What is ‘evil’?
Philosophers and theologians have asked this question for millennia. Augustine posited that if God is the creator of all things, then evil cannot be a “thing” as such, because that would implicate God as the author of evil. For Augustine then, rather than evil being a polar opposite substance than good, evil is instead the complete absence of good. In the same way that darkness is nothing in itself but an absence of light, evil is nihil—ontological nothingness.
Scientific American has a piece looking at philosopher and cognitive scientist Selmer Bringsjord, who is working with a team at Rensselaer Polytechnic Institute to develop an artificial intelligence—a computerized person—that represents pure evil.
Bringsjord says, “I’ve been working on what is evil and how to formally define it.” From the article:
To be truly evil, someone must have sought to do harm by planning to commit some morally wrong action with no prompting from others (whether this person successfully executes his or her plan is beside the point). The evil person must have tried to carry out this plan with the hope of “causing considerable harm to others,” Bringsjord says. Finally, “and most importantly,” he adds, if this evil person were willing to analyze his or her reasons for wanting to commit this morally wrong action, these reasons would either prove to be incoherent, or they would reveal that the evil person knew he or she was doing something wrong and regarded the harm caused as a good thing.
“I thought it would be interesting to come up with formal structures that define evil,” Bringsjord says, “and, ultimately, to create a purely evil character the way a creative writer would.”
The Author of Evil: Selmer Bringsjord
This research has created E: the digital representation of quintessential malevolence. By asking a series of questions to E, we learn that it describes itself as a young, white man, with dark hair and stubble on his face.
In one twisted study, E is asked how it responds in a situation where his “parents” have given him a gun that his “brother” used to commit suicide. “He” is programmed to believe that in fact he gave the gun to his brother, rather than his parents. Then they ask him why.
But of course, any exploration into the creation and manipulation of evil raises all of the ethical questions: Should we create evil? Should we make artificial intelligence at all? What is evil? How would such “evil” be controlled, if allowed to interact in a virtual world with other “avatars” controlled by humans? Bringsjord addresses only this final question:
“I wouldn’t release E or anything like it, even in purely virtual environments, without engineered safeguards,” Bringsjord says. These safeguards would be a set of ethics written into the software, something akin to author Isaac Asimov’s “Three Laws of Robotics,” that [1] prevent a robot from harming humans, [2] requires a robot to obey humans, and [3] instructs a robot to protect itself—as long as that does not violate either or both of the first two laws.
“Because I have a lot of faith in this approach,” he says, “E will be controlled.”
But the greatest blessing and curse of artificial intelligence is that it is intelligent; and as such, it can learn and adapt. Even as Asimov wrote his ethical 3 laws, he foresaw the implications they could lead to by a genuinely independent intelligence. What if to protect humans, some harm had to be inflicted on them? What if to protect humanity, humans had to be protected from themselves? Even within engineered safeguards, Asimov explored how an artificial intelligence could evolve its understanding of such boundaries and operate within them in unforeseen ways. (An entertaining introduction to this is the 2004 Will Smith film based on Asimov’s writings, I, Robot.)
An interesting field that has spawned as a result of the research into A.I. and robotics is the subfield of robo-ethics. Should robots powered by artificial intelligence be engineered to emulate humans? Should robotic intelligence exhibit “feelings,” even if only simulated? Should robots have a face?
With the evolution of humanity into a new technological age, humans and machines coexist in a symbiotic relationship. The lines between reality and virtual reality are increasingly blurred: electricity to homes and cities is controlled by digital power grids and computer networks. Conversations happen both face-to-face, and over Face-book. Some people have flesh-and-blood hearts pumping in their chest, others have artificial mechanical hearts pumping in theirs.
Not only does this raise new ethical questions, it also asks afresh all of the old: What does it mean to be “human”? What does it mean to have a will? Is the will free? And back to the original point: should we be creating evil “wills” in this kind of environment?
Bringsford says he has “a lot of faith” in his ability to create and control such an evil. Is his faith misplaced? Time will tell.
[...] Read the rest of my synopsis and thoughts on the experiment at the collaborative blog, dust and light. [...]
Under Augustine’s usage, isn’t it impossible to “create” evil … like, if evil is an absence or a privation of good, if it has no being in itself, it can’t strictly be “created” because it’s literally nothing?
Wouldn’t you then, if you wanted to make something which was pure evil, need to make something with a complete inability to do good, or to love, or show compassion or whatever, rather than trying to think up some positive characteristic which defines evil? If so, evidently J.K. Rowling was kind of right with her portrayal of Voldemort, I guess. Or, if she wasn’t ‘right’, she was at least Augustinian!
“if evil is an absence or a privation of good, if it has no being in itself, it can’t strictly be “created” because it’s literally nothing?”
Bingo.
“If so, evidently J.K. Rowling was kind of right with her portrayal of Voldemort, I guess. Or, if she wasn’t ‘right’, she was at least Augustinian!”
Definitely Augustinian. But now it begs the question—is she right?
Or more specifically, is Augustine right?
According to Cardinal Milbank, thou shalt be excommunicated for asking such questions…
But yeah, I like Augustine’s view best.
Interesting post for sure. I will be the first to admit that I’m philosophically and theologically stupid, or at least ignorant, so I won’t be making any statments regarding the Augustinian-ness of J.K. Rowling or Mr. (or presumbly Dr.) Bringsjord.
I’m going to make my argument frome the other direction. I believe it is impossible to create A.I. that does pure righteousness. Any kind of programming of the A.I. done would be done by humans,and we, of course, are falliable. Therefore it would be impossible to create something greater than we are. In addition, if we were to create the “Perfectly Righteous ‘Bot” that would undermine everything God has told us He is, and also what Christ did for us on the Cross. If we can achieve perfection by supplanting our brains with a “righteous chip”, why do we need Jesus?
Inversely, I believe the exact opposite is true as well. One can certainly design a ‘bot that is “evil”, but purely evil? No. Just as man can’t design, let alone concieve, an entity that is purely good, the same can be said about evil.
Interesting thoughts, Kyle! It’s always interesting to tackle a problem from the opposite scenario. And of course, it then begs the opposite question: what is pure righteousness? And I think that you provided a good answer.
Even if Bringsjord doesn’t make “pure” evil (and even if it’s not Augustinian), it’s still pretty twisted to “create” an “evil” “person.” But strangely, I do think it is actually worth doing, for the interesting research that could be done with it. But it’s still pretty sticky stuff!
The original Scientific American article references the virtual world “Second Life.” It’s an online community but you interact in a 3D world and create a complete avatar-character representative of you, etc. It certainly would be interesting to see what this A.I. would do in that environment
.
Awesome topic! I’ve toyed with the idea of the prototypical “evil being” as the form of “tempter” or “accuser”.
Imagine, for example, a woman who seduces men and attempts to get them to break apart their families, expend all of their money, and eventually commit suicide. She’s very good at this, achieving a 90% suicide rate among 10 men that she seduces. Only one man resists the snares, and this particular man becomes a much better person for having gone through the trials.
This woman could defend herself by explaining, “The other 9 guys were spiritually diseased pieces of garbage, and I sent them to their judgment. The 10th guy was proven righteous only by my test, so he has me to thank for his just happiness.”
Of course, neither we nor God would accept such a defense. We would condemn the woman.
Therefore, I think this would be the essential algorithm for an “evil being”. To the best of your knowledge (and perfect knowledge is not necessary — only intent matters), do all within your power to trip up other actors and cause them to do whatever they know to be wrong.
Very interesting, Josh. I think that would be very similar to the definition and model that Bringsford (from the article above) is operating on. Especially the bit about “intentions,” which is very important—regardless of success or failure
.
Simon, I agree with Milbank. Augustine is NEVER wrong.
By the way, are you enjoying your time at Nottingham? That is one of my choices for a M.A.