karpierz

Ironically, you might be increasing Google's profitability if the searches you're taking off their platform aren't really monetizable.

Ex: "what is the tallest mountain" isn't really monetizable whereas "best bbq restaurants in Nashville" is very monetizable

To be clear: bail is not for criminals, it's for people accused of a crime.

Which is not to say if you commit murder, you go to jail without bail.

It's if you're accused of murder and made to stand trial, then you go to jail without bail, regardless of if you're guilty or not.

Really messes with poor people because if they don't have bail money, they either:

1. Sit in jail for a few months, unable to work, while their bills pile up, kids go uncared for, and rent is due.

2. Pay 10% of bail to a bail bondsman, who keeps the money regardless of if the accused is innocent or guilty, just so they have the privilege of walking free while awaiting trial for a crime they may or may not have committed.

You take a lighter and light gasoline on fire, using X energy. Then once the gasoline catches on fire, it'll burn on its own, releasing Y energy.

Same principle, different means.

I think these are all good points, but they don't really apply to people categorically; but rather to some people in some situations.

For example, if half of the group wants pizza, and the other half wants BBQ, a compromise might make everyone happier if they value eating together over eating their preferred meal apart.

> How does one magically snap their fingers and change their hard-to-control brain?

There is no magic solution. But recognizing that there is a problem, it can be solved, and deciding to put effort into solving it are what it takes to fix it. This is step one.

Why do people suck to be around?

But we're not talking about long lived, historic institutions; we're talking about South Korea, a republic formed out of the ashes of Japan's colonial project, and Ghana, a new country formed from the rapid retreat of Britain from its colonial empire.

You can talk about the legacy of the colonial institutions in the respective countries, but that has little to do with how hard South Korea worked in the 1950s, and more to do with how Japan and England administered their colonies, and how the transition from colony to state happened.

They're interesting books for sure, I remember enjoying reading them. But they're very selective in the facts they tell and how they tell them, and so they're almost worse than fiction in that respect.

As for South Korea, did you look into the impact of US economic aid as well as industrial advisors, and compare it to the economic drag that the US induced via the collection of loan payments on Ghana?

https://www.kdevelopedia.org/Development-Overview/official-a...

Would RK4 work for something like this, or does it lack some stability properties?

"Why Nations Fail" comes from the Jared Diamond school of "My Model Fits My Data So It's Causal". It's a series of just-so stories, which discounts outliers (ex: Singapore, India, China). They take the scientific method (find examples to disprove your hypothesis) and invert it (find examples that match your hypothesis), which does nothing except provide people with the feeling of confidence that they've finally come across the Truth.

Your implication is that Ghana failed where South Korea succeeded because they didn't work hard enough.

Of course people in South Korea worked hard to get where they are. So did people in Ghana. But in South Korea, the US protected them and funded them. Meanwhile in Ghana, the US stripped them of their wealth.

How well off do you think South Korea would be if the US didn't intervene in the Korean War?

South Korea had a propped up dictatorship that the United States strengthened and funded because they were scared of commies.

Meanwhile Ghana had a dictator which was saddled with US loans who was overthrown. But the country still had to pay the loans back, despite them being will outside of their means.

Of course. But we aren't talking about the States, we're talking about countries that are forced into return as a condition of loan forgiveness.

Some of those conditions involves avoiding any sort of export tariffs and using the declared value of goods to calculate profits (ex: if I spent 20$ extracting copper, and sold it for 20$ to an overseas company which totally isn't related to me, then the taxable profit in your country is 0$).

They have their debts forgiven on the basis of adopting certain policies (free market reform) that allows corporations from wealthy countries to set up shop and extract raw materials from the counter while avoiding giving back to the country via taxes. The profits get booked in Singapore or the Bahamas, and the stock price keeps going up.

> Setting aside this article, where's all the hate for giving to charity coming from?

People don't hate when people give to charity.

People hate when people earn money in ways that have negative side-effects and justifying it by: promising to (at some point in the distant future) give (some portion of) their money to (select) charities.

I guess I didn't think 'use' meant 'use facts to make arguments to advance your interests'?

Like, I wouldn't say that I 'use' 'asymptotic complexity' to promote engineering priorities over product management. I read a lot of nefariousness in that phrasing, but I guess that's not what was meant.

> Intersectionality is still used primarily, in practice, to promote the interests of higher classes against lower ones.

I don't follow what this means? How does one 'use' intersectionality to promote interests? I thought intersectionality was the idea that intersections of identities have different experiences than the components.

Ex: Black women have experiences that are unique to that intersection of identity, which aren't experienced by people who are only one of women or Black.

It is nonsense, but I don't think this is the right critique. EA is very focused on answering "where should money go".

EA is notably less interested in answering "where should money come from" and "what impact does the manner in which I earn money have".

What does it mean to earn money honestly? If I'm an investment banker, is that honest? Car salesman? International arms dealer?

