• Log in with Facebook Log in with Twitter Log In with Google      Sign In    
  • Create Account
  LongeCity
              Advocacy & Research for Unlimited Lifespans


Adverts help to support the work of this non-profit organisation. To go ad-free join as a Member.


Photo

My project (following OpenAI.com's footsteps and encode.su's footsteps)

agi immortality

  • Please log in to reply
No replies to this topic

#1 Dream Big

  • Guest
  • 70 posts
  • 90
  • Location:Canada

Posted 16 June 2021 - 09:55 PM


You can see my progress towards AGI below:

https://encode.su/th...data-compressor

 

I really like openAI.com's progress. I've tried all their AIs. There's haters that hate them but these are the outcast that don't know brains need big data or simple things like how it answers many diverse prompts using basic patterns found in data. It isn't just the big data that makes DALL-E amazing though, or else I could get such results using 1GB of data, which doesn't take too long to train on. I also like the programs on The Hutter Prize and the Large Text Compression Benchmark pages, and encode.su where they discuss it all. I like how they explain more what's really going on in AI. GPT/ anything that uses Backpropagation doesn't seem so intuitive, (I still will use some things from GPT as they are good ideas), or maybe this is because the massive lack of anyone explaining how it works and only a few know how to code one up from scratch. Nonetheless there's a lot of good stuff on both sides here.

 

https://openai.com/projects/

http://mattmahoney.net/dc/text.html

http://prize.hutter1.net/

 

I know how to make DALL-E / GPT-3 more AGI/ Agent-like, but it would be overboard to say it right in public, as we are really close to AGI now and the singularity is right around the corner, for when the AGIs clone themselves to make large me_teams, think and move at fast speeds, iteratively upgrade their intelligences, and a lot more.

 

It would be great if someone else here could help me clarify some of the methods being used, as I don't see anywhere it is all laid out in easy to read English, for how GPT or other's AIs work. This would allow much faster development.

 

It's clearly visible we are just machines once you get into this stuff. All of physics/ life/ AI is just patterns as I said in my last post, there is no how death occurs, there is no difference between a man and a rock, all that immortality and evolution is/ about is resisting change, humans do it better than rocks. We die even when we bend or move but we only fear a certain type of "change" to the body, and we use patterns/ memories as experience to solve similar future issues, we use patterns to be patterns, we clone cells and bodies and minds (either by teaching or cloning hard drives). In Cryonics you don't move, and are like a statue, but upon the "thawing" you should be able to say you are still you. I'm really good at the understanding of these things after lots of research into this stuff. Some my ideas I started with remained and were good, many failed.


Edited by Dream Big, 16 June 2021 - 10:22 PM.






Also tagged with one or more of these keywords: agi, immortality

1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users