derf_

As the political arm of the government chooses to run deficits in excess of growth plus inflation, then (a) that causes more inflation, and (b) the central bank raises rates, increasing the cost of government borrowing, causing bigger deficits.

This escalates as a result of the central bank trying to control the effects of high government spending by applying a mis-matched policy tool (interest rates) in place of the politicians who have abdicated their responsibility to use a matched policy tool (taxation). Either it spirals out of control, or more and more of the government budget is devoted to interest expense (direct government transfers of wealth paid exclusively to the debt holders) and less of it is spent on providing actual government services (that benefit all taxpayers).

If the central bank does not raise rates, of course, things go even more badly.

Vibe coding does not solve this problem. If anything, it makes it worse, since you no longer have any idea if an implementation might read on someone else's patent, since you did not write it.

If your agent could go read all of the patents and then avoid them in its implementations and/or tell you where you might be infringing them (without hallucinating), that would be valuable. It still would not solve the inherent problems of vagueness in the boundaries of the property rights that patents confer (which may require expensive litigation to clarify definitively) or people playing games with continuations to rewrite claim language and explicitly move those boundaries years later, among other dubious but routine practices, but it would be something.

If South Korea had followed the standard developmental economists' advice they would still be sewing garments and growing soybeans instead of manufacturing semiconductors, consumer electronics, and appliances (among many other things).

[0] https://statranker.org/economy/leading-global-car-manufactur...

This is demonstrably false? Long-term average US inflation since 1913 is 3.1% [0]. Long-term nominal average US stock returns since 1928 are 9.94% [1]. A nearly 7% advantage compounded every year for roughly a century is not "slightly above", it is absolutely enormous. Over 60,000% enormous.

Furthermore, when inflation is high, interest rates go up, and interest rates act like gravity on stock prices. See any number of Warren Buffett shareholder letters. See also: the year 2022. Stock market returns are mildly negatively correlated with inflation (with a coefficient of -0.229 [2]).

[0] https://inflationdata.com/Inflation/Inflation_Rate/Long_Term...

[1] https://awealthofcommonsense.com/2025/01/historical-returns-...

[2] https://www.forbes.com/sites/rmiller/2024/06/20/90-years-of-...

[0] https://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD?locat...

The kind of developers who are willing to work with native GUI APIs (even though a framework) simply don't exist in enough quantity anymore.

Also, to quote Jastram [0], "In spite of the romanticism of the Spanish Main and treasure ships, the supply of gold from the New World was just a trickle by later standards. Less than 1% of what was to be 1930 world production was produced in each of the years from 1500 to 1520." Over the next 80 years, it never got much above 1.3%.

[0] Roy W. Jastram, The Golden Constant (1977), pg. 41.

Also, to start compounding, you have to have enough to compound. Most people instead start off significantly in debt (student debt, mortgage/rent, etc.). Then they try to save enough to be able to raise their children and retire before they become too infirm to work, at which point they start compounding in reverse. All the while, they are competing with everyone else trying to do the same thing for how much they are willing to pay for scarce resources and how little they are willing to receive for their employment. Having your compounding achieve escape velocity is the exception, not the rule, and it has to be that way.

[0] https://en.wikipedia.org/wiki/The_Missing_Billionaires

> ...consider a case where finance becomes much more productive... leading to fewer people becoming doctors because finance is much more attractive.

This is the opposite of what one would expect from a sector whose efficiency increases, as modeled by Baumol. See the first bullet in the article: "The share of total employment in sectors with high productivity growth decreases, while that of low productivity sectors increases" (also see the detailed analysis in the Technical Description section). It might be theoretically possible that induced demand could still increase overall employment in a sector as its efficiency increases, but I think you have to make an argument why that would be true. During the industrial revolution, automation eliminated 98% of the labor required to produce a yard of cotton cloth, but between 1830 and 1900 the number of weavers in the US increased by a factor of 4, because demand increased due to lower prices [0]... although the US population also increased by a factor of 6, so as a percentage of the workforce weavers still declined, even as people consumed much more cloth per capita.

[0] James Bessen, Learning by Doing - The Real Connection between Innovation, Wages, and Wealth (2015), pp. 96–97.

The positive news (if it can be called that), is that this level of growth cannot continue, so something will have to change.

There is literally a section that begins, "What will be the useful life of AI assets?" In bold.

