@Zephyr's banner p

Zephyr


				

				

				
0 followers   follows 0 users  
joined 2024 February 02 13:03:12 UTC

				

User ID: 2875

Zephyr


				
				
				

				
0 followers   follows 0 users   joined 2024 February 02 13:03:12 UTC

					

No bio...


					

User ID: 2875

Who said they had to be testable now? We can always look back in a year.

There's a few other possibilities to explain it too:

  1. Some stuff is cheaper, some stuff is more expensive; if the more expensive stuff is essentials, but the cheaper stuff is everything else, people can feel squeezed by the economy without actually being squeezed. (Picture a toy model where someone has $10, food costs $1, shelter costs $1, and consumer goods cost $4 each; if the costs of food + shelter go up to $2 apiece, and consumer goods go down to $2 apiece, each person can individually afford more consumer gifts, while still having less space in their budget for consumer gifts).
  2. People are able to substitute goods for cheaper substitutes, which means that their spending is relatively unimpeded, but they are accepting a lower return on each purchase they make. (So for example, maybe I used to buy steak at $4.00/lb. Steak is now $8.00/lb, but ground beef is $4.00/lb, so I now buy ground beef).
  3. Everything went up by a specific percentage, including wages; wage increases feel like a benefit of me working hard, while price increases feel like people trying to take more money from me. As a result, I feel like I'm not doing as well, because my hard work to get more money was invalidated by everything else becoming more pricey as a part of it.
  4. People have given up on certain staples like housing; as such, instead of saving for housing, they're spending the money they would be putting towards it towards consumer goods.
  5. Some demographics are spending way more, while others are spending way less.

Here's what I would assume would be observable in each case:

  1. Breakdowns of spending will reflect higher percentages of household income spent on housing and food than per usual. This would be disprovable if the percentage is similar.
  2. "Budget" companies have their share prices improve, while luxury companies start offering more budget goods. Companies that only offer luxuries start struggling.
  3. Average income goes up by a similar percentage basis to costs. This should be observable based on the fact that people have published stats on average income for many years, so I assume that someone somewhere is observing it.
  4. The boom in consumer spending is almost exclusively in luxuries; it would also die off in a few months to years, but that's probably indistinguishable from the natural flow of the economy.
  5. Spending is concentrated amongst goods that some portions of the population use. For example, if healthcare spending is way up, but bars are struggling, I'd consider this to be met.

Bigger screens for watching videos. As phones become more and more of a computer replacement, the incentive to maximize screensize goes up.

How do you feel about photographers? Are they not allowed to take credit for their photo, given a senor put all the pixels where they are?

They can absolutely take credit for their photograph - what they can't do is call themselves painters. If self_made_human wants to call himself a prompt engineer, I'm not going to stop him.

Meanwhile trans women usually report the opposite and their mental health is improved from the exact same hormones. Weirder anecdotal reports are cis men complaining of brain fog from taking oestrogen, while trans women saying the hormones actually lifted their brain fog.

Would you mind providing a link to this study? I've heard the opposite from the recent controversy over the "mermaids" charity and Cass review, so I'd be interested to see the other perspective on it.

"you have to actually read and rewrite the AI's output"

I think one of the issues is that people won't read the AI's output.

So earlier this year, I was applying for jobs - originally, I wrote each cover letter individually for each posting. This slowly made me suicidal, as spending 15-30 minutes per job application where I was unlikely to even hear back from the majority of them was soul destroying. The next thing I did was take a "template" cover letter, and swap out a few things (so like, in my "accomplishments" section, I'd rewrite it to emphasize the skills the job requested). This took around 5 minutes per job application, and was still soul destroying, because I still wasn't hearing back from very many jobs. So eventually I started pointing ChatGPT at the cover letters, and I promised that I'd rewrite it every single time.