Like it or not, the execution of EA is to make as much as possible without regard for the consequences of how you make it. The whole premise is that the ends justify the means.

What they really mean by "honestly" is in a manner that doesn't reflect poorly on EA, which is tangential to whether or not the money is earned morally.

If they're max deboosted, you won't see their posts even if you follow them; you'll only be able to see their posts by going to their page.

> You're right, there is no way to cryptographically prove that there are no contracts external to the blockchain.

But there is a very easy way to prove that there are not contracts external to the blockchain.

That's the whole point of an audit.

And it's very telling that they don't want their books audited, but repeatedly point to a clearly incomplete (from the perspective of someone who is concerned that the business might go under) as proof of solvency.

As with every other skill, practice.

Judo and skiing taught me how to fall. It's easier if you start young because you have less distance to fall.

That's measuring how much they owe clients (IE, deposits) against how much crypto they hold (IE, the crypto in wallets they hold). That doesn't at all capture the kinds of liabilities I'm asking about, which is debt external to the blockchain.

Ex: I am Kraken. Someone deposits 1 BTC with me. I then sign a contract with someone else saying "You give me 1 BTC now, I will give you 1.2 BTC in a year". I then sell 1 BTC. In my wallet, I will have 1 BTC, exactly matching my client liabilities. But my total liability is 2.2 BTC. And if those contracts start calling in early (IE, margin calls), I'll end up in a liquidity crisis.

Or more simply, what if Kraken just uses the BTC as collateral for normal loans from a bank? If they fail to pay them, then they'll fork over the client BTC.

> All liabilities were verified by the auditors and included in the merkle tree, and as mentioned, all users can independently verify that their specific liabilities are present. What more would you expect?

Where in the report do they verify off-chain liabilities?

I think the analogy is:

- you can have a gun, such that you'd be able to form a militia and fight government tyranny.

- you can advocate for collective bargaining, such that you'd be able to form a union and fight corporate tyranny.

Mergesort and binary search have a contract which defines:

- Which inputs are valid.

- For a valid input, what the constraints on the return value are.

You'd have a point if the implementations had as an input constraint: "array must be less than 2^30". But they didn't.

Otherwise, nothing is broken unless it never returns the right answer. Take:

fn add(x: u32, y:u32) -> u32 { return 1; }

This implementation works perfectly for numbers that add to 1. It just loses generality outside of that.

fn add(x: u32, y: u32) -> u32 { return (x + y) - (x >> 10) - (y >> 10); }

This implementation works for x < 2^10 and y < 2^10. Arguably this implementation is much worse than the previous one because it fails unexpectedly. At least the previous implementation would be much more obviously broken.

But these are both broken because they don't fulfill the (implicit) contract for add. You can't just say "well, it's implied that my add function only takes inputs that add to 1" unless you actually write that somewhere and make it clear.

The implication is that the statistics are what they are under the condition that women take the precautions that they do.

In other words, the 1/X statistic is conditioned on the average person's precautions (like not walking outside in a lightning storm), and by not applying on those precautions, you're using the statistic outside of the distribution that it applies to.

You could apply the same logic to any long term purchase. Dealerships only sell lemons because they want you to come back and buy more cars. Recruiters only match you with companies that you'll definitely want to leave because they want to match you again. Realtors only show you houses you'll hate because then you'll want to move again.

I worked at an online dating company. Matching people is fucking hard. It's hard enough predicting who will message who, much less who will actually date who. There's far more money in matching them well than stringing them along; if you had even a 10% success rate, you could get away with charging a silly amount of money.

I agree that a human can use the PV to get insight into why a move might be good. I also agree that non-interactive explanations provided by a chess engine will often be inferior. I think one way to break it down might be that the PV (or sets of PVs) can provide a useful explanation of tactical value (Kf3 leads to mate in 6), but struggles to explain strategic value (Kf3 leads to the positioning opening up, which broadly makes it harder for you to press your advantage).

> There is no equivalent to that in neural networks.

The first example that comes to mind is using backpropagation to determine how much influence each value had on the output result, and then rendering it in a human-readable way. Stuff like this:

https://miro.medium.com/max/4800/0*Y3Yi7cEueF0XLZP-

https://miro.medium.com/max/1100/1*IPhQ12OKnxH31AQNja6FnQ.pn...

That's punting the problem of explaining down the line. All the PV tells you is the computer is making this move because it thinks that 20 moves down the line, you'll end up in some desirable position.

It doesn't abstract or summarize any understanding. If you say, "well, what if they do Kf3 instead", all it can show is another PV. It can't tell you "knights on the rim are dim", or "this pawn needs to move now to prevent the bishop entering the position in 5 moves".

You can try to build an explanation on top of a chess engine, but it's very difficult to come up with generalizable or clear explanations beyond very broad heuristics (like "knights on the rim are dim").