this post was submitted on 22 Jul 2023
1571 points (79.2% liked)
Memes
45903 readers
1371 users here now
Rules:
- Be civil and nice.
- Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
How would such an AI be created and how would it work?
I think that such A.I. should be build with the principles of sustainability of resources, the preseevation of human and animal life inluding but not limited to mental health, phisical health, fullfilment, and perception of freedoom. When it comes to penalties while i would like to get rid of any type of penal system i understeand its nececity due to people being people, so i would like for that system to focus on rehabilyitation rather than punishment as long as its possible and if not then just on keeping bad actors from causing any more harm into society. It should distribute resources to the human populance according to their necesities when it comes to basic needs being housing, energy, plumbing, food, medical suplies and services, clothing, transport, entertainment, social fulfilment, etc. when it comes to other things like nicer furniture, software, specialiced tools or anything on the frivolous nature or customisation then it shouldnt be a problem asking the a.i. as long as requests are reasonable since resources should be exploited sustainably enough so that this tyoes of needs could be fullfiled, like what i mean is "i whant a wooden table like this design i found" would be fullfilable request and not " i want a chair that looks like this made out of plutonium" since plutonium is toxic and should be used with care and only for cientific and energy producing porpuses. It should have as a goal to automate as much labor as posible, from agriculture, metallurgy, transport,etc. all the way to accounting, administration, executive decisions, software design, and even science, in the way as to research technologies and other fields for the benefith of humanity be it medical, technological or even space travel, all of this would leave humans to pursue more recreative activities like self fullfilment, socialising, pursuing artistic endeavours like music, painting, literature, cooking, plastical arts, teather (traditional or movies) and cientific research and studying, or just living it with their loved ones, friends or alone in the woods, if anyone whants to labor because they hate themselves (i mean in an ironic way) then they very much can. Also while it would sound bad, it should have control over all weapon and defence systems in the world and have the ability to create more if the need arises and be able to operate them at discretion in case of an external attack on humanity. It should also be able to better itself and self preserve and be abble to feel love for humanity but in a good way so that it doesnt turn against us. Look what im proposing is that we create god pretty much since its the only way we are gonna survive global warming and end war between ourselves for ever. But since your comment is givving me the vibes of being in a condecending manner in the: "AHA GOTH YA DAM LIB" sence, i dont think you are gonna read any of this and if you do you are probably gonna argue that i didnt tell you exactly step by step how to build an A.I. overlod so to that i tell ya of course i wouldnt because i dont know how since im just some dumb dumb on the internet, and while i would like to work building A.I. thats kinda far away from me at this point in my life. And if the down to upvote ratio on my original comment and your response is telling of anything, then its probably gonna validate any point you are gonna trow against mine since im the unpopular one and every one is gonna be able to rule me out as some jerk with a wrong opinion so have fun with that i guess.
Sorry for bad english.
THINK AGAIN YA DAM LIB XD
In all seriousness though, I can understand your desire for utopia and for technology to solve all our problems (I desire those things too!), but I think people generally view those two desires as naive. The world's problems usually don't have simple solutions, AI is still somewhat of a new area of research with its own potential dangers, and putting all your stock into AI solving our problems ignores the very real solutions that could be implemented sooner rather than later.
But don't beat yourself up too much over meaningless internet points, learn about what you're interested in more! If AI is your cup of tea and you want to learn more about the potential issues with creating super powerful AI, I can strongly recommend the channel "Robert Moses AI Safety". Particularly his videos about The Orthogonality Thesis and Mesa-Optimizers.
that's a question for the scientists to answer, we just have to watch their incentive structures and verify that they're doing their peer review