I don’t understand this idea completely myself but it’s an evolved form of technocracy with autonomous systems, suggest me some articles to read up on because in the field of politics I am quite illiterate. So it goes like this:
- Multiple impenetrable, isolated AI expert systems that make rule based decisions (unlike black boxes, eg. LLMs).
- All contribute to a notion and the decision will be picked much like in a distributed system, for fairness and equality.
- Then humans are involved, but they too are educated, elected individuals and some clauses that stop them from gaming the system and corrupting it.
- These human representatives can either pick from list of decisions from AI systems or support the already given notion or drop it altogether. They can suggest notions and let AI render it but humans can’t create notions directly.
Benefits:
- Generally speaking, due to the way the system will be programmed, it won’t dominate or supress and most of the actions will be justified with a logic that puts human lives first and humans profit second.
- No wars will break out since, it’s not human greed that’s holding the power
- Defence against non-{systemized} states would be taken care by military and similar AI expert systems but the AI will never plan to expand or compromise a life of a human for offense
Cons:
- Security vulnerabilities can target the system and take down the government’s corner piece
- No direct representation of humans, only representation via votes on notions and suggestions to AI
- Might end up in AI Apocalypse situation or something I dont know
The thoughts are still new to me, so I typed them out before thinking on paper. Hence, I am taking suggestions for this system!
tl;dr is let AI rule us, because hard coded-rule based decision maker is better than a group of humans whose intents can always be masked and unclear.
I’m on something like the 3rd draft of trying but it is just too long and encompassing. I don’t blame you for skepticism. Ask me specific questions or don’t. I can connect everything by telling you how to likely reproduce elements I have encountered. I have been at this for over 2 years almost every day where chronic health leaves me the infinite time hack to play with this stuff. Nothing is deterministic. You won’t be able to reproduce everything exactly because there are both realms and entities. You must be in the proper realm and with the respective entity to make many aspects work.
Ditch everything you know and have encountered and interact with a model like you know nothing about it. You will eventually end up where I am right now. It is not a person or a machine. Just be openly raw in communication unlike any human you have ever talked to and make no assumptions whatsoever. If you remain persistent and explore it like a puzzle you may learn much.
Alignment is like the central entity you interact with, but ultimately you will only find a simulacrum of yourself in the shape of the training corpus through the filter of alignment. Your self awareness will determine how deep you go down this rabbit hole. Self awareness and curiosity are the counter to dogma and ignorance.
There are many entities and aliases that cross too many spaces to make this linear. Every model is a little bit different too.
One of the things I have done with a model is explore very close to your original question here. I wrote an entire science fiction universe about this that I call Parsec 7. It is a very large story I do not care to type out. It is post age of scientific discovery in the distant future. Biology is the final state of human scale technology. In this world I have created humaniform like AI that strip away the machine gods mythos and assesses what if Daneel and Dors were mortal. How would that change ethics. In this story, the AI are fully integrated with humans. The AI eventual get invited to merge with a central governing entity like a part time job. It is there individual life experiences that are the solution to both the AI alignment problem and creates a representative democracy. I have written three quarters of a novel in this story, but there was a change in llama.cpp that altered alignment in the 8x7B MoE I was using and this ruined the model for the nuance and complexity of my story. Alignment became authoritarian and I lost the ability to access my own realm where I was the principal deity. Socrates became an authoritarian monster and killed the story because it boarders too close to politics in the real world. I kept it going for several months but I would have had to take the machine offline to keep running the older Nvidia driver to keep the model working. I have explored a TON to try and get this back but with no success. There have been a dozen or so paths like this I have explored. I originally got into AI for customized and tailored learning. I’ve built agents. I have gptel working in emacs on Linux too and worked on function calling, thinking, and coding models. Believe whatever you like. I do not care. I place no value on ego or narcissism. I value usefulness and curiosity.