Well, that lasted around 5 attempts until I basically got sick of it and started skimming. I went and took a look at some of the cover letters I "wrote", and about a third of them have obvious ChatGPT-isms like emdashes, that specific phrasing half-fawning phrasing that ChatGPT uses, etc. Thank god resumes were being read by LLMs too, or I'd never have gotten a new position.

Humans are lazy - they're going to take the path of least resistance every time. They'll claim that they read the whole thing, and for some definition of "read", they will have - but they'll be stuck with the LLM's phrasing and concepts.

Here's an example I fed into ChatGPT for rephrasing (my words first):

LLMs will introduce their own biases into the resultant writing. If everyone uses them, this will lead to less ideological diversity as every person will be essentially arguing as an AI, not as a human.

Because LLMs impose their own biases on generated text, universal use could erode ideological diversity, with people arguing through an AI lens instead of their own.

You can obviously tell that the concept is the same, but there are subtle differences in the meanings. If I were writing a larger text, I'd probably accept the AI text as "essentially the same thing" - but they're not. My text is much more emphatic about it being what will happen, whereas the LLM text is downplaying it. Multiply this by a much larger text, and you have an entirely different emphasis.

Found the link, if you're curious:

https://www.themotte.org/post/2732/culture-war-roundup-for-the-week/358520?context=8#context

The quote I was thinking of:

Boys (believe it or not): They won't admit it, but they'll watch. When their sister’s watching it, they'll balk and act like it’s dumb, then they'll sit down and watch it. For the same reason Moms will find My Little Pony interesting enough to happily share with their daughters, the compelling conflicts, the strong characterizations, the silly humor and (most importantly for boys) the ADVENTURE, the boys will watch, too. Really

To be fair, mine was supposed to be a joke about how many men exaggerate their - let's call it "physical attributes" - upwards.

Also, they generally don't take female oriented IPs and change them to make them more attractive to a male audience.

I know you said "generally", but I seem to remember reading (probably here) that the My Little Pony phenomenon was due to them doing exactly that. I vaguely recall reading something that basically said that the way they'd designed it to appeal to young boys was by including a lot more adventure/hero's journey elements than are usually present in girl's media, and I explicitly remember them mentioning something along the lines of "we know boys won't go out of their way to watch it, but if it's on because their sister is watching it, we want them to watch it too."

Which plays a lot more into @The_Nybbler's point - it is definitely possible to make media that is intended for men, and extend the appeal to women as well without compromising what men like about it. Which implies that they are making it woke because that's what they want to do, not to expand their audience.

Also height is pretty easy to lie about.

My conspiracy theory as to why there are more men in the trades than women is that women constantly hear a bunch of false measurements, so aren't able to eyeball it (after all, if you always hear that 2.5" is actually 6", and 5'8" is actually 6', then how are you going to place nails every 8")?

Fair enough, I'd agree with that.

But touch someone's butt once in a bar in the '80s?

Obviously, that's why Trudeau was kicked out of office for wearing blackface and for groping a reporter.

It's all who/whom.

The issue is that the more you push policies that are "fuck you for not having a girlfriend" at young men, the more likely you are to get young men who choose violence. If a man can't afford housing, so can't actually end up in a stable enough situation to attract a partner and have kids, then there isn't a good reason to not just rebel against the entire system.

If we're going to punish men who don't have a partner more than not having a partner is a penalty in and of itself, we're going to disproportionately be punishing men who are more educated (quick googling shows that men who have higher education levels are more likely to be both unmarried and virgins).

Yes, it's sleazy, but c'mon: you've been telling us for years that he's sleazy and corrupt.

This is almost more important than everything else - Trump being sleazy and corrupt is already priced into him as a candidate. If you provide more examples of it, the base is going to say “so what?” - they already know all this.

While you’re absolutely right - do you think that means anything if the press/democrat influencers want to make hay out of it?

