aetherson

Building infrastructure that gets productively used makes you richer, if the infrastructure costs less than the productive use of it. Spending more money for the same amount of infrastructure productivity is bad. There's no magic here where, ceteris paribus, spending more money on infrastructure is itself good.

Your pronoun antecedents got confused there, but just to be clear, France is 5x larger than the UK in terms of land area.

I don't know that this really points in one direction or another. The UK might have more difficulty doing large projects due to being relatively "crowded" and having less free space, but on the other hand it might also benefit from needing to build less total mileage of road and rail, and for amenities which are mostly bound by the travel distance to them being needed in less quantity.

This is the same core observation as Scott Alexander's famous article, Meditations on Moloch.

https://slatestarcodex.com/2014/07/30/meditations-on-moloch/

The main value of stressing about sprint goals and what's "in" a sprint and finishing it is the very specific situation of there being a bunch of people who want to feed in a large amount of work to a team (usually bugs/support requests/operational work) and it results in constant prioritization issues.

If you don't have that situation, you probably don't need to be really careful about what's in a sprint.

Do you mean pool?

No. I don't know what your parent poster is talking about. D&D chargen is not in any edition of the game terribly complex.

It's not wildly simple, either -- D&D is squarely "rules medium" and there are plenty of RPGs with simpler chargen. But it's not notably hard for its space.

20 years ago there was more manual calculation of numbers on a character sheet, that now is likely to be handled by a digital sheet

I think that's a failure mode his worse books fall into, but I've found his better books enjoyable all the way through.

I think that Snow Crash is good, but those are three of his best books, too.

I appear to be a little odd in liking Reamde as much as I do, but if you want something lower-concept and more thriller-like, I really enjoy Reamde.

I admit that I assumed that "geometric browning motion" was just "a random walk with more extreme movements when it goes away from the starting place," and that was wrong. What geometric browning motion in fact is, with some handwaving simplicity, is exponential growth with some jitter.

I will note that "stocks are priced via geometric browning motion" would satisfy the EMH. And that the underlying growth trend of GBM, in a real world, is, like, "the stock market actually values real things about the company that you can find out, but which other people have already found out."

The very idea that the "market can remain irrational" implies a non-random price algorithm.

Companies never have a negative price. Also, just look at stock graphs. Companies are not priced by geometric browning methods.

This is what I mean when I say that it's usually not helpful to try to make claims about the world based on, "I'm really mad about the EMH." You end up making silly claims.

If markets just randomly assign values to companies, you can beat them in two straightforward ways:

1. Just buy when the price is low and random variation will move the price higher eventually (or short when the price is high).

2. Ignore the market except to buy when a company is underpriced, then just have the company declare dividends or whatever and directly eat the profit that the market is stipulated to have underpriced.

A very small investor in a random market might have difficulty with either of these strategies, but a reasonably well-capitalized investor would not.

If you want to think up some kind of complicated model of difficult-to-take-advantage-of company pricing that is hard to exploit, really ask yourself whether that model is grounded in anything other than, "I'm mad about the EMH."

Nobody sensible believes that financial markets are perfectly efficient.

On the other hand, nobody sensible believes they are completely inefficient, or trivially beatable. If you want evidence of this, feel free to go out and beat the market with long plays on the alpha you imagine you have.

I think it's usually helpful, when trying to understand the world, to think, "markets are pretty efficient, do they agree with me? If they don't, can I think of a real reason why this may be a case where they're wrong, or am I just engaging in wishful thinking?"

I think it is not usually helpful, when trying to understand the world, to be really mad about the EMH.

Not at all clear to me that AI models can do search-like functions cheaply enough to outcompete Google (in the very near future). Yes, quality is amazing. But cost per search needs to be very low.

However, there is the estate tax.

If you're a billionaire who does the "take out loans against your unrealized cap gains" trick, then you, you know... can't sell your stock. So then your stock passes to your kids -- who, due to the stepped up basis, yes, do not have to pay cap gains on that stock.

But there's a 40% estate tax.

Estate tax generally isn't very relevant even to the ordinarily-rich, because it has an extremely high deduction (about $27M for a married couple), but for a billionaire it's absolutely relevant.

Now, sure, if you paid both the cap gains and the estate tax you'd pay that much more taxes, but if you compare a normally-wealthy person (pays 15-20% cap gains and 0% estate tax) and a billionaire (pays 0% cap gains and 40% estate tax), it's obvious that the billionaire, eventually, pays a much higher tax rate.