> BRK-B's 5 year return is 2% per year more than SP500.

That said, 2% per year is generally considered a lot in diversified mutual fund / ETF land. You might not be able to charge hedge-fund level 2-and-20 fees for delivering that, but you could certainly charge multiples of what the low-cost indexers do (instead, Berkshire charges you approximately nothing). Now, Berkshire is a conglomerate, not a fund, and you could argue 2% is an appropriate risk premium for a single stock, even one as diverse as Berkshire (which is still less diverse than the S&P). But it is pretty impressive for something that is not a tech company. Those are the only things in the S&P that seem to be generating any returns these days (besides Berkshire, JP Morgan is the only other non-tech company with a market cap over $1tn, and arguably banks really are tech companies now, too).

> Apple is still 21% of BRK's publicly listed holdings

The public company investments are a minority of Berkshire's current value. The majority comes from wholly-owned operating companies and insurance businesses.

It does not matter what your income is if you cannot budget and save.

I don't want to discourage anyone from submitting patches, but that does not necessarily remove all (or even the bulk of) the work from the maintainers. As someone who has received numerous patches to multimedia libraries from security researchers, they still need review, they often have to be rewritten, and most importantly, the issue must be understood by someone with the appropriate domain knowledge and context to know if the patch merely papers over the symptoms or resolves the underlying issue, whether the solution breaks anything else, and whether or not there might be more, similar issues lurking. It is hard for someone not deeply involved in the project to do all of those things.

Shine on you crazy diamond, and all that, but...

> I have been continually confused as to why it is being forced down everyone's throat.

Have you never sat on public wifi and tried to open an http site? These days it is highly likely to be MITM'd by the wifi provider to inject ads (or worse). Even residential ISPs that one pays for cannot be trusted not to inject content, if given the opportunity, because they noticed that they are monopolies and most users cannot do anything about it.

You don't get to choose the threat model of those who visit your site.

Or, put another way, when the bills come in, people make their car payments first, because they have to get to work. So either there are other bills people aren't paying, or people aren't going to work. Both seem bad. One purported cause of increased auto loan defaults is that people are having to make student loan payments again, although that would suggest they aren't prioritizing the auto payments.

1. You can change the training data.

2. You can change the objective function.

3. You can change the network topology.

4. You can change various hyperparameters (learning rate, etc.).

From there, I think it is better to look at the process as one of scientific discovery rather than a software debugging task. You form hypotheses and you try to work out how test them by mutating things in one of the four categories above. The experiments are expensive and the results are noisy, since the training process is highly randomized. A lot of times the effect sizes are so small it is hard to tell if they are real. The universe of potential hypotheses is large, and if you test a lot of them, you have to correct for the chance that some will look significant just by luck. But if you can add up enough small, incremental improvements, they can produce a total effect that is large.

The good news is that science has a pretty good track record of improving things over time. The bad news is that it can take a lot of time, and there is no guarantee of success in any one area.

To quote Richard Bookstaber, "The principal reason for intraday price movement is the demand for liquidity... the role of the market is to provide immediacy for liquidity demanders. ...market crises... are the times when liquidity and immediacy matter most. ...the defining characteristic is that time is more important than price. ...diversification strategies fail. Assets that are uncorrelated suddenly become highly correlated, and all positions go down together. The reason for the lack of diversification is that in a high-energy market, all assets in fact are the same.... What matters is who holds the assets." (from A Framework for Understanding Market Crises, 1999)

Was the market drop an accurate reflection of the value that would have been destroyed by those tariffs, discounted by the probability that they would have been enacted as drafted? Nobody knew then, and I maintain that nobody even knows now. That was not the calculation that was being made.

And just as many where people explain in detail exactly how to do foo with bar without explaining why I would want to do foo in the first place and what a bar even is.

Even in a language that is not "good" by your definition... you have basically just described a function. A wrapper function around a sub-function that has early returns does everything you want. I use this pattern in C all of the time.

They do, but it is only paid by the people who took the money (instead of being forced to do so), and, more importantly, only on the difference from what they paid.

If you pay out $1mln of dividends, then everyone collectively owes (let's say at a qualified rate of 20%) a total of $200k. If you buy 20,000 shares from one guy at $50/share, you returned the same $1mln of cash to shareholders, but if he bought last year for $45, he only owes (let's say at a long-term capital gains rate of 20%) a total of $20k in taxes.