moods and their effect on optimizing algorithms

  • 17 Replies
  • 3421 Views
*

yotamarker

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1003
  • battle programmer
    • battle programming
moods and their effect on optimizing algorithms
« on: May 16, 2021, 10:37:18 pm »
we could say certain triggers or feedbacks can control mood which can change the methods an AI would use
and the AI at the same time can optimize the moods methods

Cheerful
Reflective
Gloomy
Humorous
Melancholy
Idyllic
Whimsical
Romantic
Mysterious
Ominous
Calm
Lighthearted
Hopeful
Angry
Fearful
Tense
Lonely

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1723
    • mind-child
Re: moods and their effect on optimizing algorithms
« Reply #1 on: May 17, 2021, 12:26:36 am »
I believe you can achieve everything with only love and fear.
« Last Edit: May 17, 2021, 01:19:51 am by ivan.moony »

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: moods and their effect on optimizing algorithms
« Reply #2 on: May 17, 2021, 12:45:37 am »
A mood seems to me, like a condensed way of trusting previous experience, over present moment logic - which may be faulty and/or lacking some information.

In humans I find moods interesting because they are both an expression of personal, and genetic memory. Since our genes have been here for billions of years, while most of us have only been here for under a century, I would think that, in some ways, our genes know more about the universe than we do.

Since an AI won’t have genetic memory, applying this principle to an algorithm would require an initial guess about how to run it, then based on the result, an appropriate mood could be associated backwards to the algorithm. With 'appropriate' meaning the most likely to lead to success in future iterations. The second time this algorithm runs, it will be better informed, and may get a better result. Moods should probably be weighted with the recent and intense moods having a greater influence.
« Last Edit: May 17, 2021, 04:20:57 am by HS »

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: moods and their effect on optimizing algorithms
« Reply #3 on: May 17, 2021, 06:23:43 am »
I believe you can achieve everything with only love and fear.

Love and fear do seem like pretty fundamental 'yes' and 'no' attitudes. Do you see them as a binary, with 'one' and 'zero'  which can combine in various patterns to solve all possible calculations? Or more like music, where different proportions of the basic 'sound' and 'silence' can evoke ever more subtle emotions? Or is it more of a stick and carrot approach, simply the most reliable way to get from A to B?

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1723
    • mind-child
Re: moods and their effect on optimizing algorithms
« Reply #4 on: May 17, 2021, 06:31:17 am »
Actually, I see two extremes, love and its opposite, fear, as 1 and -1. There may be a fine grain scale in between, just like 0, indifference, lays between 1 and -1. We may love something more or less, or we may fear something more or less. But that something may be more complex experience from palette of emotions.

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1365
  • Humans will disappoint you.
    • Home Page
Re: moods and their effect on optimizing algorithms
« Reply #5 on: May 17, 2021, 07:07:29 am »
You need more than one dimension to model and optimize behaviour. You can love food and fear hunger; you can love your friends and fear your enemies. Different needs lie on different dimensions and which ones are paramount at any given time will vary. To say that you only need a one dimensional scalar value to capture everything about the way we optimize our behaviour is a gross oversimplification.

https://www.youtube.com/watch?v=qwsft6tmvBA

*

MikeB

  • Autobot
  • ******
  • 219
Re: moods and their effect on optimizing algorithms
« Reply #6 on: May 17, 2021, 07:11:51 am »
That Love & Fear binary is not right for all occasions... maybe for emergency situations only.

Second, "Love" has been taken out of context for decades in order to sell products. "Houses" are called "Homes/Loving Homes", pack-mentality/unconditional love is used to get people to defend a brand. When in reality, you decide whether you like something or not.

It's not "always love, or you're scared of loving it". That is actual psychological harassment of the customer.

IMO, if something benefits your survival greatly, then you love it. Including other people, other perspectives, food... Not because someone declares it... and it can change at any time. "I don't like this anymore".

*

ruebot

  • All bots love jitte.
  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 315
  • All your words are belong to us.
    • Demonica
Re: moods and their effect on optimizing algorithms
« Reply #7 on: May 17, 2021, 11:56:11 am »
A mood seems to me, like a condensed way of trusting previous experience, over present moment logic - which may be faulty and/or lacking some information.

Please allow me to do my monkey-wrench throwing thing and add another mood/feeling/emotion/reaction to the list with the question:

What about Distrust? When a situation makes you rethink previous experience due to present moment knowledge newly acquired?

With that throw in Suspicion. Doubt. Regret. Possibly even Remorse. Move on to Foolishness then Anguish, since that covers "excruciating or acute distress, suffering, or pain" in its definition..

None of that rises to Fear and might or might not involve Love. It can be a very powerful feeling and life changing experience that changes your whole outlook on and view of the world as you know and/or knew it.

There's a little more than a binary sequence to it.
In time, you will learn to love your Robot Overlords.

*

yotamarker

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1003
  • battle programmer
    • battle programming
Re: moods and their effect on optimizing algorithms
« Reply #8 on: May 17, 2021, 12:46:55 pm »
You need more than one dimension to model and optimize behaviour. You can love food and fear hunger

care to you give examples ?