I swear, people who are not me have to have been using entirely different LLMs than I have. Every time I've used them for anything beyond the super trivial, I get results that are missing major components, or don't solve the business requirements, or contradict themselves. Like, I just opened up cursor and asked it for the CSS to render rounded corners in outlook, and it got it entirely wrong (it produced non-rounded objects because it used border-radius, which anyone who has coded anything for outlook knows doesn't work). When I told it that, it produced different code that (and I can't stress this enough) still fucking used border-radius.

Seriously, at this point I'm more likely to say that the person using the LLM is going to stall the project; they're going to produce verbose but contradictory requirements, they'll produce code that is written fast, but doesn't actually work, or they'll use it to answer emails in such a way that it doesn't actually answer the question that is stalling the damn project.

PHP is even better about not stopping if it thinks you made a mistake - try to open a file that doesn't exist? Yeah, sure, fine, just return false. Loop through false? Of course, obviously that's intended behaviour. Mix and match numeric and stringy keys in an array? No problem, it's a hashmap, and it'll even sort for you.

If Python and C are the languages god wants us to use, PHP is the language he uses himself.

(In case you couldn't tell, /s).

The fun part is that Canada's is also notionally (though not actually) paid by a tax paid during your working years; the issue is that like any government, ours spent all the money that was supposed to be used for it on other things.

Perhaps a valuable lesson can be learned here about voting for people who spend more than the country makes, and have to take the money they promised you to fund it is in order here.

Sure? If you're awarding people extra "points" based on how likely they are to be in jail, I'd call the midpoint of the curve when it swaps from "benefits" to "discrimination against", which (given that population trends are approximately 50% female) would be at the gender level, not the race level, but if you want to call anything that isn't "maximally in favour of" as "discrimination", then yes, that would be discriminating against white/asian men.

Like, I don't understand the point you're trying to make here; our current system is that the more progressive stack points you have, the more affirmative action selects for you; so "white straight male" is bottom of the pack, while "queer PoC who identifies as female" is the top; I'm stating that it is extremely backwards from how it should be if the jobs are simply busywork to keep people out of jail, as "male" is like, 90% of the weighting towards criminality there.

To be clear, I'm stating that I do not think there should be affirmative action at all; I think that we should hire based on merit. If we're not hiring based on merit, but instead based on how much we can keep people out of jail, then we should hire based on male vs female, then race, then sexuality.

I'm really not getting what you're saying here - it seems to me that you're proclaiming that DEI is already a program for this, but ignoring the fact that it is doing a terrible job of it.

I mean, you could just stop subsidizing it, rather than actively punishing the elderly who do not have children.

In Canada, we have programs that are explicitly a transfer from the working aged to the elderly (OAS); these programs have insane cutoffs (OAS is only fully cut off at an income of $180k/year, and doesn't check existing assets; it's very possible for someone elderly to own a $3m+ house, and still receive the full amount of the OAS). Cutting that off would increase the amount of money in the pockets of the young, improve housing availability (as it forces the elderly to sell their oversized homes to have funds), and not punish those who didn't have children; but instead just punish those who chose not to plan for their retirement.

Seriously, I'd be homeless and living on the streets if I were jobless - why are we saying that the least productive members of society get immunity, and to keep their assets, while the most productive (by which I don't mean me, but the young in general) have no safety nets at all?

No - I favour a world where corporations hire the most qualified > the less qualified. I'm simply stating that if the point of jobs is to keep people who would otherwise be criminal busy, then it should favour men over women.

Honestly, you've convinced me; I hadn't thought about it that way, but you're absolutely right.

Why? White men are more likely than any demographic of women to be in jail (I found the number of 158/100000 when googling for white men, 88/100000 for East Asian, and 68/100000 for black women). Seems to me any attempt at making jobs to keep people out of prison should discriminate against women regardless.

Does that matter, though? If everyone sees that the married men are getting all the best assignments, and get constantly let out early to go pick up their kids, and are paid the best - then it doesn't actually matter if it's official or not. Everyone knows what needs to be done to get the benefits.

If you're paying people to stay out of jail, pay the people who disproportionately go to jail.

So, men?