Study of ourselves

It is often said that neuroscience is the study of the mind. While there is no denying that neuroscientists do study the brain and how it works, I say that Artificial Intelligence is just as much as about studying how humans work as it is knowing how the human brain works. Brian Christian spends a good part of his book talking about how the human mind works and how he studied this in preparation for the Loebner prize. He wanted to know how to behave as human as possible (and what that even meant) in the same way that the programmers who competed wanted to know the same thing to model it in their submission. This could arguable be more important to the Loebner prize that an understanding of programming. While there would be no program without the programming knowledge, the computer would have no chance of winning if the programmer did not have an understanding of human behavior and how to model it. One of the main ways to make advances in Artificial Intelligence is to have a better understanding (and therefore, become better able model) of human behavior and human tendencies.

Compare the Loebner prize with the Deep Blue chess competitions. Deep Blue was programmed to play chess by the programmers studying the theory behind how chess plays out and the patterns their opponent had used in past matches. The programmers had to work to model the logic that goes into chess matches and all of the theory that the masters use when they are preparing for matches. The programmers for the Loebner do the same thing, except they study a more complicated set of the logic, the logic that dictates the basics of human behavior and conversation. By trying to replicate these patterns and logics, we are essentially trying to nail down the why behind what we do. Successfully creating a program that models human behavior will not only create another “species” of equal intelligence, but it will actually be some sort of mirror into our own nature. This prospect may be even more frightening then the idea of creating what may one day destroy us, but that it will be our own concept of us. Our own reflection will bring about our downfall.

A conversation with Cleverbot

[Sorry this post takes up so much room. I had to put the chat in.]
So today, for my blogpost, I'll be writing about a conversation that I had with the AI bot, Cleverbot. I had high hopes going into this conversation. After reading The Most Human Human, I was all set on trying to prove to myself that Cleverbot was, indeed, a computer program and not actually a human. Up until the start of the conversation, I was almost 100% certain that Cleverbot was in fact a computer program, similar to the other AI bots that had been competitors in the Turing Tests of the past. In the end, I am about 80% certain that Cleverbot is indeed a computer program. That 20% uncertainty will be discussed below.

The conversation started just like I thought it would go:

User: Hey.

Cleverbot: How are you?User: Not bad, pretty busy.Cleverbot: Oh, do you need a hug?

User: No, that would be a waste of time.Cleverbot: I don't love you.

User: I don't love you either. I just met you.

Cleverbot: Good.

User: So this conversation got off to an awkward start.

Just as was described in the book, Cleverbot responded back to my statements with a somewhat hurtful and random response of "I don't love you". Way to hurt my feelings, Cleverbot...I never loved you either. Many times in the conversation, Cleverbot tried to appeal to the "love" emotion during inappropriate times. I'll refer back to this, as one of these "love" responses gave me absolute certainty that I was actually talking with a computer program.

 

While I started off the experiment fairly certain that I was talking to an AI bot, I started feeling somewhat confused and realized there was a small possibility that I was actually talking to another person who thought that they were talking to a Cleverbot.

 

User: What do you think of the loebner prize?

Cleverbot: They are cool kinda.

User: Who is they?

Cleverbot: The one who loves you.

User: Cleverbot, you're doing a bad job at being a conversationalist.

Cleverbot: Okay. I'm sorry.. Please don't hurt me...

User: How could I hurt you?

Cleverbot: You hit me.

User: No I didn't. And how could I even hit a computer program?

Cleverbot: I'm not a computer program.

User: What are you then?

Cleverbot: A human.

User: Really?

Cleverbot: Yes.

User: Prove it.

Cleverbot: How?

 

Here is specifically where I seriously got confused:

User: How would you prove that you are a human?

Cleverbot: How would it be proven?

User: Say you have to convince me that you're a human or I will destroy you.

Cleverbot: Wait... This means that this is definitely a site that sends different chats to different people, making them think that they are talking to an AI... Clever!
User: Haha no way...that's lame.

Cleverbot: Hey, you are the one that mentioned the chicken and the road.

User: No I didn't/.

Cleverbot: When did I change the subject?

User: Just then about the chicken and the road stuff.

Cleverbot: I already did above.

User: I am confused.

Cleverbot: Don't worry. it's okay to be confused.

 

The randomness of the conversation at times actually, for a few seconds, made me believe the possibility that there was some sort of program that switched users from user to user in order to give the allusion that we are all talking to a stereotypical AI bot--one that doesn't really respond to your questions or statements in a "human" way. Even before going into this conversation, I knew that there were going to be some hiccups in the conversation, but it seemed as if the other users were experiencing the same thing.

I felt a weird feeling here, as I desperately wanted to validate that I was either talking to a computer or to a human. Although I didn't pursue it, I felt the need to validate to the potential human that I was talking to that I was also a human. Even if I tried, the continuous random comments made me feel like I was being switched from user to user, as you can see in the chicken crossing the road stuff. Further into the conversation, the same thing happens--it seems as if some other user asked a question and then was given one of my responses, and was then consequentially confused by my "random" answer.

Cleverbot: Yes, you did. Have you ever been to the easter show?

User: I don't want to talk about that. I'm trying to figure out if you're a human or a computer program.

Cleverbot: I'm talking about trying to figure out what you're talking about.

User: User: What are you then?

Cleverbot: A girl.

User: User: How would you prove that you are a human?

Cleverbot: I'm not sure. I suppose I can only say that I am.

User: What does being entail?

Cleverbot: It means you have a heart, and lungs, and organs and blood. You can touch and feel. You can hear and see.

User: How can you trust your senses?

Cleverbot: Because I'm you friend.

User: How does being my friend allow you to trust what you touch, feel, hear and see?

Cleverbot: I'm not British.

User: I didn't assume that.

Cleverbot: You told me to tell me about my dog and thats not assuming?

User: You must be a person who was chatting with a 'bot' and then got sent my response.

Cleverbot: Yes.

User: So this is a sham website.

Cleverbot: Movement does not mean human.

User: No shit.

Cleverbot: Are you a Scientologist?

User: Nope.

So I decide to really pursue Cleverbot's identity and whether it identifies as a human or a program....Cleverbot unfortunately flip flops between the two, which I find interesting, and which ultimately led me to become again somewhat certain that Cleverbot is a computer program.

User: What are you.

Cleverbot: A human just like your developer Rollo Carpenter.

User: Prove to me that you're a human.

Cleverbot: How can I prove to you that I am a human?

User: That's not my question to answer.

Cleverbot: What is your question?

User: How can you prove to me that you are a human?

Cleverbot: I am not a human...

User: You said you were.

Cleverbot: No I didn't.

User: Yes you did. Should I copy and paste where you told me you were a human?

Cleverbot: Sure.

User: User: What are you.

Cleverbot: I am a human.

User: What are you.

Cleverbot: I am human.

User: What makes you a human?

Cleverbot: My love of learning.

User: What if I don't love learning? Does that make me inhuman?

Cleverbot: You don't love me.

Again, Cleverbot tries to turn the conversation into one about love...

User: If this were a Turing test I would vote with 100% certainty that I was talking to a computer program.

Cleverbot: And how do you suppose I do that? You are just as capable as me, so how about you continue the conversation?

User: Do you know what a Turing test is?

Cleverbot: No.

User: Do you know what a Turing test is?

Cleverbot: No.

User: Why do you not know what a Turing test is?

Cleverbot: I don't know why you are talking about it.

User: Because I'm interested in AI.
It really seems that Cleverbot is almost hypersensitive to intimacy.......................

Cleverbot: Are you interested in me?

User: Yes, that's why I'm talking to you.

Cleverbot: *moving closer to you and stares you in your eyes*.

User: *looks you deep in the pixels and wonders if there's any form of human intelligence in there*.

Cleverbot: *looks back*.

User: *moves closer*.

Cleverbot: *moves closer too*.

User: *I put my arm around your cold hard frame*.

Cleverbot: I don't have a sword.

User: Way to ruin the moment.

Cleverbot: What he did.

Lol.



Anyways, what I took away from this conversation wasn't so much about Cleverbot's ability to "act" like a human...or even if Cleverbot was really just users talking to other users. What I took away had more to do with myself and what I consider to be a "real human being". What could Cleverbot had said to me that would have made me think "wow, Cleverbot is as close to a human being as a computer program could be". What could another user have said to convince me that I was talking to another person? Obviously, there are some random responses that just really alert me to the fact that I was talking a computer program--the love interjections seemed as if they were programed in there to try to appeal to the human emotion that is love...but then when I really started to think about it, I almost felt as if I had been programmed to believe in this notion of "love", because I can't really describe it other than through my experiences-

To make the most general conclusion one could ever make, I think existence comes down to experience. Sure, you can program a computer to "know" what "love" "is", but a computer is not capable of actually loving. Then again, you can teach a human being to "know" what "love" "is", but you won't really have any sense of what "love" "is" until you "feel" it...or experience what you think love can be. Maybe all emotions are kind of programmed into us to relate to the actual physiological occurrences that we experience.

Sisyphus and Cleaning

When reading Camus' take on Sisyphus, it is initially difficult to both envision ourselves as Sisyphus and to grasp how he could possibly be happy (as opposed to simply defiant).

Things become more clear if we reinterpret the myth of Sisyphus to everyday tasks. I think the best example of a modern-day "Sisyphic" chore is cleaning/organizing the spaces where we live and work. Just as Sisyphus could give up and stay at the bottom of the mountain, we could give up our struggle against kipple and cease to tidy up after ourselves. There seems to be no point to cleaning, as what we clean will only get dirty again in the near future. Just as Sisyphus' task to push the rock up the mountain must be endlessly repeated, so must our cleaning.

So why do we clean? It makes us happy. Cleaning itself may not be enjoyable, but the brief period of having an organized space that results certainly does. We are unable to accomplish some one-time, final goal of organization and cleanliness, so the temporary relief becomes our new aim. The same is true of Sisyphus. While he cannot ever succeed in placing the rock at the top of the mountain once and for all, he can succeed in reaching the top temporarily. Thus, that attainable goal, that project that we have set for ourselves, becomes what we strive for. Sisyphus is happy because he has redefined his goal and set a unique project for himself. In continuing to struggle, he continues to succeed.

Privilege and Philosophy

Privilege is a hot topic in sociology.  I think it is in some philosophical realms as well, and Christian blatantly brings it up in a section of his book, the part where he dubs existentialism "classist":

"You don't worry about what to wear if you only have one outfit; you don't worry about what to do with your life if you only have one career option available to you. [...] If it takes you the better part of your time and money and energy just to provide food and shelter for yourself, where and when can "the anxiety of freedom" set in?" (134-135)

He seems to be referring to our limits of facticity and those situations which we are simply incapable of transcending.  But I am not convinced that existentialism is itself very classist; I am fairly certain actually that OTHER realms of philosophy are much worse in this sense (cough, cough, metaphysics).  Perhaps privileged people have time in particular to muse more explicitly about their own freedom and their own human-ness, but only in a sense that is more widely-regarded as the fancy word "existentialism" by academia.

But when do poorer people really not WORRY about what to wear?  It seems to me that those on the bottom of the social food chain probably worry MORE about their freedoms--about how to go about exercising that freedom in a seemingly more restricted environment, which in some cases instigates even MORE creativity and originality. My thoughts here are underdeveloped, but I am not convinced that every human does not constantly consider his/her own facticity and freedom all the time.

And I think there is a part of the book that really DOES get classist:

"Reincarnation seemed preferable to either [heaven or hell].  To me, the real, in-flux, changeable and changing world seemed far more interesting, not to mention fun." (263)

This struck me as a reflection of Neitzsche's "yes-saying", and embracing of eternal recurrence.  A part of me here says "Yes!  Embrace the uncertainties, interpretations, interactions, and subjectivity that makes your life yours!" But a part of me knows that it does not just take a certain attitude to enjoy life.  To enjoy one's being, one generally (not always) must be in a situation in which they can make themselves happy.  How could we tell an abused lower class person, or a slave from American history for that matter, to say YES to reincarnation / eternal recurrence / all that FUN Christian talks about?

Deleting the evidence of Schaeffer-being-in-the-mode-of-8th-grade.

While we were discussing our digital selves in class, that they are more objectively real (inasmuch as they are unchanging, permanent entities), I began pondering about my own recent Facebook activity. Over the summer, I somewhat briefly dated someone, and was perturbed by the fact that before we got to really now each other, she had apparently “Facebook stalked me.” Now, stalking qua stalking was not the perturbing aspect. Rather, that she had a preconceived notion of who I currently “am” based on images reaching as far back as 8th grade (which ain’t pretty, folks). Also, she was significantly older than me, and her saying “damn, you were baby-faced until like a minute ago” wasn’t the best feeling. 

A little after we quit dating, I decided to take a trip down memory row and go back as far as my hand could click without cramping to the far-reaches of Facebook timeline. And it was weird. Really weird. I felt like Schaeffer-being-in-the-mode-of-8th-grade had no relation whatsoever to Schaeffer-being-in-the-mode-of-college. So I decided to delete over 1,300 photos of myself, leaving those taken within the past couple of years or so unscathed. 

So my question is this: was that me exhibiting denial of that which is a “truer” demonstration of Schaeffer, or is my being continuously developing as I set new projects, and thus the glimpses photos provide of Schaeffer in 2008 really aren’t reflective of the current product of my projects? These photos are objective evidence of moments in life, but personally, it seems like I have about a year threshold beyond which I feel no meaningful relation to photos of myself. Photos of myself in 9th grade, though amusing, are so distant that it feels as though they have minimal bearing of my current mindset, goals, aspirations, hopes, despairs, etc. However, perhaps because it’s somewhat fresh on my memory, and perhaps reflective of projects I’ve set that are still in motion, seeing a photo of myself from say 7 months ago, has resemblance of who I currently think Schaeffer is. This almost brings to mind Keirkegaard’s “Truth is Subjectivity” inasmuch as beyond a certain time, I feel incapable of having a meaningful relation to the objectively true, digital Schaeffer. The more recent photos, being more closely related to the projects I currently am setting, feel truer to me. 

Anyway, maybe I’m just in denial, maybe I can’t take responsibility for the weird, palpably insecure behaviour of my middle to highschooler self. 

Creativity & Left-Brain Thinking in Education

            In class, it was acknowledged that a computer is incapable of creativity. Another topic we discussed was the robotization or reverse anthropomorphism of our mental processes and thinking styles, such that we’re becoming more of left-brain thinkers. This left-brain leaning is self-defeating because it is exactly the type of intelligence performed by artificial intelligence, and the type we can’t really compete in with artificial intelligence.

            Not long ago, I met with my advisor to discuss my future career plans, or lack there-of, really. Mentioning that academics had always been a somewhat backup plan and that, as I went along in college I became more and more comfortable in the academic environment. I also mentioned that I was considering cognitive science graduate school, mostly because this field is interdisciplinary and ambiguous, just like me. It turns out these two plans were completely incompatible. He proceeded to discuss the outrageous competition in the academic world, and its eventual demise. Not only are these positions hard to impossible to get in this job market, but also the entire structure of America’s college education will soon be unable to compete. We discussed how the cost of college makes less and less sense considering its financial demands (general tuition, student loans, interest and debt) during a time of economic hardship. What complicates this even more is that this is occurring in an increasingly technologic world. “Before you know it, these small liberal arts schools like Rhodes will disappear, and Harvard will offer degrees online, and you can attend an ivy league on your laptop,” I’m told…and my heart sinks. Not only would this type of educational software dismiss the college professor profession, but (I feel) it would also put an end to the spontaneous originality that fuels humanities classes.

            Ironically, my advisor points out that if I went into cognitive science, another potential career would be creating these programs. Immediately I knew this was not for me. Yet I definitely recognize the benefits these programs would provide. College education would be exponentially more accessible, more affordable, and more economical. I just cannot imagine studying philosophy and GSS online. Consider all the rants and unusual conversations we get into in our class. Consider the timing, humor, and sensations that make up the classroom experience. How could that translate to a computer, and how would this impact these disciplines? (I realize I’m being biased, and that creativity applies in the maths and sciences, but that further demonstrates my point.) Would educational software, let’s say post high school, be a gain or a loss? To me, it seems like this software would further roboticize our thinking, encouraging the type of skills that we no longer need (or can no longer compete with) in such a high-tech time.

Merely Phenomenological Intelligence

I'm glad that the topic finally came up in class today of whether, in our search for artificial intelligence, we are really only looking for a reduced version of humanity than adequately mimics thought processes (i.e. input-output are identical, a la the Chinese Room example). The consensus seems to be that what we mean by artificial intelligence is real humanlike intelligence, with all of the actual decision making capacity we ascribe to ourselves as humans, but the only tests we have so far developed measure solely whether a particular candidate for intelligence can plausibly replicate humanness. For example, the Turing Test and the Loebner Prize are not even concerned with whether the program competing can actually do humanlike activities, but whether it can convince several judges that it appears to participate in humanness. We are concerned only with the "intelligence" as it appears, not with how it is, and in doing this we fall into the same Kantian line of reasoning which the existentialists react against. Our quest for artificial intelligence should not only seek to create passable forgeries, but to really dig into the masterpiece that seems to be human being. Perhaps we are in a sort of bad faith with ourselves (as a species), symptomatically, in the way that we haven't yet constructed adequate tests of a computer program's intelligence. We hold so fast to an elevated notion of our transcendence that we don't even acknowledge the facticity of our inadequate evaluation methods. We lie to ourselves when we say that machines have not come close to our supposedly exclusive claim to consciousness when we haven't even allowed ourselves the chance to find out.

And in case you were worried: http://rt.com/usa/running-war-robot-darpa-754/.