It surprises me is that there is much real tension between basic research and application development two exists in Meta. For any normal company, sure. But this is a company that spent over $70 billion dollars on VR that nobody ever wanted. And they can't fund both AI research and, separately, application development?

Yeah, that too. :) However, in practical terms, I'd assume that given enough time, willpower, and a LOT of $$$ we would both solve the environmental challenges and develop a "spacebus" to enable more efficient colonization, so the gene pool would become sufficiently diverse before it becomes a major problem. If not, we already know how that might work out from all of the historic in-breeding of the European royal families, in particular the Hapsburgs, although YMMV on what physical attributes, or personality traits for that matter, will be more likely to be "enhanced" in the Mars colony scenario.

Good luck with that. Birth rate might be declining in many countries, but we're still spawning around 100m new humans every year. That's an awful lot of human freight just to break even, and while prioritising shipping those of breeding age to transfer the newborns off-world (with all the physical development complications that likely entails) might help a bit, the reality is there are only two ways we get to point where more humans live offworld:

1. We take a long, long, long, time doing it.
2. A massive die off of those left on Earth.

Either way, the Earth-bound serfs are screwed.

Pulling repeated hardware platform switches underneath without bothering customers too much is a thing that only a very competent company could do, and Apple is a very competent company.

Apple nowadays is bound to avoid any huge missteps, because it has become a very conservative company.

Granted they blew some on the Vision Pro, but not much, for them. They folded on the electric car project, which now seems like a shame as Tesla is vulnerable.

What revolutionary product has Apple launched since Steve Jobs died? It has been 14 years, and I'm still waiting.

It was probably minimally profitable even with the Xbox 360 hardware issues. Still Xbox was a small part of MS so there was not a lot of profit pressure on them. Then MS started acquiring major developers. Activision Blizzard was a $69B purchase alone. Now MS has more profit expectations of Xbox.

Except many actions of MS says otherwise. For example, MS and Asus just released their new gaming handheld Asus ROG Xbox Ally X. It does not play Xbox games. It plays PC games. MS has been telling the consumer that ”Everything is an Xbox" from smart phones to PCs.

Yes, but in this case the US Government is the bank... China owns $700b of their debt, and can effectively either ask them to pay it back (cash out) when the bonds mature or dump some - or all - of them on the bond market at any point they choose.

Dumping them reduces their value, including any of China's remaining holdings, but would almost certainly cause the US' credit rating to be downgraded - and therefore the interest required to service its debt - all $38 trillion and counting of it! - to go up. It doesn't take many fractions of a percentage point increase to cause the US economy real problems in that scenario. If they decide to cash out mature bonds, then the US either has to pay up - which basically means balancing cutting budgets and raising taxes to raise the cash, or printing more money and devaluing the Dollar to pay for it instead; meaning prices of US imports and the cost of servicing the remaining debt both go up again. Neither are going to be popular with the US electorate.

The final option would be the nuclear one; default. That would crash the US economy completely, and bring most of the rest of the world's down with it, but especially those that have a significant slice of the global financial markets and a heavy reliance on imports (US, UK, EU...), but less so for those that, for various reasons, don't import much and are less dependant on the financial markets (BRICS). A default would also mean that the US' credit rating and value of the USD would go way down, so interest rates go through the roof and the cost of imports would skyrocket. On the plus side (if you can call it that), with the USD being worth less compared to other currencies, and perhaps significantly so, it could become *really* easy for the US to export anything it can still produce once the dust settles.

So, I take it you didn't read and understand my comment? At no point did I say China's imports are considerably smaller than their imports - I actually stated the opposite (see the last sentence of the first paragraph) - just that it's not quite the almost absolute "all or nothing" OP implied.

To re-iterate: Plenty of countries make stuff China wants, and they do indeed import in significant quantities from many different countries, contrary to what OP implied, it's just that despite all those imports their exports *still* massively outstrip their imports (as your linked data clearly indicates). Also, OP implied they were just sitting on the resulting surplus as cash (virtual or not), which was the real point of my post - that they are not doing that at all. They are, in fact, weaponising it and turning it into the economic equivalent of an aircraft carrier by buying government bonds (e.g. the national debt) of countries that they may have a problem with in the future. Many of the rest they are shackling using Belt & Road loans. If it comes to it, being able to crash an adversaries economy by dumping those bonds, then hitting it with a barrage of cyber attacks, there's a pretty good chance they'll be on their knees and have a panicing population to deal with before anyone even needs to fire off any kind of conventional weaponry.

As to VisiCalk being the real OG that started from nothing, there's an interesting comment on a VisiCalc youtube:

In 1969, we had to develop the world's true first electronic spreadsheet (LANPAR) within the limitation of 32k of memory - and we included forward referencing which didn't appear in Visicalc, TKSolver, Supercalc or even Multiplan I. Only in Lotus 13 years later. We even included the ability for sophisticated logic calculations, access to external data base data, and input of data in real time. Timesharing in those days was very similar to "cloud" computing now, except that you knew exactly which remote computer was doing the processing.

https://www.youtube.com/watch?...

Also at that scale who knows how much they are paying for each seat, I'm sure it's a bespoke deal. No doubt MS is making money, but I wouldn't just take the retail price and multiply it by 2 million.

No shit. We raised the price to $5 per letter, and now no one is writing any letters....

What makes the researchers think there are any humans on these channels anymore?

They are no longer required to fully fund pensions:
https://en.wikipedia.org/wiki/...

no problem.