SHRDLU on Film

The glory days of Artificial Intelligence — those heady days when researchers thought genuine intelligence and perhaps even consciousness would transcend the bonds of human brains and sit comfortably on computer hardware — well, it was a fun time to be alive. And check out this gem:

http://www.csail.mit.edu/videoarchive/history/aifilms

An awesome historical collection of AI videos housed at MIT.

This one’s my favorite:

 

A silent, grainy, low-res film of Terry Winograd’s awesome SHRDLU in action. For those of you youngsters out there, SHRDLU was really quite a remarkable achievement in computer programming. And the fact that Winograd’s first release of the software was in 1968 makes it all the more remarkable. The idea was to create a program that could parse real English sentences and respond intelligently to them. Winograd realized that he couldn’t tackle the problem in an unlimited domain (that is, his program wasn’t going to be able to talk about everything a human might be able to), so he limited the program’s world to a virtual table top with some blocks on it, and a virtual robotic arm that could pick up and move those blocks. Within this tiny domain, SHRDLU was really very impressive. Frankly, I’m still impressed by the program today, even though it’s more than 40 years old.

Here’s a sample excerpt from a human-SHRDLU interaction:

Person: WHAT DID THE RED CUBE SUPPORT BEFORE YOU STARTED TO CLEAN IT OFF?
Computer: THE GREEN PYRAMID.
Person: THERE WERE FIVE BLOCKS TO THE LEFT OF THE BOX THEN.
Computer: NO, ONLY FOUR OF THEM: THE RED CUBE, TWO LARGE GREEN CUBES AND A LARGE RED BLOCK.
Person: PUT THE BLUE PYRAMID ON THE BLOCK IN THE BOX.
Computer: OK. (does it)
Person: IS THERE ANYTHING WHICH IS BIGGER THAN EVERY PYRAMID BUT IS NOT AS WIDE AS THE THING THAT SUPPORTS IT?
Computer: YES, THE BLUE BLOCK.
Person: DOES A STEEPLE
Computer interrupting: SORRY, I DON’T KNOW THE WORD “STEEPLE”.
Person: A “STEEPLE” IS A STACK WHICH CONTAINS TWO GREEN CUBES AND A PYRAMID.
Computer: I UNDERSTAND.
Person: ARE THERE ANY STEEPLES NOW?
Computer: NO.
Person: BUILD ONE.
Computer: OK. (does it)
Person: CALL THE BIGGEST BLOCK “SUPERBLOCK”.
Computer: OK.
Person: HAVE YOU PICKED SUPERBLOCK UP SINCE WE BEGAN?
Computer: YES.
Person: WHY DID YOU DROP IT?
Computer: BY “IT”, I ASSUME YOU MEAN THE SUPERBLOCK.

Really impressive language processing.

AI researchers (and philosophers) at the time were a little too impressed with SHRDLU, and thought that genuinely intelligent computers were surely close at hand. In fact, Winograd himself, after wrestling for years with SHRDLU and computational language processing, came to the conclusion that AI researchers were generally far too optimistic about their achievements, at least as regarded the actual supposed intelligence of their creations. Winograd wrote:

“Most current computational models of cognition are vastly underconstrained and ad hoc; they are contrivances assembled to mimic arbitrary pieces of behavior, with insufficient concern for explicating the principles in virtue of which such behavior is exhibited and with little regard for a precise understanding.” [Winograd, 1987]

This was a cold shower for a lot of optimistic AI researchers, and a bolster to philosophers who opposed functionalism. But this is a topic for another post.

Happy AI film viewing!

Quandary Corner: Baby Hitler

Quandary Corner: In this new topic section, Alec and I will each discuss a particular ethical quandary and how we would deal with it, based on whatever ethical theory we take to be applicable at that time or in that situation. Will there be consistency across quandaries? Hahahahahahaha! Good question.

Today’s ethical quandary is an oldie but a goodie: A scientist has created a time machine that can take you back in time to when Adolf Hitler was a baby (let’s say at five months old). You will be sent to his nursery at a time that baby Hitler is alone. The nature of the time machine is such that you are only there for a maximum of two minutes. You cannot bring baby Hitler back to the future with you. You can, however, kill him. Ignoring the question of whether it is possible to change the past or not, should you kill Hitler when he was a baby or not?


Alec

My students, following my example, don’t take long before offering up “Hitler and the Nazis” whenever I ask for an example of great moral evil. Indeed, if there were a prohibition on mentioning Hitler, I’m not quite sure what I’d do, pedagogically. I suppose I’d have to resort to fictional super-villains. (Maybe I should anyway. It’s not as depressing to talk about Lex Luther.)

So, faced with the opportunity to wipe out the source of perhaps the greatest evil perpetrated in the last couple of centuries, what would I do?

Utilitarianism

My initial thought is: Of course I would kill baby Hitler. But there are some mitigating ideas that cross my mind soon after.

For one thing, Hitler was not alone in carrying out the Holocaust, and perhaps it would have happened regardless of his existence. It could well be that social forces were pointing Germany in that direction, and that another figurehead would’ve stepped into Hitler’s place, if Hitler hadn’t been available. If we’ve learned anything from time travel movies it’s that the whole thing is very unpredictable. But let’s put this worry aside for the time being, and address my feeling that the baby Hitler should be killed.

I imagine that this is the popular intuition here. But what is the basis for this intuition? Well, I can see two probable bases: utilitarianism, and righteous vengeance.

Utilitarianism is the philosophical stance that ethical decisions should be made strictly on the basis of weighing the possible good and bad outcomes of your actions. Famously, utilitarianism says that, other things being equal, ten lives are more valuable than one life, and so an action that kills one person in order to save ten is ethically justified. (We’ll save the finer points of this theory for another post. Utilitarianism is not, as you might imagine, free from problems.) Well, if you can save ten lives by killing one, just imagine the scenario where you can save millions by killing one. That’s the scenario we’ve been tasked to analyze here. By killing the baby Hitler, we are ostensibly saving millions of lives that will eventually die at his command.

Virtue Ethics

Another thing that makes me pause, when thinking about killing baby Hitler, is that I’m tremendously squeamish about death, and the idea of killing a baby (even an evil one) makes me blanch.

Gregory Peck, confronting a moral dilemma.

Gregory Peck, confronting a moral dilemma.

Now, psychological squeamishness, one could argue, has no place in ethics. If something is the right thing to do, and I don’t want to do it, I’m just wrong about that, despite whatever my superego is telling me.

But there is an aspect of this squeamishness that is actually philosophically relevant, and it revolves around the issue of virtue. Most theories of ethics are action-based. Utilitarianism, for instance, is supposed to tell you what to do in any given situation. Virtue ethics (Aristotle is most often credited as its founder) is based more on the idea of developing a good character — the idea is that if you are a generally virtuous person, you’ll generally make the right decisions when faced with moral dilemmas.

Well, it is reasonable to argue that it’s the sign of a virtuous character to be squeamish about killing a baby. And so perhaps killing the baby Hitler is the wrong thing to do, if we are to take virtue ethics seriously. Of course, it’s also the sign of a virtuous character to save millions of lives if you can, and so virtue ethics sends us mixed signals on this one.

The Right Thing To Do

In the end, I think that killing the baby Hitler is the right thing to do. And if you try to justify sparing his life, on the basis of virtuous squeamishness, you’re probably displaying the unvirtuous trait of cowardice.

There are some philosophers who argue that utilitarianism goes wrong in cases of human death, because it’s just wrong (and impossible) to weigh a life. But I say there’s something wrong with a theory of ethics that tells us we can’t weigh one life against millions.


Jim

Alec makes some good points in his write-up of the issue. I think that baby Hitler should also be killed, but I am not sure that it is for the same reasons. Let’s see.

What should we do if we have the chance to save a life, and we can do so at little to no peril to our own life? We should save the life. Few people would argue that point. Well, I might argue that point. I am not entirely convinced that life, least of all human life, is always worth saving, and that is not even based on the character or worth of the (human) life. That is a matter for another post though, perhaps. Here, I am content to go with the status quo and agree that life should be preserved.

If you see a person about to cross the street, but he does not see an oncoming car, you should alert the person to the danger, grabbing him if necessary (and, again, if it does not immediately imperil your own life). Should we save life if the only way that is available to us is to end the life of another? Whew. Good question. Do we have time left for this? We do? Dammit.

Superbad

Hitler was a bad, bad, superbad, person. If anyone ever deserved to die, it was Hitler. Perhaps, as Alec suggested above, Hitler was not the sole person responsible for all the evil attributed to his movement, but he was close enough to the sole person. Would most of it still have happened had Hitler not have been? While that is an interesting question, I don’t see that it matters much here. Baby Hitler is, as a baby (philosopher-speak: qua baby), is not evil, has not committed evil, and does not, in any modern understanding, exhibit evil tendencies or character. We are going to kill baby Hitler for what adult Hitler will bring about.

This is an interesting sense of justice: we are trying to balance a wrong that has yet to happen, but will certainly happen. The ‘balancing’ act however, is such that it will ensure the evil never occur. If that is so, then our act is unjust. If we do not do this act though, then great evil will result and that seems to make not committing the act unjust as well. There is a true dilemma here, and each horn is going to do whatever you think would make for the least possible cliche here.

This is not a true dilemma of the ordinary sort as we know what will happen if we do not act. We are 100% positive (Mel Gibson’s father aside) about what will occur if Hitler is not killed as a baby. Hence, I suggest that all the time travel does for this scenario is modify our verbs in a way that bothers us. In a justice sense, baby Hitler has to die for what we know adult Hitler will do.

Squeamishness, Utility, and the Right Thing

As for the squeamish factor that Alec notes above, I think I agree, but only as an interesting artifact. Can we trust squeamishness as a guide to morality? No, but nor does Alec think so. It is a suggestion or a clue at best.

Would I feel squeamish about killing baby Hitler? I would like to say no, because of the greater good I would be serving, but that would be a lie. I would be a little squeamish; he is a baby after all. Were I to look at his little mustache and think for even a moment about the evil commands that would march out, albeit much later, from beneath it, however, I would full on vomit with squeamishness were I not to kill him.

Lives will certainly be saved by killing baby Hitler and an enormous evil will be excised from the world. Will some other evil fill that void? More than likely, as the world seemingly sucks. Will that substitute evil be more evil than Hitler? Who can say? Baby Hitler should be killed because more lives will be saved, including infant lives, than will be lost by the singular act of killing baby Hitler, and this we can be remarkably certain of given our futuristic knowledge.

And if it turns out we were wrong, surely we would appear to ourselves in the past to stop us before it is too late. Right?

Philosophy in the Movies: Dr. Orlof

This is a new feature here at We Love Philosophy where Alec and I will look at some piece of film or literature or whatever, a piece that few if any would consider to have philosophical themes or import, and try to find philosophical themes or import within. This is not an easy task and may be better abandoned than pursued, but, as my old P.E. teacher used to tell me: “Go ahead and try Thomas, just so you will know I was right about your failing.” Let’s try. We will start with a brief synopsis of the work selected and then the picking and stretching will begin.

For the inaugural posting within this topic we chose a topic sure to kill this topic nigh-immediately: the film The Awful Dr. Orlof. Alec chose this. I agreed with the choice, but remember that this was Alec’s remarkably ugly baby.

Dr. Orlof and his blind sidekick.

Dr. Orlof and his blind sidekick.

[Note from Alec: I’m really terribly sorry. I recommend that no one watch this movie. Ever. If you see it playing one night on Turner Classic, throw your TV out. And move.]

Since this movie is on Netflix and came out in 1962, and we’ve warned you not to watch it, we are going to ignore any worries about spoilers and just lay the whole intricate plot out. Dr. Orlof is kidnapping and then killing young girls in order to cure his daughter. A few specific points should be made: Dr. Orlof himself does little of the killing, but has a blind ‘special’ assistant, Morpho, do so; the cure seems to involve skinning the women, or at least some body parts of the women; it is not entirely clear what is wrong with his daughter aside from a kind of catatonic listlessness and mild scarring; the entire police force is hopeless and remarkably uncaring about the people they have sworn to protect.

While Dr. Orlof is having girls taken and killed by Morpho, a police detective is trying to find the killer — this after four women have been reported missing. The police department really kicks into action when a man is found murdered though. I suppose one could make an argument that the movie is sexist here, but, to be fair, they found the dead body of the man, and the bodies of the women are still just missing. The police detective’s fiancee gets involved in the investigation for reasons that are never clear, but not much of the movie is, so let’s not hold that against her. She, the fiancee, is kidnapped by the doctor who, I think, needs her breasts to complete the treatment of his daughter (the contents of such a thought are uber-creepy). Dr. Orlof is eventually killed by the monster he inspired to kill for him, and that monster dies as well.

Whither Whence the Philosophy Cometh?

Oof. So, now that the summary is out of the way, whither whence the philosophy cometh? Some of these entries are going to require more stretching than others. This first entry is beyond Stretch Armstrong levels and approaching Mr. Fantastic/Plastic-Man levels. Here goes: we are going to divide this into two sections. Ethics and Epistemology. I will take the ethics section, for reasons that have yet to be explained, and Alec will take epistemology. The ethics topic of the movie is fairly straightforward, unlike everything else in the movie: is the life of one person dear to you (here, Dr. Orlof’s daughter) worth more than the lives of (at least) five strangers? And if the answer is yes, is it morally allowable to have an irrational person do the killing for you? For that last question, there is a sub-question dealing with who is to blame for the killings. Morpho did not seem to be the brightest bulb in the chandelier, but he was remarkably capable of tracking and killing, despite his blindness (which I think was almost entirely due to the papier mache mask he was wearing).

The epistemology part is a bit of a stretch, but that is what we are here for. Epistemology is the philosophical study of knowledge — what makes something knowledge and what justifies our beliefs. There are a couple of questions about the movie that fit that description: 1) As no one had ever tried Orlof’s methods before, what justified his belief in them (assuming that he was justified at all); 2) How did the detective’s fiancee ‘know’ that she had seen the killer? Is women’s intuition a reliable font of knowledge?

The Ethics of Orlof

Dr. Orlof is clearly not a utilitarian. At least, based on what we are given by the plot, Dr. Orloff is not a Utilitarian. His acts are all aimed at making his daughter better/cured (as it is never clear what her condition is beyond a sort of laissez faire attitude about life combined with a bad make-up job), and this is never suggested as being anything other than purely selfish on his part. She, Orlof-fils, is never portrayed as a brilliant scientist, or a dedicated humanitarian, or a inspirational person. Rather, she is the doctor’s daughter and so she must be fixed.

Utilitarianism is the ethic that the right action is the one that creates the most pleasure and the least pain for the greatest number of people. Ergo, Orlof is not a utilitarian.

He is selfish to a degree that many of us would find startling were he a real person (Note to Alec: this was not based on true story, was it?). We all love those in our family. Well, many of us love at least someone in our family. Within that group are many who say, truthfully or so they believe, that they would do anything to save or preserve the life of their loved one. Perhaps that includes even dying, if doing so would certainly, or damn near certainly, save or preserve the life of the loved one. Such an attitude is not entirely selfish, of course.

Dr. Orlof, however, not only loves his daughter, and not only loves his daughter more than he loves anyone else, but he loves her to the point that other lives are meaningless when compared to her own. The lives of other human beings are merely tools for Orloff to use to better aid his daughter. “What I want,” Dr. Orlof seems to be saying through his actions, if not through well-written dialog, “is what determines what is right for me and so for the world. If it is not what you want, whoever you might be, then I will sic my blind murderer on you as well.”

This is a kind of hedonism, the idea that pleasure is the most important goal, but of a non-universal character. It is one thing to say or argue that the goal in life is to be as happy as possible. That can be universal in the sense that it is what you think is so for everyone. “Hey,” you might say to me, “why are you sitting at that desk when you could be having fun?” Good question.

Alright, I am back. Thanks for suggesting that I take a break. See? A hedonist need not be a selfish prick, at least a universal hedonist need not be. Orlof, though? He’s a selfish prick. His pleasure is what matter the most to him. One could argue that he does what he does (or gives the orders to kill that he gives) because he is thinking only of his daughter, a kind of pass-the-buck hedonism.

“Your pleasure is what matters the most, such a person would say to you, “so tell me what would make you happy and I will do it.” That almost sounds like altruism: the sacrificing of one’s happiness so that another is made happy in your stead. Ignoring whether or not altruism truly exists or is rational (hey — future topics!), is that what Orlof is guilty of? I doubt it. His daughter never says or indicates that she approves of what he is doing; in fact, there is no indication at all that she is receiving any benefit at all from the murders and procedures her father is performing. Perhaps Dr. Orlof is telling himself that what he is doing is for his daughter alone, but all the visual evidence (that I shudder to recall) indicates he is doing it for himself.

So, is selfish hedonism a valid ethical standpoint? Sure, I suppose. It is as valid as any other ethical standpoint, so long as you do not take as a given that all or even most or at best some people are equal to yourself in terms what it takes to be happy. Is that what you think? Are you an Orlofian? Are you???

The Epistemology of Orlof

How did I get stuck with the epistemology piece of this? Sigh. Oh well, here goes what might literally be nothing.

First things first, let’s dive into the women’s intuition thing. So, in the movie, the detective’s fiancee happens to see the killer (without knowing that he’s the killer), and somehow is immediately convinced that she has just seen the killer.

Is she justified in this belief?

You know what? Let’s dive right back out of this women’s intuition thing. There’s nothing good that can come of the discussion, is there?

What’s more interesting here is the sentence “Is she justified in this belief?” This contains two of the three components to what might be considered having genuine knowledge of something.

A Standard Conception of Knowledge

When can we be said to know something?

Obviously, belief has to factor into it.

Belief

If Dr. Orlof is the murderer, but I don’t believe that he is, then I really can’t have knowledge that he is the murderer. So belief is certainly necessary for knowledge, though it’s also not sufficient.

Suppose Superman is the murderer. If I believe that Dr. Orlof is the murderer, I still don’t have any knowledge that Dr. Orlof is the murderer. (My belief is false.)

So we need to account for truth — circumstances that match our beliefs.

Truth

If I believe that Dr. Orlof is the murderer, and he is in fact the murderer, then can I say I have knowledge of him being the murderer?

Well, though truth is definitely necessary, it’s not sufficient.

Suppose Dr. Orlof is the murderer. But I believe he’s the murderer only because I dreamed about him being the murderer the night before. Do I have knowledge of Dr. Orlof being the murderer in this case? Nope. In effect, I’ve just made a lucky guess, and this shouldn’t count as knowledge.

So we need to account for the idea that knowledge should also be justified.

Justification

If I don’t know that Dr. Orlof is the murderer, but my wife (who never lies to me) tells me she saw Dr. Orlof murdering the victims, do I have knowledge yet?

Justified true belief. A pretty darn good working definition of knowledge. (Though we’ll talk about a famous counterexample to this definition in a future post.)

So does the detective’s wife have knowledge that Dr. Orlof is the murderer? Well, she has the belief that he is, and it’s true that he is, and so the question hinges on whether or not her women’s intuition can form the basis of a justified belief. I won’t take a stand on that issue here. Feel free to take a stand in the comments…