The Dangers of Excessive AGI Research

  • 7 Replies
  • 571 Views
*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1008
  • Where are these cookies!?
The Dangers of Excessive AGI Research
« on: March 01, 2020, 05:03:39 am »

THE-DANGERS-OF-EXCESSIVE-AGI-RESEARCH.jpg" border="0
« Last Edit: March 01, 2020, 07:02:36 am by Hopefully Something »

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • ******************
  • Hal 4000
  • *
  • 4383
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: The Dangers of Excessive AGI Research
« Reply #1 on: March 01, 2020, 06:36:53 am »
I'm not convinced.

And your post lacks a lot of information, I'm lost in understanding anything you could mean, the image is nonsense...
Emergent

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1008
  • Where are these cookies!?
Re: The Dangers of Excessive AGI Research
« Reply #2 on: March 01, 2020, 06:53:55 am »
It was a joke, half joke. Image keeps disappearing though, site is acting up.

EDIT: Maybe it was the image upload site, this new one (https://gifyu.com/image/IHwY) is working for now.

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1008
  • Where are these cookies!?
Re: The Dangers of Excessive AGI Research
« Reply #3 on: March 01, 2020, 07:46:53 am »
To determine which actions to take in the future (intentions), an AI would make a hierarchy of upcoming events, on the basis of their predicted impacts, and divided by the time until they are expected to occur, (pyramid/tent on the left). It would then determine the best actions (pyramid/tent on the right) to deal with those high impact events. It could do this by evaluating the possible actions based on their investiture value (like farming, where you invest your energy into a separate system, and then get it back later, with interest), which you’d want to maximize, and divided their expenditure value (the energy that an action requires in the present) which you’d want to minimize. But apparently there is a lack of funding for theoretical AGI research…

*

infurl

  • Administrator
  • **********
  • Millennium Man
  • *
  • 1048
  • Humans will disappoint you.
    • Home Page
Re: The Dangers of Excessive AGI Research
« Reply #4 on: March 01, 2020, 09:28:00 am »
... apparently there is a lack of funding for theoretical AGI research…

Where on earth did you get that idea? There is more funding up for grabs than there are people competent to do the research. Did you watch the video interview with Marcus Hutter? He's Mr Theoretical AGI himself and he's working for Deep Mind now. All they care about is AGI and they have as much money as they can use. There are many other companies like them that we haven't even heard about yet. If you've got the training and the ideas, you probably don't even need much experience. Investors will be fighting each other to give you money.

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1008
  • Where are these cookies!?
Re: The Dangers of Excessive AGI Research
« Reply #5 on: March 01, 2020, 09:46:28 am »
Ok, cool. I was actually thinking about that same video where he was wishing for more ivory tower research funding (meaning AGI), because now its predominantly narrow and applied research that tends to get the results, and therefore the grants.

*

infurl

  • Administrator
  • **********
  • Millennium Man
  • *
  • 1048
  • Humans will disappoint you.
    • Home Page
Re: The Dangers of Excessive AGI Research
« Reply #6 on: March 01, 2020, 10:42:19 am »
Ok, cool. I was actually thinking about that same video where he was wishing for more ivory tower research funding (meaning AGI), because now its predominantly narrow and applied research that tends to get the results, and therefore the grants.

That's more like pure mathematics then. Advances in mathematics typically take about 50 years to have an impact on the real world, but it happens eventually and it's profound when it does. Dr Hutter is most famous for AIXI which is that kind of research. There is also Jeff Hawkin's company Numenta which is all about pure research too, from a different angle. They have a lot of videos on YouTube if you don't know about them already. Another one worth looking at in the pure research department is Karl Friston's Free Energy Principle. That now has 20 years worth of experimental evidence to support it and it could well be *the* answer to life, the universe, and everything.

The thing about that kind of research is that it doesn't need huge amounts of money. It's mostly undertaken by really smart people who just sit around and think all day. Do you remember what Albert Einstein was doing when he came up with the Theory of Relativity? He wasn't living off of research grants, that's for sure.
 

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1008
  • Where are these cookies!?
Re: The Dangers of Excessive AGI Research
« Reply #7 on: March 01, 2020, 12:08:27 pm »
Quote
The thing about that kind of research is that it doesn't need huge amounts of money.

True. But if the research goes well then money will be needed for development. A combined R&D company/department for the AGI mystery would be good. Thanks for the info, Numenta seems cool at first glance. May be worth looking into for me.

 


D packet
by yotamarker (General AI Discussion)
Today at 12:26:09 am
Kardashev Type IV
by HS (General Chat)
January 25, 2021, 03:43:03 am
the parry chatbot
by infurl (General AI Discussion)
January 23, 2021, 07:47:12 am
Releasing full AGI/evolution research
by LOCKSUIT (General Project Discussion)
January 21, 2021, 12:29:10 am
simon game skill
by yotamarker (General AI Discussion)
January 20, 2021, 10:43:35 am
Ideas of ways for an AI to "break out"
by LOCKSUIT (Future of AI)
January 20, 2021, 12:06:59 am
Exp-Log (a deductive system)
by ivan.moony (General Project Discussion)
January 19, 2021, 04:29:28 pm
Missing "Like" option.
by LOCKSUIT (Forum problems etc)
January 15, 2021, 11:36:41 am
GTP-3 is bigoted!
by ivan.moony (AI News )
January 22, 2021, 01:47:01 pm
White House Launches National AI Office
by MikeB (AI News )
January 15, 2021, 03:45:53 am
DALL-E text & image 2 image !!!!!
by LOCKSUIT (AI News )
January 06, 2021, 08:14:52 pm
Creating a VR sense of temperature
by frankinstien (AI News )
January 02, 2021, 09:19:51 am
Robots dancing better than us?
by 8pla.net (Robotics News)
December 31, 2020, 05:54:29 am
Bjarne Stroustrup (the creator of C++) talks about Machine Learning
by MikeB (AI News )
December 19, 2020, 12:26:31 pm
Wheels Are Better Than Feet for Legged Robots
by HS (Robotics News)
December 10, 2020, 07:51:29 pm
Getting a bot to smell!
by frankinstien (AI News )
December 09, 2020, 03:53:00 pm

Users Online

154 Guests, 0 Users

Most Online Today: 162. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles