IMO problems are not hard. They are merely tricky. They test primarily pattern recognition capabilities, requiring that flash of insight to find the hidden clue.
So it's no wonder that AI can solve them so well. Neural networks are great at pattern recognition.
A better test is to ask the AI to come up with good Olympiad problems. I went ahead and tried, and the results are average.
> But for offshoring, I'm clueless as to how manage to "reshore" the money, so to speak, so that you can eventually... Spend it to buy stuff. (Or isn't that the purpose of hiding the money ?)
Why would you reshore the bulk of funds? We're talking about people for whom a couple of million is just spare money.
If you want to buy a company, you just use the offshore funds directly. If you want to buy a major luxury good (a yacht or a jet), you do the same and then just lease/rent it to yourself.
If you _absolutely_ need cash in the US, you can pay yourself dividends and take a hit paying taxes on this amount only.
What is more likely, that your friend knows all about thousands of tons of secrete gold mined without anyone knowing, or that with constant satellite pictures of the earth, people actually do know roughly how much gold has been mined?
And I know someone with a stash of 100000000 bitcoins. The yield of gold mines is known, and it's pretty easy to extrapolate into the past.
I understand where all these conspiracies come from. It typically starts with "there must be enough gold to back all the savings, because Bretton-Woods or something". Then they check the total cost of the world's gold. It's a small fraction of the total savings.
So obviously most of the gold is hidden somewhere.
I guess it’s the sort of thing that encourages secrecy even into the far past — from both the takers and those from which it was taken from. How to know mining yields and extrapolate past mining?
But after some research, I’m convinced. We mine a lot more gold today than in the ancient past. Not much more gold left to mine, though — at current rates, we are done in 20 some years! I find that also hard to believe, but so it goes.
We mine a lot more gold today than in the ancient past. Not much more gold left to mine, though — at current rates, we are done in 20 some years! I find that also hard to believe, but so it goes.
We're not going to stop mining gold in 20 years. That's probably an estimate derived by dividing known reserves by current production rates. But that number is just an artifact of how "reserves" are defined. See this publication from the United States Geological Survey:
"Mineral Reserves, Resources, Resource Potential, And Certainty"
Reserve: That portion of an identified resource from which a usable mineral or energy commodity can be economically and legally extracted at the time of determination.
Miners continually quantify geological features to turn them into known "reserves." Some identified ore bodies also get converted into reserves by rising commodity prices or improved extraction techniques. No new ore bodies have formed in the past 60 years, but new reserves have continually been identified.
This misunderstanding is why people keep (wrongly) predicting that the world will run out of e.g. indium; even people who are otherwise educated make this mistake:
"Augsberg University Calculate When Our Materials Run Out - Soon" (June 4, 2007)
Armin Reller, a materials chemist at the University of Augsburg in Germany, and his colleagues are among the few groups who have been investigating the problem. He estimates that we have, at best, 10 years before we run out of indium.
While (human) vision is 3 colors, reviews of visual arts obviously can't just describe the colors of the thing. It also has shape, depth, style, etc.
Food reviewers don't note the levels of salt, sour, etc. They describe flavors and textures and parings.
But also, I don't buy that taste is just the composition of 5 components. This sounds like a reference to that diagram of the tongue we've all seen. It's as complex as scent is. There's no way you can define the taste of cinnamon by specifying some sort of 5-tuple.
I believe he is correct. The misunderstanding is from the old chart that showed certain tastes were only detected by certain parts of the tongue.
It’s still true that we can only taste salty, sweet, sour, bitter, and umami. All other flavor complexities come from scent simultaneously giving us information. It’s why everything tastes so boring when you have a head cold.
Think about this, suppose you're on a Zoom call and you want a person on the other side of the call to match a color that you're seeing. You can say "make it more blue", "make it brighter", "shinier", etc.
You can get pretty close to what you're seeing this way.
reply