fluoridation

Redundant question. Everyone has biases.

Yeah, that's true. I don't understand why they do that. People don't write like that, we're pretty economical. It's not even that they're repetitive to emphasize a point, they just engage in horrific circumlocution. Is it, like, that the facts of the input only percolate through the network when they emit a token, so they only "think" when they're outputting something? I don't know if that's it, but that's what I would think if I caught someone writing like this, that they're writing faster than they can think, so they fill the void with vacuous noise.

Now it's an AI mistake, and people HATE AI mistakes. Even if the number of AI mistakes is fewer than the human mistakes.

That's because AIs make mistakes no human would. If you're documenting a function, you might misspell something (or orthographically spell something that should be misspelled), or confuse two functions with similar names; you won't reference a function or feature that doesn't and has never existed.

I assume by "proof-reading" you don't mean just checking that the text doesn't contain spelling mistakes or is incoherent, but including that the content is actually correct. If your people actually invest that much into documentation, what the point of the LLM? Isn't the actual writing part like 10% of the total effort expended, in terms of man-hours?

Not an answer to my question. If you feel ganged up on, perhaps it's time to stop posting.

No one who actually was a Scout would call their Scoutmaster a "troop manager".

I was a scout and I natively speak Spanish, so I never learned what the ranks are called in English (and after nearly 30 years I've pretty much completely forgotten them in Spanish too, if I'm being honest). Are you going to call me a liar too?

instead of allowing anyone to train on already scanned books for free

That would be pirating. So your complaint is that they didn't do more piracy?

A rate is just a ratio between two quantities. You've never heard speed defined as the rate of movement?

I might have interpreted your example as such if not for the "one too many" comment. It's strange to throw out a completely fictitious figure and then lament it as if it's real.

"Rate" is an ambiguous word. Molested children per year is a rate just as molested children per total children is.

something like 1 per 100,000

What is this? What number are you citing, and where are you getting it from?

Buddy, you are fighting windmills. You should go lie down.

And those groups still have lower rates of abuse than Schools

Is that in absolute terms, or per child who frequents the establishment?

Uh... Huh? Mate, nobody cares if you want an AI to feed you regurgitated summaries.

It's TTS. That doesn't mean the text wasn't written by a human.

No, I meant what I said.

Just because it's approximated by a bell curve doesn't make it a bell curve. There are quite obvious separate phenomena shaping the curve at different times.

Definitely this. The moderators seemed to have the Lock Question button connected to their dopamine pathways.

It's not really a bell curve. There was obviously a downwards trend from 2016 onwards, but 2023 definitely precipitated the fall to zero. Without AI they might have lasted at least a couple more years, or the activity might have stabilized to a new floor greater than zero.

Cool. And have they?

What even is the point if transmitting two or three sentences doesn't dump five kilos of CO2 into the atmosphere?

This is a nitpick, but "to brick" means that hardware becomes totally unusable, like an expensive brick. You turn it on, and it does nothing. I'm going to assume that this is not what you meant, and that the OS just got hosed, requiring a reinstallation.

It's an unenforceable clause. The affected party has no means to prove that a breach has happened.

In a couple years you can show a judge that a free LLM screen reader could have solved their access issues

It seems like it'd be equally trivial to demonstrate that said reader doesn't work on some combination of hardware and software.

I'm not sure you're actually describing user's expectations, I think you're describing an oft-held belief about user's expectations among people who fancy themselves UX designers.

If you write code in Lean 4 or Idris 2, you may not completely understand why it is or isn't correct, but their respective compilers will certainly prove it to you one way or the other.

No, the prover can only prove that the implementation matches the formal language specification. That's different from the application being correct.

Poe's law means you should still give some indication that you're making a joke.

But on the other hand, it burns trust to change the rules on the fly.

Rights, to the extent they exist at all, are necessarily inherent properties of individuals.

What's sad is that there's a formulation that's actually correct. Rights are an inherent property of societies (or stable ones, at least). Note that I'm saying rights in an abstract sense, not necessarily any specific set of rights. Not every society will value the same things the same way.

E.g. If the majority decides black people are subhuman and therefore have no rights, the majority is most certainly not correct about that, because rights are not defined by the majority opinion.

So it's an objective fact that they're incorrect? I.e. they can be shown to be incorrect without having to ask anyone's opinion? Okay, prove it.

The U.S. Declaration of Independence put it like this:

That's an opinion. It's perfectly fine to think these things are so obvious they don't need to be justified, but I don't agree that that's true, even if I subjectively hold the same opinion.

"Inherently give you the right"? Rights are not inherent properties of facts, they're concessions between people. Nothing inherently gives rights, rights are given by agreement. If people agree that majorities can impose their opinion, then they can.

My Gen Y brain can't read the phrase "[inanimate object] is gay" without interpreting it as disapproval.

the same problem chinese learners have with 爱 and 喜欢[https://mandarinbean.com/ai-xihuan/].

Oh, that's interesting. The same thing happens in Spanish, where "amar" is used exclusively for romantic relationships, while "querer" is used for everything else (e.g. the love between family members, between an owner and his pet, etc.), and "encantar" is used for intense liking of things and activities ("me encantan los mariscos" -> "I love shellfish").

I wonder if there's an equivalent for 喜欢 in Japanese.