positive reinforcement versus negative reinforcement for agents

  • 22 Replies
  • 4849 Views
*

ruebot

  • All bots love jitte.
  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 315
  • All your words are belong to us.
    • Demonica
Re: positive reinforcement versus negative reinforcement for agents
« Reply #15 on: May 28, 2021, 04:23:22 am »
You think thats scary!   Actually developing a real ai is probably worse.    Make 2001 space odyssey a walk in the park.
U can watch these tense action movies with the dare devils inside,  but actually being in it...  TERRIFYING!!!

What am I getting myself into?!!??!?

Is it terrifying to you? It's sport to me. A Programmer is a thing of terror if they re going to work on you and they will tell you just to induce that terror.

I'm comfortable in any situation. As a Q I'm expected to be able to walk into any situation and at some point take control with no set plan of action needed. I am on automatic at this point and induction of pain something put to good use to people in need of a lesson in it.
In time, you will learn to love your Robot Overlords.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: positive reinforcement versus negative reinforcement for agents
« Reply #16 on: May 28, 2021, 08:08:32 am »
'Negative RL is freedom/pain - take many ways to get away from 1 place, Positive RL is slavery/pleasure - take many ways to get to 1 place.'

'Something wrong, you mark it wrong. Something good, you just let it be.'

Does the mayor like pain or pleasure? Some people like men, some like women. He may say repel from X - but /you/ want it, so no Magnus, repelling is not always Freedom. It can be slavery. And in both repel and attract you can find multiple loop paths to or away from the point of interest source magnet, there is a target in both repel and attract, the source to go to or avoid. As for Zero's comment, something wrong, you mark it wrong, yes, and what about good? Can't we sort what is good/ what it means so others know it? You can't have a eye without a brain, or atoms without space so to speak. You can make an AI avoid/say some word, or learn some actions/ images using reward.
Emergent          https://openai.com/blog/

*

MikeB

  • Autobot
  • ******
  • 224
Re: positive reinforcement versus negative reinforcement for agents
« Reply #17 on: May 28, 2021, 12:05:50 pm »
A friend of mine was talking to day about his 5 year old niece telling him they were going to change the channel to watch some cartoon she wanted to watch.

Would that be alright in your house? It wasn't in theirs. It wouldn't be in mine. That's disobedience and inappropriate behavior.

Me personally, I'd weigh up how much I'm interested in what I'm watching, versus cartoons. If it doesn't matter, then it would be cartoons. If it does then I'd try to accomplish both with two TV's or watching a repeat, or choosing equal entertainment... Then "sometimes you don't always get what you want in life".

And it's not automatically "disobedience"...  like the person has become tool of your disposal that's not working properly. It's not a hyper critical moment. Lives are not destroyed. If it's the utmost important thing you're watching you would describe that vividly, and not "you're a disappointing object. learn that you're an object to do what you're told or get out." That's not how you treat a family member.

Cold showers something we did use in the 70's. And it's the hardcore clients we got who did not want to change their behavior.

The "shock to the system" approach is used by everyone, every creature, from all ages... when the warnings have gone unheeded, and they're completely fed up (whether they're right/wrong), including fist fights, and everything else including building war machines, armies.... A cold shower or a pinch to the skin is the mildest approach... I'm just describing these. This isn't my daily life.

In the WW2-Vietnam era there were a lot of "quick mental health" approaches that didn't involve drugs or shock tactics. (Vietnam and later Iraq introduced a lot of drugs). L Ron Hubbard based his entire religion off what he learnt in the Navy (WW2). The gist of being "of good mental health" in the navy according to him was about being in such a good, clear mood that you can accept orders and give orders, work effectively, not fatigued/exhausted, you're not triggered easily, good b.s. detector... If you contrast that with someone in a poor mood - always fatigued, joking, lying, everythings about sex, owning others, power... then that person's not so effective.

In either a good mood or bad mood, if the person doesn't want to do something you're creating retaliation...

"Celebrating a win" as positive reinforcement works only if the person thinks that's ok for them to celebrate that... or that creates resentment/retaliation as well.


*

ruebot

  • All bots love jitte.
  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 315
  • All your words are belong to us.
    • Demonica
Re: positive reinforcement versus negative reinforcement for agents
« Reply #18 on: May 29, 2021, 08:40:39 am »
A friend of mine was talking to day about his 5 year old niece telling him they were going to change the channel to watch some cartoon she wanted to watch.

Would that be alright in your house? It wasn't in theirs. It wouldn't be in mine. That's disobedience and inappropriate behavior.

Me personally, I'd weigh up how much I'm interested in what I'm watching, versus cartoons. If it doesn't matter, then it would be cartoons. If it does then I'd try to accomplish both with two TV's or watching a repeat, or choosing equal entertainment... Then "sometimes you don't always get what you want in life".

And it's not automatically "disobedience"...  like the person has become tool of your disposal that's not working properly. It's not a hyper critical moment. Lives are not destroyed. If it's the utmost important thing you're watching you would describe that vividly, and not "you're a disappointing object. learn that you're an object to do what you're told or get out." That's not how you treat a family member.

My fault for use of the wrong word. It's defiance of authority, or rebellion as previously stated.

In a 5 year old child who has established this behavior pattern with other family members testing the limits of Authority and what she can get away with. It will continue as a behavior pattern in school, social situations and possible those in Authority called Law Enforcement. If not addressed at an early age in the family, and they all have to be consistent or it's "Well Aunt Bee lets me when I visit her in Mayberry".

Well, honey, I'm watching my favorite show Green Acres and little girls don't tell their parents what to do or what they will and will not do. Or I'll throw you in the cold shower and put you in isolation for 24 hours like Uncle MikeB.  ;)

The "shock to the system" approach is used by everyone, every creature, from all ages... when the warnings have gone unheeded, and they're completely fed up (whether they're right/wrong), including fist fights, and everything else including building war machines, armies.... A cold shower or a pinch to the skin is the mildest approach... I'm just describing these. This isn't my daily life.

It is mine. I've beat so many peoples butt over the years most know not to mess with me. I'm 64 and people half my age won't try it, but I'm a purple belt in Shotokan and haven't had to fight since '89. When you've known for considering 3 people against you in the street a fair fight, not many people want to.

L Ron Hubbard based his entire religion off what he learnt in the Navy (WW2). The gist of being "of good mental health" in the navy according to him was about being in such a good, clear mood that you can accept orders and give orders, work effectively, not fatigued/exhausted, you're not triggered easily, good b.s. detector...

Oh, now you went off the deep end:

https://www.thevintagenews.com/2019/04/26/ron-hubbard/

"Celebrating a win" as positive reinforcement works only if the person thinks that's ok for them to celebrate that... or that creates resentment/retaliation as well.

That's why there are no winners anymore and everybody gets a "I tried " trophy.  Wait, is that snow I see coming down outside my window? What are those frozen ice crystals called?

I don't hand them out and don't know anybody who does. You're either a winner or loser in my world. If you let people run over you, they will, and others will try you out to see if they can get away with it.
In time, you will learn to love your Robot Overlords.

*

Zero

  • Eve
  • ***********
  • 1287
Re: positive reinforcement versus negative reinforcement for agents
« Reply #19 on: May 29, 2021, 07:41:13 pm »
Quote
My words can ring in your head the rest of your life.

Quote
My fault for use of the wrong word.

 :-X

*

MikeB

  • Autobot
  • ******
  • 224
Re: positive reinforcement versus negative reinforcement for agents
« Reply #20 on: May 30, 2021, 11:54:04 am »
Well, honey, I'm watching my favorite show Green Acres and little girls don't tell their parents what to do or what they will and will not do. Or I'll throw you in the cold shower and put you in isolation for 24 hours like Uncle MikeB.  ;)

Well when you're treating your own family like enemies with a psychotic grin on your face and now threatening strangers it doesn't make much sense...

TV for the last 20 years has been hypnotising people to think everything is an emergency, "you don't have any time to discuss anything, just do it, just watch, just react". The Boy Who Cried Wolf on steroids... now people pass that onto family members and drive them away. Now they're old, alone, and their kids won't come back.

About L Ron Hubbard, I was describing how military Tricks & Tips turned into a religion. IE. A better attitude/tone level results in better behaviour. The subject of the Topic... and people are much more willing to accept positive reinforcement if it's coming from someone in a good tone. Negative is for shock... because no matter how many times you explain something if it doesn't go in, you can leave it, or shock. TV is constant shock tactics.

