gold internally to take rubles off the market in a bid to both strengthen the currency and provide rubles to redistribute without printing more, with the National Wealth Fund selling off over half its 400 tons of gold to oligarchs. There's still almost 2000 tons of gold left inside Russia, but any significant dip into that is going to cause even more economic problems. A combination of Trump's bumbling and Sunni Arab OPEC states upping their output has resulted in a drop in the price of oil (WTI is trading around $57 today while Brent is around $60). Russia sells at a discount from the market price, and its budget is based on being able to sell oil at $70+ per barrel. It's getting $50ish now, and that may drop even further. Ships are becoming less likely to take on Russian contracts due to the increased activity by the US Navy and attacks by Ukraine. On top of that, while the Ukrainian drone strikes against Russian petrochemical plants isn't decimating the output quite like some think, it is heavily impacting refining capabilities such that Russia put a moratorium on refined products exports, further squeezing their foreign cash income, and they've put limits on internal distribution, which is raising prices, especially outside the major cities. Enlistment bonuses are dropping as local governments run out of money to pay them, with even the relatively wealthy St. Petersburg reducing or delaying bonuses. The federal government is lavishing at least some veterans with education and health benefits as well as some financial benefits, but that appears to not be universal. This is causing a complex situation resulting in high inflation (officially around 7%, probably much higher). The only reason that Russia isn't in hyperinflation right now is tight controls and high core interest rates (16% now, down from 25% because it was absolutely strangling growth). And Russian defense contractors are filing bankruptcy at alarming rates, putting key supply chains at risk.

Despite its population advantage, Russia does have a manpower issue. They're largely not sending conscripts into battle. By law, under the current circumstances, only contract soldiers can be sent to fight (there may be some exceptions, but this is the overwhelming majority of Russian forces fighting). They don't have the money to pay the big bonuses to keep the contract force up, and as veterans return missing limbs and living with severe PTSD, more people are becoming aware of what it's like at the front. Criminals whose sentences were commuted are also returning from the front lines (journalist Emily Hoge has been following this for some time, including the results when hardened criminals become hardened soldiers and return to the streets), pushing up crime rates. Escalating to legal war conditions to allow conscripts to be sent into battle would be an enormous political gamble for Putin.

Russia's problems aren't a lack of physical production capacity (they have that in spades) or people (they have many more they can conscript if necessary). Their problems are that they're running out of money to spend on the war and men they can legally send to fight. Changing the circumstances on the ground requires changing significant political positions inside Russia that have a very strong chance of provoking political backlash, regardless of what levels of control Putin has been able to exert so far.

Maybe Russia can force a ceasefire or even a peace, though it's hard to see what terms would work in such a way that both sides would accept, as both have become skilled at finding something the other side absolutely will not accept, so the war continues. Maybe Ukraine can eventually push the last Russians out of its territory, possibly even out of Crimea, though that, too, is a tall order. Then again, it seemed in the early 1980s that dislodging the USSR from Afghanistan was a tall order. Right now, there's not enough information on either to hazard an educated bet.

The name on the Kennedy Center is slightly longer than that:

"The Donald J. Trump and
The John F. Kennedy Memorial Center for the Performing Arts"

Restoring the name just involves removing the first line.