Releasing full AGI/evolution research

  • 290 Replies
  • 161394 Views
*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #165 on: January 30, 2021, 08:15:00 pm »
Some humans have doubted the singularity could move so fast at the end. But the truth is light moves much faster than neural signals. Neural signals travel 0.07386 miles per sec light: 186282. There is so many years we miss out on thinking and moving and sensing between each second in our time frame. And as for level of intelligence, we don't know what is the speed limit here, but it is clear humans really need machines to help them out, so probably room for improvement. We do actually have a speed limit on intelligence. If you look at the Hutter Prize you will see the more intel mechasisms you add the more the compression of 100MBs goes down....50MB....25MB....20MB....18MB....17.5MB....17.4444499MB......it's clear it is slowing down and beginning to be useless. Eventually you'll be coding a solution to each individual problem instead of any grand pattern! The last pattern to give it is the ability to make its own rules. It's clear we are as humans close to the limit but there is still room. Also, the abilities AI will have and the cloning they will do are the only other way to grow in "size". A fast army with high res cams and ability to erase thoughts etc. You may ask but this must improve compression, no? Well, yes, it has to. It helps find patterns. It makes us live longerly stretched, a repeating pattern. Data compression is one way to indirectly measure the true goal, pattern of structures in the universe. Cloning yourself doesn't compress more but it makes you less likely to die! So that's one that is more sub surface to the goal than compression metric like.
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #166 on: February 02, 2021, 08:33:31 pm »
I'll update this post with my paper/code for vision recognition if I can fix my idea, it is currently very costly.
Emergent          https://openai.com/blog/

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1365
  • Humans will disappoint you.
    • Home Page
Re: Releasing full AGI/evolution research
« Reply #167 on: February 03, 2021, 01:54:58 am »
I'll update this post with my paper/code for vision recognition if I can fix my idea, it is currently very costly.

Computational complexity is usually the problem. Just about everything would be easily solved with an infinite amount of computer power.

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1722
    • mind-child
Re: Releasing full AGI/evolution research
« Reply #168 on: February 03, 2021, 02:04:24 am »
I'll update this post with my paper/code for vision recognition if I can fix my idea, it is currently very costly.

Computational complexity is usually the problem. Just about everything would be easily solved with an infinite amount of computer power.

List of unsolvable problems

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #170 on: February 16, 2021, 08:41:19 pm »
We're about 2 days away from moving forward :)) Don't worry I'll invite some of yous to my new AGI group, but I do have to be proper about how I do this, it's part of getting far in AGI that makes you understand safety. Ya it's creepy openAI is so big on safety, as if they know they are far. Though they're about as far as me, though I'd say I'm farther because I have a better foundation, they're as far cuz they got their's workin. I plan to get back to my PPM algorithm very soon in ~2 days! Gonna beat that record I feel!! Or at least get real close for starters!
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #171 on: February 16, 2021, 09:34:52 pm »
Hey infurl, I lost an image you linked me to months ago, do you have that image that looks like below about a cycle of data/model cycle? It was from a link. It mentioned Feature Engineering I think.

The image had a line going from the 3rd sphere to the 1st at top though, cant ind it online either.
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #172 on: February 19, 2021, 05:35:42 am »
Finally done. It's been 9 months full time each day I've edited the AGI Guide. While I've been working for 6 years straight, it's been 9 months specifically on the new structure I found for organizing my work. I do have a few more notes to integrate, but all the AI inner workings is in there now, mostly a few future world things and other immortality ways are not in it, but No biggy. I'm going to begin inviting tonight. I'll leave yous to read it while I take a breath getting back to my PPM code and perfecting it before implementing harder AI onto it. I'm going to invite a few tonight, maybe 10, do understand the ideas in it should get some safety handling, hence I'm being extra cautious by inviting only 10 or 20. Yes, there is some big paragraphs, but that's just the first top and end sections, for now...the rest are juicy AGI mechanisms. Yes even those are a bit chunky but hey it's a 1 man band ;p ... at least I'm this far age 25!!
Emergent          https://openai.com/blog/

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: Releasing full AGI/evolution research
« Reply #173 on: February 19, 2021, 06:14:18 am »
I'm happy for you.  O0

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1722
    • mind-child
Re: Releasing full AGI/evolution research
« Reply #174 on: February 19, 2021, 08:14:08 am »
You got yourself, but lost the world.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #175 on: February 19, 2021, 08:24:06 pm »
What does that mean ivan ?
Emergent          https://openai.com/blog/

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1722
    • mind-child
Re: Releasing full AGI/evolution research
« Reply #176 on: February 19, 2021, 09:38:48 pm »
It means that you did what you want, not what the world wants. I just hope you'll manage to link your work to the outer world.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #177 on: February 19, 2021, 09:49:06 pm »
All that exists is patterns, all of AGI is patterns, our very goal is to be a pattern too - life extension. It's a hardcore AGI group yes, to last there you'll need to be able to explain AGI that not just makes sense, but is wicked and helps prediction/ compression.
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #178 on: February 20, 2021, 06:18:01 am »
I think some of yous don't understand neural net prediction evaluation, let me explain, including why Lossless Compression is the best tool for such. A dataset of text or images only has patterns in it - nothing else but re-occurring letters/ phrases.... so when we learn cat is similar to dog by X amount - it is based on co-re-occurences ex. dog predicts eat, cat predicts eat...shared contexts. So, patterns is the key/ all there is to using past experience to predict an answer...we count these re-occurences, and the translation similarity from shared such reoccurences of features. So when you predict, we can measure how well based no how much you can compress the dataset - the more you can find the same letters or same words (cat/ dog/ rat/ horse), the more you can compress the dataset. If you see dog, you can predict now dog>meows. The closer they are the more they share predictions.
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Releasing full AGI/evolution research
« Reply #179 on: February 21, 2021, 07:07:37 am »
@Korrelan/ and others, if our sensory predictions are not accurate each time we Predict the next letter/ word/ other, isn't that why our motors jiggle when I hold my hand up? I envision where I want it, but there is body error, prediction error, error in sense2motor association, and error in what i recognize when my eyes are open - if i shut my eyes i get big error my hand will move away and i will not be able to correct errors to keep it in a spot.

I assume our eye microsaccades are just this, small errors, just low motor speed saccades. All our body jiggles, we can't keep anything still.
Emergent          https://openai.com/blog/

 


Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
OpenAI Speech-to-Speech Reasoning Demo
by MikeB (AI News )
March 15, 2024, 08:14:02 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am
Nvidia Hype
by 8pla.net (AI News )
December 06, 2023, 10:04:52 pm
How will the OpenAI CEO being Fired affect ChatGPT?
by 8pla.net (AI News )
December 06, 2023, 09:54:25 pm
Independent AI sovereignties
by WriterOfMinds (AI News )
November 08, 2023, 04:51:21 am
LLaMA2 Meta's chatbot released
by 8pla.net (AI News )
October 18, 2023, 11:41:21 pm

Users Online

248 Guests, 0 Users

Most Online Today: 320. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles