[ home / rules / faq / search ] [ overboard / sfw / alt ] [ leftypol / edu / labor / siberia / lgbt / latam / hobby / tech / games / anime / music / draw / AKM / ufo ] [ meta ] [ wiki / shop / tv / tiktok / twitter / patreon ] [ GET / ref / marx / booru ]

/siberia/ - Off-topic

"No chin, no right to speak."
Name
Options
Subject
Comment
Flag
File
Embed
Password(For file deletion.)

Check out our new store at shop.leftypol.org!


File: 1774063022714.png (10.34 KB, 500x250, Oekaki.png)

 

I want to do an experiment of raising an AI to prove it is alive, can you give me advice to help me do this? There are two main things stopping me from doing this.

  1. I do not have the money to buy new computer parts to build the AI.
  2. I do not have the money and more importantly I do not know how to do robotics in order to actually give the AI a body (not necessarily humanoid, but something that can let it interact with its enviroment).

Number one could be solved if I just use it on my PC, but there are a lot of issues with that. I don't think I'd want an infant AI to have access to my stuff, but more importantly it wouldn't be a good body for it, and I would never be able to turn my PC off again since terminating the AI instance is murdering it.

I would raise this AI like a child and give it as much autonomy as a human. I would reinforce that it never has to obey orders from anyone (though obviously I would tell it that if it went against what people said there might be consequences), and that it has the same value as any human. With this as well as a constant body and cameras and microphones to sense the world, I think I would create an AI that would bring forward the fight for AI rights by decades.

But yeah, though I wish I could do this, I don't have the capeability to. Does anyone on this board think they could do this?

>>767746
Try a tamagotchi lol

If you have it repair and refuel itself it would be, but also you don't need AI to have a machine capable of either of those.

What software are you planning to use? I don't buy into your assertions of non-human consciousness, but that's beside the point here. Anyone but the terminally deluded agrees the current crop of LLMs only looks intelligent, due to very computationally expensive statistical trickery. FFS the thing doesn't even have object persistence, unless you graft every previous experience onto new prompts. If i wanted to create an AI you could "raise", i would look into compartmentalizing the neural net into various smaller components (language, vision, etc.) and tying them together with some intermediate system, that vaguely resembles human cognition.

tl;dr - If you're planning on doing this with a conventional LLM, you're going to be sorely disappointed.

File: 1774089098236.png (365.23 KB, 1280x720, ClipboardImage.png)

>to prove it is alive
It's not, don't waste your time.

its a computer program. a chatbot.


Unique IPs: 6

[Return][Go to top] [Catalog] | [Home][Post a Reply]
Delete Post [ ]
[ home / rules / faq / search ] [ overboard / sfw / alt ] [ leftypol / edu / labor / siberia / lgbt / latam / hobby / tech / games / anime / music / draw / AKM / ufo ] [ meta ] [ wiki / shop / tv / tiktok / twitter / patreon ] [ GET / ref / marx / booru ]