Yeah. Capitalism has never been less about actual profits and efficiency than it is right now. They just tout whatever random thing they're currently doing, regardless of why they're doing it or whether it was voluntary or not, is great and producing amazing results.
RTO is primarily pushed by the capitalist class, and managers are just stuck in the middle. No good manager wants pissed off employees, but managers who push back on the capitalist class do not stay managers long.
"If being in the office conforms to the interest of the capital class, it implies that WFH is inherently less efficient"
only if the capital class is solely motivated by efficiency. I think this is trivially demonstrable to be not the case.
The capital class's primary interest is self-preservation - both of their capital, of course, but also preserving their place in the pecking order. And they'll spend a LOT of the former to maintain the latter because the latter is how they got the former.
Through that lens, GP's point is perfectly coherent.
"They seem to imply that business owners just want employees to suffer as a goal in itself."
Have you met... people? Yes there are literally many owners who do want employees to suffer. Or, perhaps worse, will tolerate tremendous amounts of suffering in the pursuit of minor other gains. (Amazon pee bottles come to mind.) It would somehow be a comforting kind of moustache-twirling comic book evil to say they just want people to suffer. Another to say they simply don't value human happiness (or lack of suffering) enough to not trade large amounts of it for small things they do care about.
I had a boss who was only willing to hire non-whites because he could inflict undesirable work on them, leaving more desirable work for the white employees.
I just want to end this by remarking that this presumption of owners being perfectly optimal, morally clean agents of free markets is absurd and honestly disgusting to bring to an argument.
Shorting stocks has a very high "you must be at least this right" bar in order to make money. And given the uncapped nature of losses - the market can remain irrational longer than you can remain solvent - you need to also be really correct about how high it'll go before you're right, and also, you're borrowing these shares you've sold, so you're on the hook for the borrow fee and also on the hook for paying dividends paid out to those shares you're borrowing but not holding. Plus you have to pay for the margin loans you're using.
That's a very high set of both static and scaling costs that eat away any profit you made by being nominally correct. Combined with the risk profile... you can't "just" go short a stock.
And yes, you can hedge losses with options or construct complicated options positions to try to hone in on a specific price movement you're anticpating. Now you have to deal with entering and exiting a complicated multi-instrument position without price slippage, AND you have theta decay and volatility-related price movements also eating away at the core money you're making by being nominally right.
Have people made money? Yes, for sure. There's also a lot of dead bodies and people who barely broke even despite theoretically having been right.
"yes exactly. Too many people ask AI to one-shot complex tasks, and wonder it behaves like a junior asked to rush something."
Because this version of AI is worth 10 trillion dollars.
While the pragmatic versions from realists you can find all over this thread are ultimately probably less of a speed boost than just having your CEO/local micromanager be conveniently on vacation during critical periods when the work actually gets done.
It's funny to me to imagine that the whole time humans were doing basically anything on this planet, nuclear fission was also already happening in a few places around the world. I wonder how much science would've been jump-started if we'd found any of the natural nuclear reactors prior to having figured fission out already.
Somehow one of the most fascinating aspects of LLMs is that they were able to basically both reveal corporate capitalism as primarily being about virtualized feudalism, and also casually ruin the entire world, just by their ability to generate decently grammatically correct English.
Why is it that all of these bigheaded corpo type people react to "anti-AI" sentiments with this strongman crap?
Are they surprised that people, especially young people who are going to head into the world to try to make their way, are NOT supportive of the idea they'll just be eternal slaves, or even unemployed nuisances, to feudal lords?
The whole fucking point of this system is for it to support people. "They" need to remember that or they're going to be reminded of it. Capitalism was cool as an evolutionary step but it's pretty fucking clear we're going to need another innovation here sometime soon.
Honestly there are so many levers we could pull off within Capitalism to both meaningfully improve the lives of people while keeping together all the things that are genuinely good about it (such as the central thesis of rewarding risk and innovation).
- better safety nets paid through higher taxes and closing down loopholes (such as buybacks)
- stronger regulatory enforcement to reduce the power of monopolies / monopsonies
- stricter rules around tax-payer investments / subsidies into companies (sure SpaceX gets to have billions to shore up our competitiveness but there must be a mechanism for the tax payer to have board representation)
etc. etc.
The list goes on. Honestly it doesn't take a lot (besides political will and escaping the capture of Citizens United) to move the needle so that the median income or bottom 10th percentile lifestyle meaningfully improves while we no longer have wealth horded to the level of having centibillionaires.
It does require a government that operates with much less corruption than we have, so it is more idealistic than not, but very easy to imagine.
This feels very adjacent to the story about the whole town in debt, and the rich guy leaves a $100 bill on the table, [and so on], in a way that I can't quite put my finger on.