Feels like LLMs could be used to scalably grade more complicated assignments than multiple choice tests.

Hard to tell the size, but based on the tree doesn't seem all that giant either.

We should just use fucking normal plastic straws. This is environmental priority number 1,000,000, it just appeals to people instead of vastly more impactful interventions because roughly 50% of the environmental movement is looking for minor inconveniences that they can personally endure so they get a feeling of atonement.

You don't drive economic growth by rebuilding infrastructure that already existed, but now in the wrong place.

You might "create jobs" that way in the general sense of stimulating an economy by borrowing against future monies, and that may be helpful in smoothing out a business cycle or avoiding a negative feedback loop, but "creating jobs" is only useful when the economy is understimulated (which it isn't at this particular moment, though I'm sure it will be again at various times between now and say 2100).

You increase the size of the economy, rather than just smooth out the business cycle, by increasing the net productivity of the population.

I think that the case that Patrick describes leads straight into the intended case for adverse possession. You need to extend the timeline a litte.

Okay, so Bob owns a house. Bob marries Jane in an ill-advised, quickly ignored ceremony, but never gets formally divorced. Bob sells his house (in a community property state) to Ted. Ted lives in the house for 20 years. Jane dies. Jane's son Rick goes through her stuff, realizes that she was still married to Bob, realizes that Jane had a legitimate claim on the house that Ted has been living in for 20 years.

This is a classic adverse possession situation. Ted has been openly and notoriously living in this house, acting as owner. If Rick presses his claim, it is likely that Ted can win an adverse possession claim (assuming that Ted lives in a place where adverse possession still works the way it basically did in common law).

I am not a lawyer, this is not legal advice.

Insurance has negative expected value in return for risk pooling.

If something has a high level of importance in your finances or for other reasons, it makes sense to buy insurance. If you can handle the risk yourself (typically because the investment is not a terribly large amount of your total investments, and you do not think risk for this investment is correlated with risk to your other investments), then it doesn't make sense to buy insurance.

House flippers could be anything from very small one-person operations who would be wiped out if they had to clear a lien, to fairly large operations who can absorb the occasional risk into their costs of doing business.

We'll see GPT-5 in a few months and that will be vastly more useful information to update your sense of whether the current approach will continue to work than anyone's speculation today.

There are freeways within the geofence in SF. My understanding is that Waymos will not drive on those freeways without a safety driver for now.

Maybe they are being stupid! You can sell your Apple stock or short it if you like. But people getting indignant about Apple's decisions are not, it seems to me, mostly offering dispassionate market analysis.

Companies have decided that some markets aren't worth the headache since forever. Countries have fine-tuned their laws to be more or less open to foreign products since forever.

The EU is acting like it has a natural right to have American companies bring every product to its market no matter how burdensome its regulations are: it doesn't.

If you were an astronaut and your mission was scheduled to use Starliner instead of Dragon, would you think:

a. Mission control doesn't care if I live or die

Or

b. Mission control is actively trying to kill me

I think you're talking about different things. It's true, the Earth is substantially non-spherical. But things like mountains are tiny irregularities on a smooth (not-perfectly-spherical) surface.

I think Dan performs a mild sleight-of-hand trick here: he asks why we don't consider this a bug when any other software would consider it a bug. But in fairness, the question was not, "Did this prompt have a bugged output," it's "Did it have a racially biased output," and that's a more emotionally charged question.

If I wrote software that choked on non-ASCII inputs for say a name, and then someone said, "Hey, that's a bug," cool, yes, fair. If someone said, "This is evidence of racial bias," I mean... I'd probably object. Even if there is some nugget of truth to the idea that there may be some implicit language bias which relates to race in there.

I think Dan does a decent job showing that there is some level of racial bias -- not on the level of "the model refuses to show asian people," but on a level of "the model conflates certain jobs with certain races," and that's fair. But I just found the lede of, "Why don't people admit there's a bug in AI" to be a little obtuse.

Certainly. But some are much more common than others, and if you're playing the odds, as you must in these kinds of events, we can say that it's much more likely for an inexperienced driver to have an accident on pitch black wet roads, than that they attempted suicide in this unusual way.

I think very few people commit suicide by driving a car into a river.