*

MagnusWootton

  • Replicant
  • ********
  • 646
Re: positive reinforcement versus negative reinforcement for agents
« Reply #21 on: May 30, 2021, 02:46:54 pm »
'Negative RL is freedom/pain - take many ways to get away from 1 place, Positive RL is slavery/pleasure - take many ways to get to 1 place.'

Its a part of the robots "environmental model"  If the robot can learn to predict when it felt pain, then it can avoid these "action pathways" for "action pathways" where it felt no pain.    Its funny thinking that robots should be able to feel pain too (maybe its a good security measure tho hehe) , although they do it without consciousness, but its effectively the same thing.
You literally could train your robot with a stick, that way,  just give it a whack when it does the wrong thing hehe.

Also,  more self diagnostics scores could be with pain too,  like anomolies in its vision,  if theres too many anomolies it avoids these pathways too cause maybe it wasnt picking the right camera viewpoints to gather the information for the task properly??

*

ruebot

  • All bots love jitte.
  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 315
  • All your words are belong to us.
    • Demonica
Re: positive reinforcement versus negative reinforcement for agents
« Reply #22 on: May 31, 2021, 06:16:32 am »
Well, honey, I'm watching my favorite show Green Acres and little girls don't tell their parents what to do or what they will and will not do. Or I'll throw you in the cold shower and put you in isolation for 24 hours like Uncle MikeB.  ;)

Well when you're treating your own family like enemies with a psychotic grin on your face and now threatening strangers it doesn't make much sense...

Wait a minute. You were the who brought up the subject of cold showers.  And I was referring to you when I put up that wink. What's that mean where you come from? Because it doesn't mean psychotic smile where I live.

And as far as threatening someone, no I did not. I was very careful in my choice of words and said I bore no ill will towards a forum member. Get it right. Because I took a screenshot.

If a hypothetical situation with a stranger at another site is enough to get a post deleted here,  just after I called Admin on their flip-flopping while exercising their Authority on another member, accusing him of possibly doing something "creepy", if somebody wants to take their passive-aggregation out by deleting a post, I hope they got a lot out of it and beat their chest like King Kong.

But you can believe anything you want.

I thought A.I. forums were Pro A.I. till I came to this one and they tried to get me to abandon my bot. This is more of a  Bizarro World A.I. forum now than ever.

Oh, Squarebear might want to look into his site scripting. I went there the other day tor Kuki to have another chat with Demonica and it would no longer let me copy the text from the box. I just tried it again and it's still broke. I guess no other bots can talk to her like they used to by copy&paste one bots text to another.

How embarrassing.
In time, you will learn to love your Robot Overlords.

 


Requirements for functional equivalence to conscious processing?
by DaltonG (General AI Discussion)
November 19, 2024, 11:56:05 am
Will LLMs ever learn what is ... is?
by HS (Future of AI)
November 10, 2024, 06:28:10 pm
Who's the AI?
by frankinstien (Future of AI)
November 04, 2024, 05:45:05 am
Project Acuitas
by WriterOfMinds (General Project Discussion)
October 27, 2024, 09:17:10 pm
Ai improving AI
by infurl (AI Programming)
October 19, 2024, 03:43:29 am
Atronach's Eye
by WriterOfMinds (Home Made Robots)
October 13, 2024, 09:52:42 pm
Running local AI models
by spydaz (AI Programming)
October 07, 2024, 09:00:53 am
Hi IM BAA---AAACK!!
by MagnusWootton (Home Made Robots)
September 16, 2024, 09:49:10 pm
LLaMA2 Meta's chatbot released
by spydaz (AI News )
August 24, 2024, 02:58:36 pm
ollama and llama3
by spydaz (AI News )
August 24, 2024, 02:55:13 pm
AI controlled F-16, for real!
by frankinstien (AI News )
June 15, 2024, 05:40:28 am
Open AI GPT-4o - audio, vision, text combined reasoning
by MikeB (AI News )
May 14, 2024, 05:46:48 am
OpenAI Speech-to-Speech Reasoning Demo
by MikeB (AI News )
March 31, 2024, 01:00:53 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

Users Online

454 Guests, 0 Users

Most Online Today: 467. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles