🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

The plan? Convince the media that superintelligent AI has the potential to wipe out all life on earth, to attract media attention to our new and improved chatbot.

following this pyramid when modeling other people is a great way to get exploited plus, following incentives is often unacceptable. suppose someone punches you in the face and steals your insulin. well, they were "only" following incentives!

The solution to this problem is to stop judging things by whether they're normal. Plenty of normal things are utterly unacceptable. Whether or not an abusive behavior is normal should be irrelevant.

i fuckin love these guys man

the fact that self-insert fiction is seen as incapable of being high-quality or "respectable" by the vast majority of people is an incredible psyop

myst-ass book

people really think that being the "best parent feasibly possible" is about controlling all of the information that a child has access to

PSA: There are going to be people who literally YELL at ChatGPT-4o as a "prompting technique" to get what they want. These people are evil and should not be trusted. You could already see this with people acting abusively toward previous versions, but it'll be even starker now.

If you see a potentially trapped chest which says "THIS IS A TRAP" on it, it is not valid reasoning to say, "if it is a trap then the trapmaker wouldn't have wanted to reveal that, so this must not be a trap", because if you reason that way, then the trapmaker can exploit that.

this person has never seen a human baby

chickens laying eggs isn't ugly or gross, it's beautiful and eggs are beautiful. and taking those eggs from chickens and eating them is wrong.