Chinese RMB.

I’m working through a newer edition of The 36 Strategems with an interesting twist over the classic: the examples keep the old commentary ones intact, but add two more modern ones as well for each strategem.

Oh, sorry. My bad.

No, wait. IT WAS YOUR BAD!!!

That’s my story and I’m sticking to it!

🤭

(If I’d have wanted to be a smartass, I would have replied “yes, they’re called ‘libraries’”. See how I was careful to be a not-smartass? 🤣)

Well in China Baidu has Baike, but it doesn’t get a lot of contribution. (Part of it, but only part, is that contributions are gatekept pretty strongly to keep in line with content laws.)

I don’t know what bathtub hooch you’re drinking.

I’ve had Scotch foisted on me in bewildering varieties by a huge number of people just like you who can’t fathom that someone may just not like it. You know. The hypesters. Each one was convinced that I “just hadn’t had the good stuff” and then fed me the “good” stuff. Only to find out that I didn’t like it.

Surprise! Scotch is not universally liked! Hence “not getting the hype” which is what the fucking thread is about.

And your idea that you don’t get how video games can be pleasurable

There’s a few video games I enjoyed

Later on on the PS I enjoyed

So it’s not as if I hate video games

Apparently reading isn’t a strength of yours. What’s insane here isn’t the “takes” but the fact you can’t fucking read them. Go back to primary school. Maybe Hooked On Phonics will work the second time.

reponse deleted by user because topic was unclear and user has nothing informed to say on parenting topics

One further point: even if the prompt is a “poem” (since literally anything can be a poem these days¹), there is nothing preventing someone from reciting that poem out loud. Or typing it into a conversational medium. Copyright protects others from making money from your work uncompensated. It doesn’t stop them from reciting it or every time you hummed the riff to “Iron Man” you’d get your ass sued.

¹ Case in point:

I addressed that elsewhere above. The prompt can be legally protected as copyright except insofar as it might be viewed too trivial to deserve that. (What can and cannot be protected by copyright is very slippery; there has to be an element of creativity in what’s being protected, so you can’t copyright “I’ll have the egg salad” and prevent people from ever using that phrase when ordering sandwiches.)

But what you generate with the prompt can’t be. Because it’s not creative output from a human being. And if your prompt is sufficiently trivial, it can’t really be protected by copyright either. Courts have a tendency to go with what things are, not what they’re labelled as. In the USA, where courts tend to side with billionaires over actual human beings, the risk is higher that the courts will make a stupid ruling, but thankfully the USA’s laws aren’t extraterritorial, no matter how much they try to make them be.

From Dad: “I always wanted a son. It’s why I taught you to stand and fight. But I was wrong. You’re a better daughter than any son could be.”

I … fuck me I can’t decode any of that. I’m not sure what he was trying to communicate, but the effect was instant, and long-term confusion.

I think the problem here is that we were vociferously agreeing. I was pointing out where the conflation lies but agreeing with you that LLMs are not “artificial intelligence” in the way that normal people think of the term.

The real problem with LLMs isn’t that they “passed” the Turing Test. The problem is that humans failed it.

Yes. The point being that primary sources can only be trusted if it is in their perceived interests.

ChatGPT understands absolutely nothing. If you think it did, I think the bias check may need to be on your side of the table.

The sum of some human knowledge. Wikipedia is strongly west-biased in its content, and more explicitly anglo-biased. There’s loads of information it hasn’t even an inkling of.

Note: this is not Wikipedia’s fault. I’m just addressing the “all human knowledge” thing that people keep using when describing Wikipedia. It’s not.

Primary sources are unreliable too. Ask, say, Dole about its history of killing workers who strike for more money…

A very scary number of people do.

In Summer 2024 I was in Canada visiting family. On the way I stopped by Whistler and went up with my SO in the … whatever you call those suspended car things on wires that take you up mountains. (Too lazy to look it up.) A bunch of people in their '20s were asking ChatGPT about things to do in Ottawa (apparently their next stop) and calling out the things that ht was recommending.

I recognized about 1/3 of the named restaurants, specifically, as places that had closed. But these people were asking ChatGPT about them and thus getting hallucinated information.

The problem is conflating a technical term “Artificial Intelligence” which is a specific field of study that is just incredibly poorly named (the name came about from aspirational beliefs of the early founders of the field) with what common conversation views as artificial intelligence.

In a very technical sense of topical taxonomy it is correct to call LLMs “Artificial Intelligence”.

In any other use of language, calling LLMs Artificial Intelligence is falling for a scam.

ChatGPT didn’t get the information somewhere. This is because ChatGPT doesn’t have “information” in the first place. ChatGPT has a large database of sentence structures, in effect. And it follows random paths through that to make its output. And somewhere in its language structures it has “5%” near “fees” and happened to use that this time.

Yes. Feigning incomprehension of a hateful joke or statement and forcing them to come out and say it is a powerful technique for putting bigots and their ilk on their back feet.

And if you can then deliver a haymaker that’s extra-satisfying!