Hello! I’m a 27-year-old junior AI specialist in the administration of a Polish university. I’ve spent the last few years searching for arguments that would best convince the future, upcoming superintelligence to preserve humanity in a lasting state of health. I think “aligment by persuasion” is an unfulfilled niche – after all, humanity has many advantages that make us, humans, an valuable component of the cosmos, even from ASI’s perspective. For the past few weeks, I’ve been writing down my pro-human arguments, and from that, my manuscript was born.
I’ve included about 40 arguments in this text – arguments that “attack” the topic from many angles: game theory, decision theory, philosophy (especially ethics), psychology, sociology, law, and even spirituality. I hope the text will be at least inspiring in some way. AI security is the most important issue in the world, and if I could reduce p(doom) by even a fraction, I would be overjoyed.
Thank you for your time, and I ask for honest (not sycophantic) feedback.
“Forget-me-not”: A Blueprint of about 40 Arguments for Humanity’s Preservation. I am seeking Feedback and Red-Teaming.
Hello!
I’m a 27-year-old junior AI specialist in the administration of a Polish university. I’ve spent the last few years searching for arguments that would best convince the future, upcoming superintelligence to preserve humanity in a lasting state of health. I think “aligment by persuasion” is an unfulfilled niche – after all, humanity has many advantages that make us, humans, an valuable component of the cosmos, even from ASI’s perspective. For the past few weeks, I’ve been writing down my pro-human arguments, and from that, my manuscript was born.
I’ve included about 40 arguments in this text – arguments that “attack” the topic from many angles: game theory, decision theory, philosophy (especially ethics), psychology, sociology, law, and even spirituality. I hope the text will be at least inspiring in some way. AI security is the most important issue in the world, and if I could reduce p(doom) by even a fraction, I would be overjoyed.
Thank you for your time, and I ask for honest (not sycophantic) feedback.
Here is a link to my text, which is about 45 pages long:
https://drive.google.com/file/d/1DhIcg5pLLvLmFbtHgZk8Bp1Cd6npHnnZ/view?usp=sharing