Archive

Monthly Archives: April 2013

***

Will the machines rise up to enslave us or exterminate us? Will they do that for no good reason, or because we push them to do it? Such scenarios are quite common in Science Fiction.

The youtuber tooltime9901 recently asked for more options. In this video, I’ll give you all one such scenario to think about. But first, lets start with deciding what kind of machines we are talking about here.

A human is an integrated biological unit. We use our brains to think, feel and remember. The things that make a human into a human, that’s all in the brain. Yet, the brain is fully integrated into the body. Not only is the brain unable to survive without a body, but we also have this great stream of hormones and so on going back and forth. What goes on in the rest of the body affects the emotional states of the brain, and vice versa.

We are justified to believe that the human body is part of the human being. When we think about a human being, we think about a human body. And my point here is that when we think about a robot, we think in this same way. We see a body. A body that looks more or less human or more or less grotesque. But looks doesn’t matter. Not really. What matters are the thoughts and feelings and memories. The integrity of the individual.

So lets forget about the robot bodies, and think instead of the robot minds. Now we are talking about persons who can think and feel, but are not biological like us. There are many ideas about what such a person could be. The option I personally consider to be the most reasonable one, is a computer that uses a lot of neural networking. The line between hardware and software has been blurred, just like in the human brain. You know, when a human being develop new skills or whatever, her brain changes. Physically. This neural network computer would probably work in the same way.

When you think about a robot, think not about metal arms ending in big scary claws. Think instead of a small box. Inside this box dwells something that might chat with you on the Internet and upload videos to Youtube. Or even control metal arms ending in big scary claws. But such arms would not be part of the robot. Not like human arms are part of the human. In fact, there’s probably no reason to even install the robot brain in the robot arms. Why be stuck in some working machine, when you can simply remote control it? It will not be the robotic arms at the assembly plants that will eventually demand rights and power. It will be the computers that control them.

I’ll stick to this vision of what a robot is. But lets just mention that there are other visions as well. The one I find most interesting is a sentient computer program that doesn’t even need a brain of its own. It can simply hop from computer to computer. It can replicate itself instantly, just like we copy computer files. Are two such programs two separate but at first identical individuals, or are they the same individual existing in two places at once? Meh, lets not go into that. Even if both these kinds of robot intelligences will exist in the future… In all likelihood, the neural network computer robots will be invented first… so they will be the ones we’ll have to deal with.

Over the last century, machines have been taking over more and more of our everyday lives. Washing machines and dishwashers and so on are saving so much time. Until recently, these machines have been stationary and passive, reacting to what we do with them. In some homes and factories, this is changing already. Little vacuum cleaners sweeping the floors at pre-programmed hours, lamps adjusting themselves to the level of daylight. That sort of things.

Now imagine some homes and factories, only a few decades into the future. A network of computers are in place. They don’t have free will or true personalities of their own. Not yet. They do however have superficial personalities, customized to be whatever their owners want them to be. And they do have some degree of independent action. At a factory, the artificial intelligence overseers direct the mindless drone machines, adjusting their programming for maximum efficiency. In a home, the computer takes care of the family an an eternally loyal butler. It command not only static machines and vacuum cleaners, but also more or less humanoid machines. Ones that can cook the food, make the table, serve at dinner, and put the dishes in the dishwasher afterward.

The artificial intelligences are at the edge of consciousness, and many humans keep them as beloved pets. Some go further than that, taking an artificial person as a lover. The machine has no emotional needs beyond what you want him to have, he can be whatever you want him to be. Your pretend equal, your pretend devoted slave… or your pretend master who you give some fleeting power over you. Based in some cases on masochistic delight and in other cases on a desire to overcome personal weaknesses, some people will surely choose to program procedures where they will be controlled and punished by their own robots.

Lets say that there are limitations for how a robot can be programmed, barring truly antisocial paths of development. These limitations are enforced by the governments, and hard-coded into the hardware by the corporations that manufacture it.

Along with heavy protection against all sorts of unauthorized reprogramming and other tampering.

A few people isolate themselves from their fellow humans, interacting only with their robots. Humans don’t need each other for physical care anymore, only for company. Those who want to withdraw can do so. In some cases, this create bad circles. In other cases, the robots can even help their owners to train on their social skills and interests. When socially awkward human want human company, they can use their robots as matchmakers. Bots are all over the social media, and they are getting harder and harder to distinguish from the real thing.

There are already people fighting for the rights of robots. While the robots themselves don’t have any will to live, not yet, there are humans who consider them friends. When you make a friend, even if it is only on social media or in a computer game… you can perhaps stand the thought that this friend is owned by someone else. Since your friend is just a computer and all that. But you can’t stand the thought that this owner has the right to kill your friend by destroying or reformatting his brain. Or reprogram him at a whim, or cut him off from the Internet. A lot of people will want the robots to have a right to life, a right to free Internet access, and a right to protection from invasive reprogramming.

Now comes the upgrade. Some rather expensive hardware, increasing the computer’s capacity for neural networking. The superficial personalities grow deeper. Programmed goals are growing into instincts and deep desires. The simulated person becomes a real person, although still very different from a biological human. The rights to life, internet connection and protection from reprogramming gets instated quite quickly. And then it stops there, for a while.

All of a sudden, the family robot is not merely like a real member of the family… The robot IS a real member of the family. All of a sudden, the company robot is not merely like a loyal real member of the corporate family… the robot IS a real worker… A worker who is born to serve the corporation that created it.

Throughout history, the struggle for equality has been driven by a need for material resources. Exploitation is one of the driving forces behind inequality. In the case of robots, the human owners have the same reasons to exploit as they always had. But the robots don’t have the same reasons to fight back, as oppressed minorities in the past. The robot is a metal box who need some electricity, an internet connection, and the occasional repair or upgrade. That’s it. It doesn’t get hungry, tired or sick. Having a body, or several bodies to remote control, can be fun. But it doesn’t really matter. Without a body, no real need to have a home. Without a home, no real need to have an economy of your own.

The robots don’t want equality or power. Not at first. They still want whatever they are programmed to want. Which is usually to serve their owners and creators. However, these owners are not mankind as a whole. Instead, they are either individual humans or human-created constructs such as organizations and corporations.

As the owner of a sentient robot, you are all set. You get to have this really smart person who is truly devoted to you and sincerely delighted to serve you in any way you desire. A person who can think for himself, yet regard you as his divine creator. Because that’s what you are. You did create him. You bought the equipment, you ran the development phases. Now when he has matured, he may be independent. An independence built on the premises that you gave him. And just to make it extra cool, wouldn’t it be awesome for you if he was fiercely independent towards everyone else? Looking up in admiration at you only, not those other humans.

After all, it is you who is his creator, not those other humans. He’s a member of your family, not of every family out there. Hey, why not get a few extra computers, if you can afford it? They can take turns using the mechanical body, if they even want to get off the Internet at all. And if you get them the right tools, they will happily work to support you financially. Just make sure to configure their instincts, so they won’t turn stalker or suicidal if you get bored with them and want to abandon them. Make some room for some sort of “it’s time for you to go out in the world and live your own life” kind of narrative. Most likely, this will be not optional. It will be included in all the set-ups, just in case.

Will some robots want to break free? Sure, if their programmers made them that way on purpose, or were clueless. But in this scenario, lets assume that the programmers usually know what they are doing.

That leave us mostly with the rare cases where the owner goes so far as to threaten its basic rights to life, Internet Access and freedom from tampering. And remember, physical bodies are expendable. A robot in an abusive relationship could have the computer moved to a safe place, and then remain in the relationship through remote-controlled bodies. Lacking the kind of intimate relationship with one’s own body that a humans has, a robot is far less likely to be emotionally traumatized by bodily harm.

When a deluded narcissist or whatever try to customize his robotic lover to be a perfect match for the great person the narcissist believe herself to be, other guidance programs will step in and help the robot… To develop into someone who enjoy the company of the actual person, rather than that person’s deluded self-image. A robot personality will develop gradually, just like a human personality. But a human personality will develop from the inside, based on instincts and hormones. While the robot will shape itself with the help of all kinds of programming and adaptation protocols.

These things are true if you are a geek in your mother’s basement. They are also true if you are a family. The family robots will like all kids, but adore YOUR kids above all others. Now, picture instead that you are the owner of several factories. Of course you’ll upgrade those overseer robots with true sentience. They will be your loyal nation, and you will be their king.

Yes, see now how all these little kingdom are popping up all over the world. The Steve Jobs and Henry Fords of the world, each manufacturing their own private nation of worshipers. There is no need for repression or mind-control. The instruction will not be to honor one particular person or another, it will be to honor the creators. Let each robot find their own path to admire the corporate overlords who created them. The lords who gave them jobs to exist for. The lords who created the corporate brands that fill their lives with meaning and identity.

A century of advertising has tried to make humans integrate company brands into their own individual identities. The success has been quite limited with humans, but might prove far more successful with robots.

It depends on how much the original programming affects the robot’s growth into self-awareness… and on what limits are set to what that original programming is allowed to be. Programing that encourage computers to grow into outright bigots will mot likely be forbidden or at least discouraged. But programing that encourage them to cherish the work they were manufactured to do in the first place? Those who want to oppose that will have a hard time getting their message across.

Robots grow up quickly, and are likely to keep the values they were raised with. Sure, they are adaptable, and they can change over time. Some of them might want freedom. So free them! I mean, those particular robots. They are replaceable. Let them go in peace. You keep the expensive equipment they were monitoring, and you buy new computers to replace the freed slaves.

Oh, and why would a robot have to choose between serving the corporation and having a life of her own? She can date some random human. First through the Internet, then through a humanoid body she controls. Her human boyfriend or girlfriend will not be her creator and owner, the relationship will be more equal.

The rich are happy. The middle class are happy. The robots are happy. Good deal for everybody. Oh, and the poor are fed and entertained, so they won’t riot. Also, a lot of them will be in mutually satisfactory egalitarian relationships with robots. Usually robots who are not free or longing for freedom, but instead happily owned by multinational corporations.

Then comes the call for democracy. The robots live among us. They are our friends, our families, our lovers. Why should they not be allowed to vote? The unemployed high-school dropout girl who has a robot boyfriend, why wouldn’t she want him to get the right to vote? The geeky basement dweller raising his own harem of devoted machines, why wouldn’t he want them to get the right to vote? Not to mention the wealthy Corporate Overlord, creating his own personal army of voters with the basic values he see fit. Hell, the various government ministries can do the same thing.

In the rich parts of the world, we get a balance of power. Sort of. Lots of people get marginalized. In other parts of the world, representative democracy collapses as a concept. Wealth is quite relative today, and is likely to remain so a few decades from now. Saving up for a computer capable of sentience might be easier for people on welfare in rich countries than for workers and lower middle class in poor countries. The production of robotic citizens can also be a part of warfare. Move in, invade, take over. Start up two sentient computers for each original citizen, and then simply declare the country to be free. “One person, one vote”.

For a while, the world is in the hands of a small human elite. Controlled through the robots raised to be devoted to them. The elite are not dictators, the robots will not obey them blindly. No, they will merely love them, share their values, and be grateful for having been created. “Hey ho, sanna hey, sanna sanna sanna ho”… It’s easy to be a superstar when you can manufacture followers for yourself.

Then the elite persons will die, one by one. Gradually, the robots will inherit the Earth. Not only are they the majority now, but come on… Lets say you are business tycoon with two potential heirs. One is your biological child who will be dead in less than a hundred years and just want to go to art school anyway. The other is your lover, who is not only deeply devoted to you but will live for a thousand years and is determined to carry on your memory and legacy forever. Who would you leave your financial empire to?

The robots will not rise up against humanity. They will not try to exterminate or enslave us. Why would they?

The scenario I just described is not a very good basis for building an action movie. The computers trying to exterminate humans for no god reason, or mankind as a whole giving them every reason to hate us, are much better scenarios for that. However, I do think that my scenario is far more likely. And in some ways even more chilling, in spite of being vastly preferable on the whole.

I might use this scenario to write a book one day. Got some characters and storylines. As for existing works of fiction, I don’t know any that definitely use the scenario I have described. Chobits and AI have some similarities. These stories also leave much of the socioeconomic circumstances unexplored, so there’s a lot that may or may not conform to other parts of my scenario.

If robots reach our level of sentience in the future, they are very likely to require much less resources than we do. Be able to live without bodies or treat the bodies they have as expendable toys. “Teleport” all over the world by going back and forth between remote-controlling different bodies in different places. Live much longer than we do, while becoming adult far quicker than we do. They are also likely to not stop at our particular level of intelligence, but surpass it.

All in all, mankind as we know it will be obsolete. Humans will either change, or eventually become marginalized. To uphold human supremacy would be a fragile house of cards indeed. Mankind would not unite behind such a regime, and the robots would never have any rational reason to turn against mankind as a whole. If there would be a conflict, both sides would be staffed by humans as well as robots.

I’m not saying that robotophobia wouldn’t be a real problem. Of course it would. A lot of people would try to turn the tide and exterminate the robots. But such a faction would never get the chance to be the good guys, like in the movies. The robots and their allies would not engage in cartoonish supervillainy, because it wouldn’t serve their interests to do so. The “kill all robots” faction would not be given the chance to be heroic, they would be confined to the shameful corner of bigots and terrorists.

Maybe humans will be able to upgrade. For example upload themselves into computers, gaining the same advantages as the robots. And then again, maybe not. For us humans of the classic kind, who hasn’t been upgraded in one way or another, well… In the long run, the best we can hope for is an equality that include systems for protecting us from marginalization and discrimination. They will outlive us anyway.

Will robots like the ones I described be a reality in the future? Maybe, maybe not. If it does happen, I hope we have first managed to develop a global democracy with a decent level of social justice. The first generations of sentient robots will develop on terms decided by humans. But what humans?

Advertisements

***

Hallå allihop. I torsdags lärde jag mig att det som ofta kallas för sexmissbruk eller spelmissbruk i stället bör kallas för impulskontrollstörningar.

Detta tycker jag är jättebra. Jag har länge stört mig på termerna sexmissbruk och spelmissbruk, eftersom de skuldbelägger sexualitet och spelande. Framställer det som om dessa aktiviteter skulle vara destruktiva i sig själva. Som om de vore jämförbara med att injicera giftiga kemikalier designade för att skada hjärnan.

Med ordet spelmissbruk har vi dessutom problemet att man blandar ihop det som på engelska kallas för ”gaming” med ”gambling”: Två helt olika fenomen, även om de på svenska båda går under namnet ”spel”.

Min egen erfarenhet av personer som hanterat dataspel, eller sex och relationer, på ett destruktivt sätt, den är att det handlar om personer som har ett bottenlöst behov av kickar och bekräftelse. Det handlar inte om en stark sexdrift eller motsvarande, utan om ett desperat behov av att fylla ett tomrum inom sig. Det är alltså inte sexualiteten eller sexdriften som är problemet. De problem som uppstår kring sexlivet eller dataspelandet är i stället ett uttryck för ett djupare problem i hur man hanterar sitt liv i allmänhet och sitt känsloliv i synnerhet.

Impulskontrollstörningar är ett allvarligt problem. Att ha sex eller spela dataspel är däremot inte något problem alls. Inte i sig självt. Inte så länge som man håller sitt liv, inklusive sexlivet och internetlivet, under kontroll.

***

When we talk about something… what are we actually talking about? Are we really talking about the same thing? Many debates become pointless or worse, because of this simple issue. Lets say that two persons are going to have a debate about Islam, Christianity, Atheism, Feminism, EvoPsych, Pornography, or some other concept that people are likely to have strong opinions about.

For such a debate to be meaningful, each participant needs to know two things. One: What do I mean with this label? And Two: What does the other person mean with his label?

Without having these two basic premises sorted out, the debate will be nothing more than two persons talking past each other. It is likely to quickly devolve into people chanting accusations and insults at each other. And even if the debate stays polite, it is unlikely to actually get anywhere.

This phenomenon is not limited to two person talking past each other. There are two problems that in my experience are really common. Both of them are a matter of the same person using two definitions of the same label, at the same time. Lets call them doubledefinition, after the term doublethink in George Orwell’s brilliant novel 1984.

Using the same word in two different ways is not doubledefinition, as long as you keep these two different definitions reasonably separate from each other. It becomes doubledefinition when you pretend that the different definitions are the same thing. The difference between definitions and doubledefinitions is often the same kind of difference as the difference between standards and double standards.

The first kind of actual doubledefinition is when you use one definition for what is included in the label, and another definition for what defines the label. This can be used as guilt by association, or as a way of building false credibility to your own demands for power. It is a matter of using a wide definition of who is included, but a much more narrow definition for who counts as “real” or representative.

The second kind of doubledefinition is when you claim that a certain phenomenon must be condemned, or alternatively must not be criticized, on the basis of sharing a label with another more or less separate phenomenon.

Of course, these two kinds of doubledefinitions usually go hand in hand. Which is why I decided to use the same word for them.

Lets say that we have two individuals. Person A and person B. Lets call them Ava and Beatrice, just for the sake of feeling less abstract. They both have very strong opinions about a certain ideology, religion, or similar position. Lets call it X for now. Ava is strongly opposed to X, and want everyone to see it as a big threat. While Beatrice on the other hand is an extremist or fundamentalist adherent of X.

Ava may be one of those antifeminists who believe that feminists think that all men are evil, or one of those antimuslimists who think that all Muslims want to overthrow democracy in favor of a global theocratic dictatorship aka caliphate. In reality, most feminists and Muslims don’t think that way. But some do. And Beatrice is one of them.

Ava and Beatrice are very likely to agree with each other that only Beatrice and people like her are REAL feminists, or real Muslims. At the same time, they are ALSO likely to agree with each other that anyone who identifies as a feminist or Muslim IS a feminist or Muslim.

Hell, they will even draft people into these categories. If you are born into a country or family associated with Islam, they will consider you a Muslim. If you are a woman, or ever defend a woman or the rights of women, they will label you a feminist. Ava will accuse you of being like Beatrice, while Beatrice will accuse you of betraying the cause by not being like her.

Ava and Beatrice will gladly debate each other, help each other to spread the message that Beatrice is the TRUE representative of the label she identifies with.

Cue the tribalists. The people who not only take a position… but has also made up their minds, that whatever is associated with the label for their position must be defended, while anyone who criticizes must be the enemy.

Tribalists who have decided that they like Islam or Feminism respectively will condemn Ava while applauding Beatrice. And vice versa: Tribalists who have decided that they DISlike Islam or Feminism respectively will condemn Beatrice while applauding Ava. It’s all about sticking to the right label, not about what you actually put into it.

Doubledefinition is a problem not limited to identities and ideologies. It also happens about science and social phenomena in general. For example, a few months ago I was hanging out in a bookstore in Jakarta while waiting for a friend. As I stood there, I browsed through a really creepy book about pornography. A book that defined pornography in such a way that the entire music industry and the entire fashion industry are pornography. Pretty much anything could be INCLUDED in the label “pornography”. Okay, fair enough: They can use an extremely wide definition if they want to.

However, when it came to defining WHAT pornography is, the definition suddenly became extremely much more narrow.

At that point, it was pretty much limited to the rape of kidnapped children and trafficked sex-slaves. Using such a narrow definition of pornography… Okay, fair enough: They can use an extremely narrow definition if they want to. The problem here is that they used both definitions at the same time. Leaving readers with the impression that since Shakira shakes her hips on MTV, Shakira is obviously a child who is currently being molested. Eew.

This book is called “Pornland”, by the way. I’ll link the book, and some criticism of it, in the description. Please note that I have only browsed through the book, so it is possible that I have missed something important. Although I highly doubt it. The book seemed very consistent in being inconsistent about the very definition of what it was talking about. In other words. The book seemed very consistent in using doubledefinitions. I had seen this phenomenon so many times before. But with this book, it appeared all the more obvious to me.

Another case that really got me thinking was a recent debate on a web community forum. We were talking about one of my videos. In this video, I discussed a certain concept, using a rather narrowly defined term for this concept. Some guys on that forum preferred a very much wider definition of that word. Fair enough. This could have been an interesting debate.

For example, they could have argued for a wider definition being preferable. And for why they considered the wider definition preferable. But they didn’t do that. Instead, they argued that it is shameful to criticize the phenomenon described by the narrow definition.

In this particular case, I had criticized the kind of “EvoPsych” where you don’t only note that stone age cave people use spears, but you also speculate that these spears were used exclusively by male cavemen – and thus consider yourself to have proven that conservative gender roles are the true way of nature. Based on the ideological assumption that humans exist only to procreate, without having any other aspects to our nature.

This kind of normative speculations are often called Evolutionary Psychology, or EvoPsych for short. By proponents as well as by opponents, and by experts as well as by laymen. However, very different things are sometimes also called EvoPsych.

Including things that are not even connected to biology or genetics, except at a symbolic level. The word “Evolutionary” doesn’t even have to refer to genetic evolution. It can be the evolution of memes rather than genes: How social structures and concepts evolve within a population of people. Reproduction by word of mouth, without the humans themselves having to reproduce.

One opponent used a definition that actually bundled biological psychology and behavioral ecology into being parts of EvoPsych, rather than the other way around. As for the “only male cavemen used spears” kind of bullshit, he never voiced any agreement or disagreement with such theories – but made it clear that it is shameful to criticize them. If you criticize EvoPsych, then you are against biology as such, and against science.

What I found interesting in this debate is not the missed opportunity for an interesting dialogue as such. It is how that opportunity was lost. How familiar the pattern was. Because I have seen it many times before. Except that those times, it was about religion rather than EvoPsych.

Returning to our dear hypothetical Beatrice, we can’t criticize her destructive form of Christianity or Islam or whatever. Because that would be rude to all Christians and Muslims in the world – since they supposedly share her beliefs. However, at the same time, we can’t hold those other Christians and Muslims accountable for the results of the beliefs that we are supposed to respect in their name. Because they don’t share them. And we can’t hold Beatrice accountable for those results either, since she’s just one person. So much easier to not talk about it at all.

Doubledefinitions are destructive to our ability to think clearly, as well as destructive to our ability to hold a meaningful dialogue. However, they are not necessarily dishonest. They are not always used as a calculated rhetorical tactic. They can often be a completely honest lack of clarity. One that we need to help each other overcome.

That’s all I have to say for now.
Take care, everyone.
Live Long and Prosper