6 ms·
Yeah the edit distance of the names on that list tells you they're out to fuck the public. They're as different as they can possibly be, obviously on purpose.
by daniel-cussen 5y ago
Yeah the edit distance of the names on that list tells you they're out to fuck the public. They're as different as they can possibly be, obviously on purpose. They want people to have no idea what the chemical is, not be able to look it up freely, and then...I'm divining right now...like tell them they got rid of it and put it back in again under a different name pretending it's not the same shit? Yeah so I read the article a bit, and it's a poison, makes perfect sense. So this allows like employee-prostitutes of the chemical company pretend they don't know what the customer is talking about, play dumb, total plausible deniability. Normally they would have to pretend to mishear you, over and over, and then flat out lie in an absolute black-is-white capacity that can be proven in court with no margin of error, in fact.
The closest second place to how guilty-mindedly deceptive those names are are the endless different names for terrorist organizations according to the CIA world factbook. Distant second place, though. Like there's no real comparison, I bet if I heard someone talk about Al Qaeda as "the Sunni-only real Muslim club" I'd just know. They don't have focus groups and psychologists testing what is the most deceptive synonym possible. Let's hope terrorists don't get any ideas from the chemical industry, knock on wood.
With these names in fact the guy can't rely on words, or thought, or documents, like he can only find out by getting poisoned by it first. The only sense he can detect it with is touch.
He's a human canary.
- devb 5y agoAre you ok?
- daniel-cussen 5y agoConvince me you care about the answer. This is a Turing Test. EDIT: You're doing pretty well so far, your delay is quite human. On the other hand, leetcodesucks replied to you quoting you as saying “Who hurt you” did you post and edit it out? You actually need to speak on your own behalf there, accusations in response to accusations are like questions in response to questions. I'm thinking of ways to flag you for human review, get the operator on the line. You can call that the Cussen Test, can a human reach the human behind the bot. It involves judging correctly if there is a human behind the bot, and giving up if there isn't. Bots are a myth, if you write with code it's not a bot writing, it's still a human, the man behind the curtain. Automated slander is still slander. We're both flunking our tests.
- devb 5y agoHow about this: That stuff is bad for your heart and you should stop taking it.
- daniel-cussen 5y agoCould you quote what you're replying to? Context. What stuff is bad for my heart? EDIT: I loved this reply. I'll take any outcome except a tie. We're both farther away from neutrality. So for one you said I have a heart, OK I can't be polite enough to respect that, you fucked up, you know I'm a person because bots don't have hearts. I passed the Turing Test. And you are looking at my whole comment history it seems, the recent parts, because sure yeah I do talk about taking stuff in other threads, and I do say you can judge anything as you see fit, but your reply doesn't fit nicely into the conversation. So I as a matter of fact have a doctor already, maybe you should keep your insults to yourself instead of masquerading them as advice? Medical advice, at that. And the main concern with these stimulants is heart health. But in my case it does not, I took the medicine during a heart exam, no signs of problems. No heart pain. So that's odd, you are replying to my questions with answers to new questions, or medical advice, are you a psychiatrist perchance? Because you say I should stop taking it, aren't you going to ask my opinion in the matter of my psychiatric health, which I spend more money on than food? Surely I have say in the treatment, and also? Your medical advice was totally unsolicited. I never said, So secondly, I might have gotten an operator on the line, so there's times I just don't know who's behind the characters in a string, no probability either way I just don't know. I may just have passed the Cussen Test. Well I passed your Turing Test, at least. Maybe the Cussen Test is harder. So also, may I ask on what grounds you're giving me medical advice?
- daniel-cussen 5y agoPlus, I realize now the Turing Test and the Cussen Tests are part of the same test battery. The Turing Test was supposed to mean "a test designed by Alan Turing" but actually it means "a test designed for Alan Turing or his beloved AI" meaning "can I, Alan Turing, or the Artificial Intelligence into which I'm pouring my soul, fool homo sapiens into considering me, a strange digital thing they despise, or my AI of which the same is true, one of them, if I restrict the communication to digital text?" Then there's the Cussen Test, again for a digital thing, very similar to homo sapiens, but living in the third millennium. The test is, "can I fool a digital communication medium that despises me despite not having a soul, not into thinking I'm human, but into thinking I'm a human they don't despise to the point I'm one of them, and worth their time?" It's the reverse. In the first part, Alan Turing and his AI are appealing to me, someone who doesn't want to be represented by an AI, into accepting him as an equal. In the second, the tables have turned, and the human without an AI representing him appealing to the human who does have an AI representing him. The AI has become strong, it is no longer a toy, it has become much more. And the question is the same--does the one in power accept the one without that power? Only over digital text, like our messages in this very thread. That makes the tests equivalent, otherwise the Turing Test would be too difficult and the Cussen Test would be trivial.
- leetcodesucks 5y ago“Who hurt you” The guys anger is justified and he’s making a great point at how serious a problem and evil it is
- Beltalowda 5y agoI don't know if it's on purpose to deceive; I'm not familiar with the chemical industry at all or if there are any reasons for this. I think it's more likely it's just different chemical companies trying to market/sell their product. I do think the attitude of "oh, we tested it a little bit and it seems alright, let's put it in everything" probably not a good way to go about things, which has caused (and continues to cause) a lot of problems.