> True, at the beginning each resident has a $100 liability. But each also has an offsetting financial asset of $100. At the end, they all have neither. So the $100 bill acts as a clearing mechanism
You can't put your finger on it because money is merely an accumulator and medium of exchange of economic performance. The performance of services in exchange for other services without money is a perfectly valid economic exchange that can and should be booked to revenue of each of the parties, if actually performed.
Loans without any economic performance of services generate circular meaningless cash flows yeah, but that's not the case when services are actually performed.
Loans are promises to pay. Business deals are promises to perform services or deliver goods. The difference is easily lost in the details even for accountants and economists.
When comparing promises between businesses to pay versus promises between businesses to perform services, it is irrelevant that fiat currency is a federal reserve note rather than, say, bottle caps. Irrelevant.
Not quite. At least the one I found is some trickle down economics myth.
The one op is referencing is more like the dollar is used to pay off the waitstaff, who pay their rent to the landlord, who pay their over due taxes, so that the government can issue a refund to the cafe owner. The dollar ends up back in the hands of the cafe owner, who puts it back down on the table with all the debts paid off.
The problem is that aquifers are really cool natural filters, and only refill as fast as groundwater moves through the soil. So they're a finite resource. Instead of depleting them, people who want to farm in deserts should probably start desalinating or whatever themselves instead of assuming subsequent generations will do it.
The government made it literally the only way to claim much of the land out west[]. They require that you come up with an agricultural land including plan for watering crops on that acreage in order to claim the land. And you're required to execute the plan to get the deed.
In fact, this is the only remaining way I know of to more or less 'homestead' federal land in a way that results in a permanent deed. The rest of the homesteading type stuff was revoked back in like the 70s or 80s.
AFAIK it's been available since the 1870s but after the 20s they clamped down a lot harder on ensuring you were actually irrigating it and had agricultural plans.
I'm not sure if the BLM has relaxed their discretion under Trump.
Do you think laws go away just because they're old?
The Colorado River compact came into effect in 1922 and I'm almost surprised literal fist fights haven't erupted over it during the modern negotiations.
I’m no lawyer, but according to my perusing (sourced directly from the BLM), the 1877 Desert Land Act is very much obsolete.
The age of a law does not matter with regards to its validity, you are pedantically correct. But it very much matters to its relevancy, which was my argument. Laws regarding horse traffic in Manhattan may still be valid, but a lot less relevant than they were hundreds of years ago - assuming they haven’t been repealed.
Regarding the “Colorado River compact”, I would say my qualifier of “quite niche” is important. Ownership and water rights over the second largest watershed in the US by affected population is far from “niche”.
On the other hand, how settlers can claim public land in the desert (which happens incredibly rarely now, by design) is quite niche.
Even if you presume it's obsolete, the fact remains the land was distributed under a model where it had to be claimed, developed, and irrigated for agriculture. The water distribution and well systems alone required for agriculture far surpass use for residential purposes, for example, but once you have wells and irrigation suitable (not cheap in the desert, and also valuable) for agriculture it may be more economical to just keep going with it. The momentum of that remains and the economic incentives aligned with those policy choices remain with us as to part of the reason why desert land is used as it is. The first developed use of land can have a lot of influence on what happens next, it's quite different from starting from an undeveloped state with pure free market forces on development.
I find it does not do very well with go at all. I speculate that it's partly because go is going to be a language you find a lot of concurrent programming happening in. And sure enough, I find even the best claude is nearly useless at anything beyond copy-pasting examples out of the docs for goroutines.
My own experience with agents, I'd summarize as "the more the world model (which the LLM does not have) is not concretely represented by the text, the worse LLMs are at it."
So it's _great_ at HTML, CSS, markdown, and most cursory-inspected English. Good at javascript. OK at most languages. Then very bad at concurrent programming and closely-inspected English.
I also don't think your top-line conclusion is right at all. I'm quite the opposite opinion. The types "working out" does not actually give me hardly any conviction that the code actually works. And notably, LLMs seem good at making types work out (they're in the text!) but then still have code that's not actually at all right (for the world model).
I also find that types are not worth the often COPIOUS amounts of boilerplate that comes with them. Some of the worst code I've seen is using reflection to make something happen that would otherwise barely be metaprogramming in Python or Ruby.
But that's not to say types are useless. I just think rigorous static typing is not worth it. My current favorite way to program is Python, with an enthusiastic use of type hints, enforced by a good type checker (pyright). It gets you 99% of the benefits of traditional static typing, but you can also just tell the type checker to just look the other way for a moment if you're going to commit a dynamic typing.
my experience with elixir is that if you tell the LLM "let it crash" it is productive. If you don't tell it this it hides errors and writes crappy tests. So, maybe it is go and not concurrency programming.
reply