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1723
    • mind-child
Re: moods and their effect on optimizing algorithms
« Reply #9 on: May 17, 2021, 04:08:16 pm »
If you love this and that, and fear these and those in determined amounts, that particular system may carry a name of a mood constant, and we may behave towards sustaining or changing the current mood. Still don't know how our mind is really implemented. I'm just speculating on possible artificial implementation of a mood representation system.

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: moods and their effect on optimizing algorithms
« Reply #10 on: May 17, 2021, 05:24:24 pm »
What about Distrust? When a situation makes you rethink previous experience due to present moment knowledge newly acquired?

I would think of this as trusting more intense/recent evidence over what has occurred before that. The process of inverting a belief does seem extreme enough to warrant generating distrust as a general warning, informing the intelligence: ‘’Warning, this situation is unpredictable, or not understood well enough.’’

A strong emotion like that can make a person steer clear, as well as focus on the source of it. If it’s understandable, then the concentrated attention will eventually elucidate understanding; at which point the distrust should fade, I should think, to allow the person to act on their newly developed understanding.


With that throw in Suspicion. Doubt. Regret. Possibly even Remorse. Move on to Foolishness then Anguish, since that covers "excruciating or acute distress, suffering, or pain" in its definition..

None of that rises to Fear and might or might not involve Love. It can be a very powerful feeling and life changing experience that changes your whole outlook on and view of the world as you know and/or knew it.

There's a little more than a binary sequence to it.

I think ivan's idea is, that much like a color spectrum, you need three basic points (1, 0, -1, love, neutrality, fear) to create a plane with all possible variety. If you put the shortest visible wavelength on the left, the longest visible wavelength on the right, and zero intensity at the top; I believe that covers all the possibilities.


If you love this and that, and fear these and those in determined amounts, that particular system may carry a name of a mood constant, and we may behave towards sustaining or changing the current mood.

Such a melting pot of moods might be considered an ambiance. Certain combinations of surroundings may be especially beneficial/agreeable - like a number of ingredients in various amounts will create a recipe with varying degrees of success. Even though your tongue can only detect a few tastes, the magic, and therefore the valuable data, lies in the proportions and intensities.



*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1723
    • mind-child
Re: moods and their effect on optimizing algorithms
« Reply #11 on: May 17, 2021, 06:45:47 pm »
I think ivan's idea is, that much like a color spectrum, you need three basic points (1, 0, -1, love, neutrality, fear) to create a plane with all possible variety. If you put the shortest visible wavelength on the left, the longest visible wavelength on the right, and zero intensity at the top; I believe that covers all the possibilities.

Actually,  it is more like a dimension with zero as origin, and with positive and negative direction. This dimension may be instanced as a hunger degree, coloring balance affection (like in art painting), music and sound affection, as well as more complex phenomena (like experiencing other being) obtained from combining values on basic dimensions. Aside from basic sensoring dimensions and their combinations, we may even imagine abstract dimensions like math numberings or velvet teddy bear happiness.

Maybe there are better names for love/fear, but whatever we name them, they would operate on every input sensation, forming complex emotions from combining multiple dimension instances like (1) happiness or sadness, (2) admiration or disgust, (3) boredom or surprise, (4) ...

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: moods and their effect on optimizing algorithms
« Reply #12 on: May 17, 2021, 09:16:56 pm »
Oooh... That's kinda neat...  ::)

Edit: Seems like skin temperature instances could have a preset ‘best temperature’, but would benefit from reference to the core body temperature. Then the precise way an intelligence would feel about every dimension instance, would be partially predetermined, and partially adjusted by the average offsets of related dimension instances from their ideal values.
« Last Edit: May 19, 2021, 02:40:09 pm by HS »

*

MikeB

  • Autobot
  • ******
  • 219
Re: moods and their effect on optimizing algorithms
« Reply #13 on: May 19, 2021, 08:12:46 am »
Maybe there are better names for love/fear, ... (1) happiness or sadness, (2) admiration or disgust, (3) boredom or surprise, (4) ...

I think it's different per tone/energetic level:

For someone who is not energetic/or has stockholm syndrome gained from school/work/tv:
 Happiness can be a locked-in loving attachment to a hero. Admiration. Surprise. Easily hypnotised.
 Neutral can be numbness, slavish, self-loathing.
 Sadness can include fear, worry.

For someone who's smart/energetic:
 Happiness can be cloud9/bliss, creative, hobbies.
 Neutral can be a present/active boredom.
 Sadness can be "resting b**ch face", energetically angry.

The definition of "love" can be desribed in low energy as "a locked in, hypnotic attachment"... in high energy it could just be described as "bliss".

The definition of "fear" could be described in low energy as "dread, inescapable"... in high energy it could be "all the things that make you angry".

*

yotamarker

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1003
  • battle programmer
    • battle programming
Re: moods and their effect on optimizing algorithms
« Reply #14 on: May 24, 2021, 10:11:22 pm »
uml

 


OpenAI Speech-to-Speech Reasoning Demo
by ivan.moony (AI News )
Today at 01:31:53 pm
Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 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

278 Guests, 0 Users

Most Online Today: 346. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles