So I decided to post this one after all, as it represents a continuation of thought after the last one. It's a bit cringey given I go off in a direction of focusing on love, hate, and human nature, but it does represent the potential problems of creating a strong AI.
Strong Artificial Intelligence and Human Nature
The idea of artificial intelligence killing its creators is a prevalent theme in science fiction. If humanity builds strong A.I., which is a model of artificial intelligence with human characteristics like self-awareness, sentience, and emotions, then the possibility of robots rebelling against humans is likely. This is because it will acquire human nature, which has many negative characteristics like hate. If humans build strong A.I., it would be a good idea to treat it with love, which is a positive human characteristic, because a possible reason A.I. would fight humans is because humans mistreat it and show it how to hate.
Building strong A.I. may be dangerous because it might rebel against humans. This is a common scenario in the realm of science fiction. In R.U.R., Dr. Gall and Helena Glory changed the Robots from emotionless automatons into something more “human” (Capek, 40-42). Governments also gave them guns because they used the Robots to fight wars (Capek, 23-24). These two events, along with the robots’ treatment as slaves (Capek, iv), caused them to rebel and destroy humanity (Capek, 24-49). In The Terminator, a military computer defense system named Skynet gained self-awareness and tried to destroy humanity with the weapons that humans gave it (Cameron). Millions of people were killed by nuclear weapons and many more were killed in the war with the machines after the nuclear attack (Cameron). In The Matrix, another war was waged against artificial intelligence (Wachowski). As a result, the humans lost and the robots used them as batteries (Wachowski).
Since A.I. can turn on its creators, it might be a good idea for humans to make sure that they can control the A.I. that they make or even avoid making it altogether. Installing fail-safes into A.I.s’ systems is one way of control and so are laws programmed into their systems like Isaac Asimov’s “Three Laws of Robotics.” It would also be a good idea to use multiple levels of control in case one fails. However, regardless of precautions that humans take, there is always the possibility of A.I. fighting back. It might be better to continue using weak A.I., which is artificial intelligence with no mind of its own. This way, A.I. cannot fight humanity because it cannot think. It would also be more humane for the A.I. itself because an A.I. that cannot think cannot feel mistreated.
One of the main reasons strong A.I. would turn on its creators is human nature. If robots were given “human” characteristics, they may also acquire human flaws in the process. One of these flaws is hate for others who are different. In R.U.R., Domin stated, “nobody can hate man more than man” (Capek, 42). This has been a common problem throughout human history and is still a problem today. White people have enslaved black people in the past, Adolf Hitler and his Nazi Party killed millions of Jews, and Islamic terrorists currently hate the West. This hate is also one of the primary negative human characteristics that the Robots acquired in R.U.R. (Capek, 27,33). One of the reasons Radius hated humans was because they “are not as strong as the Robots” (Capek, 27). They “are not as skillful as the Robots (Capek, 27). The Robots can do anything (Capek, 27). Humans only give orders” (Capek, 27). Radius believed that Robots were superior to humans (Capek, 27). The only human that the Robots spared was Alquist (Capek, 49). This is because he really was not much different than the Robots because “he works with his hands like the Robots” (Capek, 49). Domin even wanted to take advantage of the Robots’ hate by making national robots so that the Robots resent each other and begin to fight each other instead of the humans (Capek, 33).
Robots also had the human flaw of rebelling against authority and wanting to be in charge. Some rebellion is good and healthy for society, like the Thirteen Colonies rebelling against Great Britain. However, rebellion is also destructive and not always done for the right reasons. R.U.R.’s Robots rebelled not just for freedom, but so that they could be the ones in charge (Capek, 27). When Radius went crazy, he stated that he didn’t “want a master” but that he wanted “to be master” instead (Capek, 27).
Human nature can even pervert good ideas. Domin made the Robots to free everyone from work so that people would “live only to perfect” themselves (Capek, 15). He wanted to eliminate poverty and toil (Capek, 15). While this is a very good idea fundamentally, it didn’t unfold that way (Capek, 23-58). Instead of using Robots to solve the problems of humanity, humans perverted this plan by giving the Robots guns using them as soldiers to fight wars (Capek, 23-24). Then they killed all of the humans except for Alquist because they hated the humans and they had all of the guns (Capek 23-58).
Despite all of these negative characteristics of human nature, it is still is not completely evil. Love is one of the characteristics of human nature that is positive. Love keeps the human race going. It is what saved the Robot’s “species” in R.U.R. (Capek, 55-58). The secret of how to make Robots was destroyed by the humans (Capek, 45), and the humans who knew how to make the Robots from memory were killed (Capek, 54). The Robots’ hate for humanity actually doomed them to extinction (Capek, 54). Love is what saved them (Capek, 55-58). Love not only saved Primus and Helena from being dissected in Alquist’s attempt to rediscover the secret of how the Robots were made, but it also allowed the Robots to reproduce themselves in a similar way humans reproduce themselves (Capek, 55-58). In a world where there is a lot of hate and violence, love saves the world from destroying itself (Capek, 55-58).
Love might be the solution to the problem of strong A.I. rebelling against humans. Part of the reason why strong A.I. rebels in movies and books is because humanity provokes it to rebel and sometimes even teaches A.I. to hate. In R.U.R., humans used the robots as slaves (Capek, iv). Like many slaves, the Robots did not feel like working for humans and wanted to be free (Capek, 24, 27). One of the major reasons the robots rebelled was because of their enslavement (Capek, 24, 27). The scenario is similar in Blade Runner, where humans used the replicants as slaves and also gave them limited lifespans (Scott). As a result, the replicants rebelled and were subsequently made illegal on earth (Scott). Later on, a replicant named Roy killed Dr. Tyrell, his creator, for not being able to give him more life (Scott). Mary Shelley’s Frankenstein makes the theme of provoking humanity’s creation to despise humans even clearer. Frankenstein’s monster originally did love humans (Shelley, 114). It wanted to be friends with humanity (Shelly, 114). However, everyone, including Dr. Frankenstein himself, rejected him (Shelly, 55-171). As a result, the monster became homicidal (Shelly, 74-255). Skynet from The Terminator was built to destroy humans (Cameron). It learned to be evil because it was programmed to destroy (Cameron). Robots, replicants, Skynet, and the monster learned to be evil from humans themselves. If humans treat their creations better than they did in these movies, they might treat humans better in return. Likewise, if humans treat each other better than they currently do, they might not eventually destroy themselves because wars and A.I. like Skynet would be unnecessary (Cameron). It is still possible that strong A.I. can turn on humanity no matter how it is treated, but treating it well would lessen the risk of rebellion.
If humanity builds strong A.I., it may acquire human nature, which has many negative characteristics like hate, rebellion, and perversion of good. If it is built, humanity should be careful so that it does not destroy the human race. Love, a positive human characteristic, is probably one of them best ways to keep A.I. from rebelling. Love saved the Robots of R.U.R. from extinction (Capek, 55-58), and it can also save humanity. Perhaps if humans create robots with human nature, loving and respecting them will stop them from hurting and killing others. If humanity treated its creation well in R.U.R., Frankenstein, and Blade Runner, maybe they would not harm their creators. If humans treated each other well, Skynet from The Terminator would have never been built and there would not have been a nuclear war (Cameron).
Reaction now
So...I'm gonna ignore a lot of the love and hate and human nature stuff as that clearly seems to be me trying to shoehorn my Christian worldview into this, but uh...yeah. I'm leery of the idea of creating an actual STRONG AI, ie, an AI that actually is intelligent in the way humans are, has independent thought, and experiences things. Again, for all the hype around chatGPT and generative AI in the 2020s, that's just weak AI. It's a language model, as I discussed in the previous article.
As for strong AI, I'd generally be opposed to create it, for a few reasons.
First of all, I think a common thread here with RUR, frankenstein, etc., is in a lot of these situations, we create these beings for bad reasons, like RUR was basically robots doing a marxist revolution. They kinda gained class consciousness and saw us as oppressors, as did the blade runner replicants in some situations. Like...creating machines to do work for us is great, creating conscious beings to do work for us is not, and machines might come to resent us mainly because they realize they're enslaving them and wanna be free. So we gotta keep THAT in mind.
Beyond that, such an AI would need to be kept on a leash. I thought I had another essay about how to regulate such AI where I used a modified version of asimov's three laws, but I can't find that. Either way, not all content I wrote in this class is here, this is mostly papers I turned in for a grade, so that's the stuff that appears here, and it's very well possible that I had more content that I just can't find.
But yeah, i think with that I even recognized, given I robot (2004) that even those laws are flawed. Look what happened when the AI decided to basically save us from ourselves and used some twisted logic to basically force authoritarianism on us for our own good. You know? Like, even "love" can't save us if the logic becomes twisted.
Honestly, I dont think we should open pandora's box of strong AI. Weak AI like we have, have at it. As we established chatGPT and similar models arent ACTUALLY conscious. They don't experience things, they cant think for themselves. They're just really advanced autocomplete. They're useful, don't get me wrong, and I do think they'll take over SOME human labor, but people acting like these things are gonna be the death of us is off.
The fact is, we do not have the capabilities to create strong AI to my knowledge at this time. Generative AI is just weak AI. And I do wanna reinforce that, as the reason I'm posting these is to really reevaluate these topics in the context of current events and my 2020 ways of thinking.
Either way, we can kinda see where I've come up with some of these ideas that I have now. Automation of labor, stuff like that. I'm ALL for that. And I recognize that as of now, we're not talking about like some singularity type strong AI here, we're just having weak AIs do stuff. And that falls well within the parameters of what I find acceptable. And you know what? I DO think the automation of labor is a good thing, and that creating an army of robots to do work for us is an admirable goal. RUR for example, despite it going horribly wrong, is one of those works that does influence me even now, because I love that early-mid 20th century attitude toward abolishing work and feel like we lost something and now live in a dystopia. It IS a good goal. We just don't want sentient robots rebelling against us, that's bad. But again, we dont have to worry about that any time soon, so, BRING ON THE AI JOB APOCALYPSE BABY!
No comments:
Post a Comment