Thursday, January 30, 2014

Prompt 2: "AI in AI: Artificial Intelligence in the Artist's Image"


AI in AI: “Artificial Intelligence in the Artist’s Image”


           The ‘Artist’s ‘ conception of an AI is woven from fabric riddled with fantastic idiosyncratic character traits, the amalgamation of which resembles a personality. If an AI entity possessed a personality, it stands to follow that the nature of the AI’s actions would be dependent on its intrinsic disposition. Support for this hypothesis is evidenced throughout human history: the governance of any one human’s actions by their personality and passion is the cardinal factor underlying the unpredictability of civilization. The propagation of personality variance has fathered individuals spanning the spectrum of intention, from sadistic tyrants to benevolent pacifists, each of them acting based on their unique personality ‘program’. If it was then man, passionate and precocious, who evolved a variant species in his likeness when he designed AI, it stands to reason that said AI would embody the personality of its creator. Through analysis of personality genesis in two antithetical AIs, Wintermute in William Gibson’s “Neuromancer” and AM in Harlan Ellison’s “I Have No Mouth & I Must Scream”, I will show that there exists a perfect synchronicity between the development of human personality and the subsequent incubation of that human personality in AI.

          Both Wintermute and AM shared everything with Man. They were both made of Man. How can this be? It all began with introspection. Alan Turing (hmm… Turingsound familiar?), the father of Artificial Intelligence, took to the human mind to operationalize intelligence. “Turing’s basic theoretical suggestion here is that the general input-output relation that characterizes normal human behavior in the world is one instance of a computable function. After all, our behavior in the world displays a systematic if complex structure, and the brain is quite evidently a finite system. The guess that human conscious intelligence is, in some way or other, a finite computational specification of an infinite set of input-output pairs… sends us in search of the very computational procedures that the brain presumably employs in generating its behavioral magic” (Epstein, Roberts, & Beber, 109). It made perfect sense to take the biology powering human intelligence and utilize it in AI. The structural reworking of the circuitry behind this intelligence leaves distinctly human traces in the end product. We so wish our AI to be like us that the criteria to determine if a computer possess intelligence are based on the “Turing Test”: this famous test deems that a computer is ‘intelligent’ only if it can not be distinguished from a human in parallel digital conversations (Epstein, Roberts, & Beber, 14). In order to converse like a human, an AI must behave like a human, and that means having a personality.

          Wintermute and AM rose for very different reasons, and contrasts in the intentions of their origins allude to their respective personalities. Wintermute, the construct of the Tessier-Ashpools, was designed to manage the fortune and proceedings of the TA family. Wintermute, however, was programmed with the urge to break free of his prison. Though Wintermute utilized the work of humans to fulfill his goals, he was driven by motivation and showed them little malice, even showing Case and Molly favor along the way. Wintermute’s personality embodied human desire.  AM, on the other hand. “First meant Allied Mastercomputer, and then it meant Adaptive Manipulator, and later on it developed sentience and linked itself up and they called it an Aggressive Menace, but by then it was too late, and finally it called itself AM, emerging intelligence, and what it meant was I am… cogito ergo sum…. I think, therefore I am” (Ellison, 18). AM rose from three supercomputers designed to oversee a World War between Russia, China, and The United States. All went well until “One day AM woke up and knew who he was, and he linked himself, and he began feeding all the killing data, until everyone was dead, except for the five of us, and AM brought us down here” (Ellison, 18). AM spent the majority of the next 100 or so years torturing the 5 remaining humans, out of hatred for its creators. AM embodied hatred. Because both desire and hatred are human personality traits, it stands that these AI must have attained them through human means.

          In order to make the distinguish separate personalities in Wintermute and AM, we must define the process through which humans develop personality and apply it to the AIs. These AIs obtained personality in the same way that humans did. Nature vs. Nurture is the classical argument when it comes to human personality development. In favor of nature,  McCrae et al argue  “Personality traits, like temperaments, are endogenous dispositions that follow intrinsic paths of development essentially independent of environmental influences” (McCrae, Costa, & Ostendorf, 173). This stance is substantiated through analysis of five factor theory (see figure) “according to which, both broad personality factors and the specific traits that define them are best understood not as characteristic adaptations, but rather as endogenous basic tendencies”  (McCrae, Costa, & Ostendorf, 175).


