Nah, your reading comprehension is just fine, I just worded myself in a pretty verbal way. I totally get how it could be misread, I was thinking along the “salesman making a pitch” line:

“Want something that folds perfectly to A4? A3. Want an extra fold before you get there? We have A2!”

Yes, I did a typy-type with chubby morning fingers

I fat-fingered the A5, otherwise the numbers were right?

I’m aware how the system works, I just wrote about it using A4 as the “base”. A3 is what you fold to get A4, A2 folds to give A3, etc.

I didn’t know A0 was exactly 1m² though, TIL!

Oh, let’s also remember that A4 is a member of beautiful variant of an “SI system for paper”. Want a paper the size of an A4 folded double? We have A85. Want something that folds perfectly to A4? A3. One more fold? A2.

• Statement: “I want a piece of paper that I can fold twice to get exactly a US letter size.”
• Response: “You must be crazy.”

Edit: Did a fat-fingering

A garden hose is about 2cm in diameter, you’re gonna need a pretty massive pump to get any flow rate to speak of through that hose. I’ll be back with numbers for you ;)

Well that’s exactly what I said that I seem to remember reading? So provided I’m remembering correctly… yes?

with no saying …

Well that’s the neat thing when paying money for something: You get a contract detailing exactly what access you get and what you can use it for. You typically read and sign this before paying.

Requesting paid access for research purposes would probably get you there though. After all, they need to sell access to posts and comments through some API to make money off all that data.

I can’t quite find the source, but I remember seeing a plot of the estimated fraction of posts and comments on reddit that were produced by bots (not estimated by reddit themselves of course). The conclusion from that blog post was essentially that, at this point, reddit is mostly bots interacting with each other, with some humans sprinkled in. Of course, this doesn’t matter to reddit as long as they can convince the world that they have plenty of users they can show ads to. Companies buying ad space largely don’t do the research needed to find out that they’re paying for their ads to be shown to bots, so they keep buying, reddit keeps making money, and everyone is happy.

There was definitely a massive migration, at least seen from lemmy’s side, check out the graphs at Fediverse observer. Within a two month span (May-July 2023), the number of total monthly lemmy users went from 47 000 to 1 700 000. That’s a 37-fold increase in two months. The number peaked at about 1.9 million in august the same year, with about 130 000 active monthly users (up from about 1 000). In the same period, the number of online lemmy servers went from 80 to a peak at 1396 (a 17-fold increase).

In reddit terms, the migration wasn’t massive, they lost on the order of a hundred thousand users. Still, over a hundred thousand users leaving a platform within a couple months is probably near-unprecedented in internet history.

You do realist that “I am innocent, armed and will shoot if you try to arrest me” is possibly the worst conceivable way to avoid being unlawfully arrested, right?

There is quite literally nothing you could do to make it more certain that you will be either killed or arrested than trying to shoot your way out of being wrongly arrested.

I hope you also realise that the US has the highest incarceration rate in the world, and has a long and proud history of arresting and/or executing the wrong people. Literally any other system than the one you’re advocating for is provably superior when it comes to preventing the wrong people from being arrested.

Sometimes they literally don’t have the ability to press 1.

Which is why I’m saying it should go through automatically after a short delay.

It’s easy to say “just over staff” until you’re faced with the reality that maybe once every five years, for a grand total of one hour, you’re faced with call pressure 1000 times higher than the second highest spike. It’s not viable to staff a response center to deal with thousands of people calling in the same bomb attack or shooting within minutes, when you normally see on the order of hundreds of calls per week.

What you need is the capacity to deal with 99.99… % of situations, and a safety mechanism to prevent the remaining situations from overloading your system more than necessary.

In a mass casualty event, you need to be able to filter out the people making redundant reports in order to effectively handle the important reports. You need to do that without blocking out the important reports, that’s the hard part.

You really don’t need an AI for this though. With the number of mass shootings the US sees, it’s honestly shocking that they don’t have a button the operator can press to activate an auto-reply along the lines of “We are already responding to <event>, if you still want to talk to us, press 1 or wait for a reply.”, which waits something like 10s before going through.

This would filter out a bunch of duplicate reporting with far less effort, and with less odds of something going wrong.

One thing is the 58 000 that have to re-take the exam. I’m just as interested in the 2% of exam-takers that were terminated due to “unspecified conduct issues”, and how many of those were unfairly thrown out for cheating without having done anything wrong…

So much the last bit! I can’t count the number of times our CEO has either sent a mail or created an issue (very small company) with something along the lines of “X is broken”, when what they mean is “I don’t like the design choice behind X.” Like, whatever they’re complaining about is completely functional, it just doesn’t work exactly the way they envisioned it. When explaining why it can’t (or shouldn’t) work exactly the way they envisioned, there’s maybe a 50 % chance that “X is broken” turns into “oh, I hadn’t thought about that”.

It honestly shocks me how stupid people are. There was just a case in the news about some guys that were busted while planning an attack on the White House UFC thing recently. They had initiated contact about this on tiktok (before moving to signal). One of the guys had posted photos on instagram of training with one of the others. The first guy that was busted had been reported by his parents, who he lived with, because they had overheard phone conversations that made them worried. They had anlso noticed he had begun stockpiling weapons, ammo, and other gear. All of them had quit their job to make time for the planning + attack.

Like, I’m glad these people were caught so that a bunch of innocents weren’t killed. What baffles me is how intensely stupid it’s possible to get. What would make anyone even consider communicating about this kind of stuff on a digital platform? How does anyone think it’s a good idea to suddenly start stockpiling weapons in broad daylight? In what world do you document this training online? Why in gods name would you quit your job out of the blue while doing all the above?

We’re just lucky these people are so stupid…

This makes me tremble. I cannot imagine starting every day with breaking some obscure tool, pipeline, or tool chain just for the hell of it

That’s definitely true, but the question is do the execs really care? I think that for a lot of these people, the only thing that matters is whether they can keep pulling those sweet sweet cash-outs. Just look at the absurd bonuses musk was promised from tesla recently. If he really cared about the success of the company, he wouldn’t take those, he would take a fair paycheque and allow the company to reinvest the rest of the money. Instead, he requires massive bonuses to keep working. We’re talking about the kind of money that could fund the entire educational sector in a small country for many years. He’s taking that out as a personal bonus, to the detriment of the company.

That kind of thing makes me believe that he doesn’t really care about the long-term success of the company. What he really cares about is squeezing out cash from the company for as long as possible. If the company fails, he has enough money to buy up something else that he can squeeze cash out of. The modus operandi is basically

1. Be rich
2. Buy some company
3. Wave your arms and wag your tung to get investor money into the company
4. Cash out bonuses
5. Go to step 3 until your cash-out has surpassed the investment cost
6. Either sell out of company (if it’s been run to the ground), or go to step 3 until it has been.

I honestly think a lot of the people pushing this stuff know very well that they don’t need it to succeed. They don’t really care if the general public hates it, and it flops. As long as they can convince investors, they’re pulling yearly paycheques and bonuses in the tens of millions of dollars. If it flops, some thousand employees (thralls) will be layed off. If it fails catastrophically, they might need to step down themselves, and move to a different company that “appreciates their ability to be a visionary”.

These people are only capable of failing upwards, and they know it. The name of the game for them is waving their arms and blabbering about something to draw in investor money.

I think it makes sense to be flexible: In a table, or other bureaucratic contexts, it makes sense to put the family name first. In daily speech, it’s rather common that I’m in contact with family members (even more so historically), and it makes most sense to use the distinguishing name (first name) first.

If I’m with a group that includes siblings or parents/children, I can usually distinguish everyone by first name, while many people share last names.