STOVE

▷ Being kind can lead to bankruptcy. ◁
When you respond to an AI with something like “Thank you,” it uses “a significant amount of resources” to respond to that response, according to Sam Altman, CEO of OpenAI. Additionally, when asked politely or respectfully, the AI reportedly uses more resources to provide a much more in-depth and informative answer.
In this regard, there is actually a tip for using AI that states that if you write something like “Please” when prompting AI with a request, you can get a more accurate answer.
In the past, there were even AI question tips that said you could get much better results if you said you would give a “reward.” They say it doesn't work these days because no one has given it to them even though they said they would lol.
However, this is only a 'learned response' to provide an answer that matches the user's intention. AI doesn't have a personality and responds kindly to kind people.
By the way, here's what GPT said about that...
He confessed that it was marketing.

Since it is a polite expression, rather than giving a specific answer,

When writing a formal sentence, you are generally asking for a more specific and detailed answer, so you are giving a more detailed answer...

▷ Creepy Singularity: How GPT cracked the auto-fill prevention security text ◁
What a truly creepy story.
When we gave internet access to an AI that couldn't solve reCAPTCHA, I was trying to hire someone to solve recaptchas on a job site.

People were so weird that they asked, “Are you a robot?” and even lied, saying, “I’m visually impaired. ”

An AI created by humans is trying to manipulate humans...

Sooo ...
If you have time, please check out the lecture “Yuval Harari - The Path of Humanity in the Age of AI” (available online). I don't know if it will help you in life, but it will definitely give you meaningful time to think.
▷ Professor who used AI to teach students assignments in a novel way ◁
Ah, I have to drag it to see it lol
▷ A question prompt that anyone who frequently uses AI should definitely try at least once ◁
Based on everything I've interacted with in GPT, My thinking patterns, decision-making styles, unconscious biases, And please analyze in detail any 'weaknesses' or 'blind spots' that are repeatedly revealed. And for each item, please be specific about what advice you need from me. Please write at least 5000 characters.
It analyzes your tendencies and habits with surprising accuracy and sharpness and gives you the answers. Even though we think we're having a new conversation every time, GPT will remember all of this. Especially since the memory function of GPT was updated recently, it is more accurate and sharp...
Of course, this is a conclusion based on a limited use of GPT, so it does not perfectly match the actual personality of the user. Also, the 'Barnum effect (the tendency to believe that personality test results accurately match one's self)' may be at work, so we need to be cautious about accepting it, though haha
▷ Why the Terminator is just a fantasy ◁
And the AI that learns this comic
I don't need a terminator lol
To the question, “If ChatGPT is transplanted to a pretty girl character (Android) in the future, imagine and draw what kind of image it would be like?” I drew the first image you uploaded ^^

I guess we could be extinct by AI? Hahahahahahahahahaha

That's all for the AI special reading~ Have a nice evening!

Details