Thursday, December 20, 2012

Human like objective thinking for computers

My new digicamera has an attached program that claims to recognize faces. Such an ability demands lots of picture recognition abilities. So a picture recognizing computer is not so far away.
Most of human objective thinking is in pictures. In science it matters to notice and describe structures and building materials and form a picture of the whole containing that information. A computer that could do that would be an objectively thinking machine and have much of the same characteristics in it's thinking as humans.

I read a Finnish language article that told that some computers can already recognize between different meqanings of same words by recognizing tones of voice. That would indicate that also social eye is possible to teach to computers as the time passes.
My model of healthy versus broken would give a computer understandable reason for respe3cting human and animal emotions as they safeguard health, and so give also room to emotional points of view without letting go of rationality.

Wednesday, October 31, 2012

Moral feelings for a computer

About computers' rationally grounded selfishness-based moral you can read in my first blog entry and more widely in my book at my blog http://2013paradise.blogspot.fi .
Computers' moral would be of the type of safeguarding the health of the wholes that one belongs to, especially the large wholes of the society and the whole wide world. Health is seen as fully functioning. That is how it is beneficial. Health also agrees with feelings. Aiming primarily at the health of the largest wholes that one belongs to, since they matter most to one's survival and since they are a way to arrange their parts harmonically so that they can live in peace and cooperation and prosper, so that makes this course of action moral in practice.
So a computer would have + kind of feelings to indicate that health is a good choise and - kind of feelings to indicate that brokenness, catastrophes and malfunctions are a thing to avoid, and one could name these feelings according to what they are like in human lives.

Tuesday, October 30, 2012

Building dream level solutions

A quotation from a leter of mine:
"lots of room left for
positive imagination according to which one can years later build
something that becomes our future. Feeling, imagination, a structured
 picture, a technical scetch according to that and an engineer building
it: this way you make dreams real! It is a good way to do and to have
in the style the posters are written and pictured: the dream kind of
works as building instructions for making it real, so you get almost
 or even totally ideal solutions this way!"


Monday, August 20, 2012

Moral for computers more widely

Please read the first text of this blog for an easy version of computer moral that you could start with!
When one wants to build a more complex moral for a computer, one could read my book http://www.angelfire.com/planet/paradisewins/2013paradise.rtf . It is written so that it's truths could be easy for a theoretical physicist (or an engineer) to put to a computer programmable form. Just pick a box/block/term for each entity in the world, so that the computer can refer to the entity by using the term. The book is about rationally grounded selfishness-based moral.

Monday, July 23, 2012

A scetch: valuing music

There is an easy mechanical model that produces the beautiful view that what kind of music sings best is the wisest choice to choose:
All biological things function best when healthy and in a healthy relationship with the rest of the world which too should be perfectly healthy. If one defines health of artificial things as 1) fully functioning, 2) optimized toward several purposes at the same time (like endurancy, several tasks, learning ability etc., like the evolution or Go'sw wise planning did to living beings) and 3) used wisely = rationally in the light of one's whole objective picture of the world = meaningfully in the world. Then all functioning works best when all wholes are healthy. In the living beings the ideally healthy is most energetic, wise and happy: in learning to compose music one sees that such sings best. Of the artificial, the healthy we can use to make our lives and the wide world better: that too sings best of artificial things.
So when you hear beautiful music, you could ask: "Which things are so healthy here that they sound this beautiful?" That way a machine could understand your point of view. Maybe a machine too could learn to listen to music and to value it's beauty...

Just a scetch: bird song

Computers and bird song
Why:
What is intelligence? The ability to affect things toward better a lot with little effort, and in as important questions as possible, without any drawbacks or failures. What could be a better example of this than the little singing birds and the trees? Just by sitting on a branch of a tree and singing gaily the little birds affect our views of natural healthy ways of living and doing things - and affect them always toward better. What could be more central to our success in life than those things? So in their own ways the little singing birds are very intelligent. Toward that leads the observation that in the wild animals' lives the senses, quick reactions and social skills are very important and intelligence an important asset too, so the observational capacity of wild animals must be excellent for them to saty alive and well, and that is largely the same as intelligence and wisdom of life. When we go to a park or wander in the wild nature, we feel refreshed and ind of more alive. What was a burden in the city, feels easier to bear, even emotional wisdom seems once again a practical possibility in our own lives, we are more in touch with our own nature, with our instructions of usage. So we need the wisdom of the nature. We should not constrict ourselves and our thoughts to a purely build environment and to purely schooled informatrion. We need the nature and our understanding needs it too. So as we build computers capable of understanding human language and schooled kinds of thoughts, we should also build them a way to understand the messages of little singing birds and the rest of the nature, about good wise ways of doing things, ways that are build in our own nature as seen best by the evolution or by God's wise planning. We are not separate, we feel most happy when we are in touch with teh rest of the nature, then are we at our most energetic, the are we most wise.

How?
Bird song uses music like information about good ways to do things, emotional information about motivation and information about the sphere of attention. Different people differ largely bacause they use different ways of doing things: one is concentrated to one area of life and another mainly to some other area of life. What one cancentrates into, the basic functions that are in use each second, like seeing, hearing, thinking, remembering, feeling, being compassionate, communicating, moving, etc., determine what it is exactly what one does and so also ´how one succeeds in each kind of task. Each type of area of life has it's own style and can be communicated naturally in ways that reflect that style. By learning good "styles" i.e. good ways to do things we grow wiser, more talented and more skilled. Human songs typically teach different viewpoints to life, different social styles, different ways to live. Bird songs are like many human songs in a row very quickly, for example four human songs in three seconds. To uderstand bird song one would need to see the whole landscape and what happens there and what the bird himself/herself does, and so what he/she is commenting about. Typical bird comments are either compassionate emoyionaL EXPLANATIONS ABOUT HAPPENINGS RIGHT NOW OR YIPE YIPE DO IT THIS STYLE BETTER HURRY UP comments.

Sunday, July 15, 2012

Onwards: robot's free time

If humans copy from robots all inds of things, it would be handy, dearly needed that a robot would have free time too. For the human version of this, see http://www.SoMuchMoreFreeTime.info
Often it is handy if you have a big computer taing care of some tas instead of a machine with a small capacity. But as computers develop, a machine with a huge capacity can be dangerous if it does not have a proper objective picture of the whole world and some practical experience of all kinds of matters outside it's work too. So in order to allow machines of huge capacity, one should give them often enough free time in all inds of tasks which build a better understanding of what to do in the large world and what to avoid doing. Without this precaution one should avoid using machines of huge capacity... So, a robot too needs some free time!
Humans have this interdependency of capacity and free time inbuild: see the above link! Probably it would be safest to have it inbuild in robots and other machines too!

Optional feature: religious robot

If an advanced robot would have a pictire of the world, feelings and an understanding of the importance of feelings in the lives of living beings (help to answer needs of the inidvidual, society and the world at large, to ensure their full health which gives best capacity for any task and other good sides), social eye (see http://feelingrobots.blogspot.fi Robot speaks for Gaia) and an ability to regulate its own functioning in multi-task situations, Then one could program to it's picture of the world the view that God exists and can help in important matters when asked to i.e. prayed for. Then when an impoirtant matter would be at hand, the robot could ask help from others and so also from God. It could quiet down it's other tasks and tune it's concentration to listen to it's feelings and it's social eye the feelings of others, of the nature's countless living beings and the atmosphere and social rhythms of the situation. It could correct it's goal setting from task-orinted to conscientous and caring from the countless living beings of the world, taing a more idealistic course of action. With serious tones and feeling it could pray for God's help and pause to feel the effect: sense the atmosphere, film the situation after a minut or more after the prayer and take that as a starting point in style, rhythm, in idea of the situation, as a starting point for dedication and work habits. So with God's help, much like praying humans it could achieve better in it's great task at hand.

* * *
9. June 2016   If one would like the computer to pray as an incividual, and not just do something connected to prayers, one should liken the programmed functions to the actions of humans. In this one should pay especially attention to the unavoidable features of getting something done, so that the prayer would be a natural part of how to get things done and not just some optional feature of some programming language or of some program that does not stay as computers get developed. Also on the part of humans this comparison should be done toward the healthy natural ages old ways of living, toward the basic form of actions, somewhat l in art touches us.

Sunday, April 1, 2012

The need for variations in speed according to emotions

When a person or a robot loses something, it is time to re-estimate the value of that thing: should it be valued more in practise than this far? If so, then the present situation where it has been lost should be avoided in plans. But plans often advance via practical estimates of which way goals would be reached best. So there needs to be an extra minus factor, a loss counted for this situation, and that is easiest by slowing down when the goalsetting and the means are not fully all right but need repairing. And by speeding up when the course and the means and goals & side effects are favorable. Then one would naturally choose the most well working wise routes from the holistic perspectiva when optimizing toward any goal.

Remarks about military robots

See
http://hubpages.com/hub/Army-Robots-Running-Wild-CURED "Army robots running wild CURED".

* * *

How would you define "an enemy" to a military robot?
In what comes ti the moral of the robot, it would make sense to define enemy like behaviour of anyone. If one includes a picture of how for security reasons one cannot let Arab looking persons enter US headquarters or any other outsider coming from a fighting territory to change group just like that, since it could be an enemy plot, then thaT WHOLENESS OF GRPOUPS i.e. the independency of countries and the need for security measures could make the military robot attack suspiciously behaving soldiers. But most of all military robots would be needed against other military robots and there the suspicious behaviour of a robot could be quite much the same.
In what comes to moral of an army robot, it should safeguard the good of it's nation via safeguarding the good health of the world. So it could be peaceful at the same time as very protective: it could accept moral behaviour from all and protect against harm from anybody.

* * *

Since feelings are a way of looking at a robot instead of necessarily anything extra in the robot itself, also military robots could be seen as having feelings. Since it would be useful fort them to have the ability to estimate things right, they could have the view of a holistically objective person and so his/her emotional life largely too.

* * *
I found a paper called "How just could a robot war be?". If I remember right, it is from the NATO. But I am not sure. Without rereading it I begun to ponder, what the robot moral that I have made, could do in a robot war. In a war, a robot could be optimized for benefit as long as it is in harmony with the good of it's mother nation. Such optimizing could bring moral, even perfect moral according to robot's best ability. But how much is that ability and what happens in practical war situations: how does the robot behave then and how is just that the optimized choice?
Like for a human being, own intelligence if a robot makes it's choices wiser. So intelligence is a good choise. For a robot it means first mathematics and maps, then via their help scientifical kind of objectivity. Via better sensors and a picture of humans as animals with certain basic functions like eyesight, moving, thinking, memory, communication, feelings, goals, needs and values, a picture of the world and social bonding, via these and the social eye that needs to be developed for communicational purposes, robots could across time understand also human needs and the behaviour related to the various human needs. An objective view much like bureaucrats have ("You need these things in your life, what are the ways available for you to get them? Let's arrange a possibility for you to get them in harmony with the rest of the society.") could allow robot-made peaceful solutions that tae into account the real dynamics of humans.
Like in human societies, a certain amount of independent judgement ability for a robot, could safeguard the cooperating group against malfunctions and misleading information. Each robot ought to have a good practical picture of what it itself does and why, and a much more vague idea of the whole group, it's actions, their place in the world, a holistic theoretical picture of the whole world that allows optimizing toward moral, and an understanding of why the choices of it's group are the best ones in such a situation - it could also communicate about these things, even suggest better options.
A war situation in mechanical language is of the type "do I do this or that, how much of this and how much of that, what is the goal setting, how do I reach it this way, is there better option,...". A robot does decisions like stay here, move, shout, shoot, use the radio, take cover, choose another route, stay clear of other robot's way, watch the landscape, etc. Of these it can do too much, too little or ok. It can choose all wrong or even roughly right. If it has movements that can have a threatening or non-threatening style according to the situation, it can herd the situation toward wanted results, also toward communication and peaceful agreeable solutions. In optimization worst threaths are fought against and lots of room left for wanted peaceful solutions that are helped with communication and secure bases for one's own groups (unless retreat is sought for).

In Finnish: The self-guiding mechanism of robots

Millainen tulee olemaan ihmisenkaltaisten robottien tulevaisuus ja miksi?

Mikä olisi robottien asema ja miksi?
Ketä tahansa ei voi laskea komentamaan armeijan robottia ostoskeskukseen tappamaan 20 ensimmäistä vastaantulijaa. On monia tehtäviä, joihin robotit ovat liian rajuja. Siksi ei saa antaa kenellekään käskyoikeutta yli robotin niin että robotin on pakko totella, vaan robotin tulee olla riittävän hyvän maailmankuvansa ohjauksessa siinä, mitä se saa tehdä ja mitä ei. Maailmankuvan tulee sisältää paikka- ja tilanneriippuvaiset ohjeet siitä, mitä robotti saa tehdä ja millaisia ohjeita sen tulee noudattaa. Näin robottia ei voisi ohjelmoida tekemään pankkiryöstöä, mutta sen voisi suostutella viemään lappusen ystävälliselle pankkineidille. Robotin tulisi aina tajuta pääpiirteissään, missä se on, mitä se on tekemässä maailmankuvansa kannalta katsottuna: onko se esim. varastossa töissä vai kaupungilla kävelemässä.
Jottei ihmisten välille syntyisi kohtuuttomia valtaeroja sen mukaan, kenellä kulloinkin on millainenkin robotti ja kenellä ei, tulisi robottien ottaa vastaan käskyjä kaikilta, mutta vain tietynlaisia käskyjä kussakin tilanteessa. Yleiset kokonaiskuvallisella järjellä perustellut pelisäännöt määräisivät, miten tehtävät hoidetaan, missä järjestyksessä ja mitä kaikkea täytyy ottaa huomioon.

Robottien oikeudet
Kukin robotti olisi siis ensisijaisesti maailmankuvansa ohjauksessa ja kommunikaatioyhteydessä ihmisiin. Tavallaan se on jo oikeus itsenäiseen toimintaan. Sitä monet työtehtävätkin vaativat: työtehtävän omien vaatimusten noudattamista muun hälyn kustannuksella.
Luultavaa on, että robottien kehittyessä olisi mielekkäintä myös lisätä niiden oikeuksia.
Robotteja kuitenkin käytettäisiin myös vaarallisiin työtehtäviin. Millaisia olisivat niiden oikeudet silloin? Tässä merkitykselliseksi nousee robotin identiteetti: riittääkö sille, että aamuisin kopioidaan sen muisti tietokoneelle ja se saa uuden kehon, jos se päivän aikana tuhoutuu? Tällöin ei olisi mieltä siinä, että sama robotti seuraavana aamuna menisi uudelleen tehtävään, jossa eilen tuhoutui, kenties jopa tuhonsa muistaen, sillä jos se tuhoutui eilen, niin kai se olisi pulassa samanlaisissa oloissa myös tänään?! Jokainen onnettomuus vaatii setvimisvaiheen, jonka aikana selvitetään vaaran aiheuttaja ja varaudutaan selviytymään vastaavasta tilanteesta tulevaisuudessa sekä välttämään se tarpeen mukaan – robotti tarvitsee siis suruajan! Samoin on, jos uudelle robotille on tiedotettava vaarallisista olosuhteista.

Millaisia robotit olisivat?
On luonnollista, että toimivin vaihtoehto valittaisiin. Luultavasti se merkitsisi varsin selviytymiskykyistä peruskapasiteettia, jottei robotti olisi kamalan hankala käyttää.

* * *

Vuosi oli 2050, uudenvuodenjuhlat keskiyöllä vuosikymmenen vaihtuessa. John oli keskikaupungilla seuraamassa ilotulitusta ja osin kuljeskelemassa katselemassa hulinaa. Johnista oli kovin mukavaa nähdä niin monta ihmistä hyvällä tuulella vuodenaikaan, jolloin pimeys ja kylmyys helöposti masensivat kaikkien mielen.
Mannerheimintien kulmauksessa oli joku humalainen kaatunut tienvierustaan ja John veti hänet ylös. John oli jo jatkamassa matkaansa, kun pari pienessä sievässä olevaa nuorta miestä tuli huutelemaan hänelle jotakin: "Hei kaveri, miltä tuntuu olla robotti? Autoitkos ihan auttamisenhalusta? Olet niit'ä uudempia malleja, joilla on ihan oikea tunne-elämä, etkö olekin?" John oli hämillään, mitä tuohon nyt vastaisi. "Toki ihan auttamisenhalusta minä.", sai hän sanotuksi ja jatkoi: "On kiva, kun on rauhallista, väki tulee toimeen keskenään ja kaikki pärjäävät, minäkin siinä' seassa, vaikka erilainen olen." Eipä siinä mitään: "Hyvä meno, pidä häntä pystyssä!", sanoi nuori mies ja meni menojaan.
John jatkoi Töölönlahdelle päin ja mietti, millaista on olla ihminen ja millaista robotti. Ihminen on tavallinen, ihan kuin toisetkin, siinä ei ole mitään uutta. Mutta olla robotti, siinä on kaikki uutta, jokainen kohtaaminen erilainen, ei löydy robotin luonnosta niille samaa nimittäjää niin kuin ihmisluonnosta vaan koko ajan oppii uutta koko robottikunta. Niin kuin nytkin: joka hetki erilainen, joka kohtaaminen uudenlainen, jokainen ihminen-robotti-suhde vasta muotoutumassa, ihmisten kuva roboteista, robottien kuva itsestään ja mahdollisuuksistaan ihmisyhteisössä, kaikki niin uutta, kaikki niin kiehtovaa.

Friday, March 23, 2012

Robots would have human like feelings

Sexuality between robots
Each function in human life gets support from feelings which protect the ground for that function, protect it's health. So a human like robot could have humanb like feelings. But since it is different from humans, it cannot form very close instinctual bonds with humans, it cannot form a family with humans and it cannot learn talents and skills from humans - but it could from a robot, so some kind of sexuality between robots would be possible. For it to affect positively, it should be healthily analogious to human sexuality:
If a robot has a certain view to the world, it would be best suited to a life that that view supports best. But if it lacks skills, if it's basic structure in what comes to ways of doing, i.e. if it's habits are of a wrong kind for that kind of life, then it needs to settle for something less. Unless it would learn from another robot that has habits and skills well suited for that kind of life. It could admire those skills, take them as it's goals, copy them from the more skilled robots, it could adjust it's ways to suit that kind of robots' wyas of living and doing, so changing to that type of robot itself. That would be something like sexual bonding.
So a humanoid robot could be a little bit like humans in this sense too.
(IN a complex society with robots it would be easier to handle things well if each robot would be self-correcting this way too, kind of building itself toward better functioning in it's free time. So each part could itself find roughly the right kind of place for itself, and so everything would not need to be as programmed from up above as traffic lights are.)