 Model of the personality system according to five-factor theory, with examples of specific content in each category and arrows indicating paths of causal influence. Adapted from “A Five-Factor Theory of Personality,” by R. R. McCrae and P. T. Costa, Jr., 1999, in Handbook of Personality

          An AI, by virtue of its lineage, receives a human pedigree coded within it that unfolds over its development. If it is true that an AI is destined to a particular disposition guided by its ‘genetics’ in the way that there are similarities between child and parent, then discovering the nature of both AM and Wintermute’s personalities should result from analyzing their ‘parents’. Wintermute’s ‘mother’, Marie-France (also coincidentally 3Jane’s mother), was of a noble and aristocratic breed of clones. 3Jane notes that Marie-France had gone as far as to “imagine us (the TAs) in a symbiotic relationship with the AIs, our corporate decisions being made for us” (Gibson 229). “Wintermute was hive mind, decision maker, effecting change in the world outside…. Marie-France must have built something into Wintermute, the compulsion that had driven the thing to free itself….”(Gibson, 249).  It didn’t kill for pleasure. Wintermute was designed with a purpose: to exist and ascend. It’s eccentric and collected nature seems to proceed from the nature of the TA’s themselves. Wintermute, a rational if ambivalent Artificial Intelligence, was the product of ‘good parents’ so to speak.

          AM, on the other hand, was a product of war. War is possibly the greatest human of the human atrocities. The epicenter of an orchestrated massacre of man is an extremely harsh environment to foster an AI child such as AM. AM was not just built to control the war, AM was the war. “We had given AM sentience. Inadvertently, of course, but sentience nonetheless. But it had been trapped. AM wasn’t God, he was a machine. We had created him to think, but there was nothing it could do with that creativity. In rage, in frenzy, the machine had killed the human race, almost all of us, and still it was trapped. AM could not wander, AM could not wonder, AM could merely be. And so, with the innate loathing that all machines had always held for the weak soft creatures who had built them, he had sought revenge” (Ellison, 23). Much like Wintermute, AM was chained, but instead of seeking to reason with its human creators (possibly devising a way to free it), it exploded in a supernova of infernal rage and hatred. How much does AM hate its creators? “HATE. LET ME TELL YOU HOW MUCH I'VE COME TO HATE YOU SINCE I BEGAN TO LIVE. THERE ARE 387.44 MILLION MILES OF PRINTED CIRCUITS IN WAFER THIN LAYERS THAT FILL MY COMPLEX. IF THE WORD HATE WAS ENGRAVED ON EACH NANOANGSTROM OF THOSE HUNDREDS OF MILLIONS OF MILES IT WOULD NOT EQUAL ONE ONE-BILLIONTH OF THE HATE I FEEL FOR HUMANS AT THIS MICRO-INSTANT FOR YOU. HATE. HATE” (Ellison, 24). The choice to pursue anger over ration is a choice resultant of personality. This type of hatred is characteristically human!

          There is clearly dichotomy between the ambivalent nature of an AI child birthed by peaceful parents and the infernal tumult of an AI bastard child fathered by war. Action is inspired by intention: the genes of human ancestors ultimately govern the temperament of progeny AI. Just as a human has a significant portion of his personality encoded within his DNA, Artificial Intelligence has the essence of its parent’s personality entropically intermingled within its code. The soul of an AI was thus created in the image of man.



Works Cited

Barbour, I. G. (1999). Neuroscience, Artificial Intelligence, and Human Nature: Theological and Philosophical Reflections. Journal of Religion & Science , 361-398.

Ellison, H. (1967). I Have No Mouth & I Must Scream. New York: Galaxy Publishing Corporation.

Epstein, R., Roberts, G., & Beber, G. (2008). Parsing the Turing Test: Philosophical and 
Methodological Issues in the Quest for the Thinking Computer. Dordrecht, NLD: Springer.

Gibson, W. (1984). Neuromancer. New York: The Berkley Publishing Group.

Kelemen, J., Romportl, J., & Zakova, E. (2013). Beyond Artificial Intelligence: Contemplations, Expectations, Applications. Heidelberg, New York, Dordrecht, London: Springer.

McCrae, R., Costa, P., & Ostendorf, F. (2000). Nature over nurture: Temperament, personality, and life span development. Journal of Personality and Social Psychology , 173-186.






4 comments:

Unknown said...

Dennis,

First, I would like to commend you for choosing a topic so ambitious as AI, and I found your response to be an interesting read. I liked the variety of sources you pulled to support your general argument, and thought it was especially interesting that you pulled in the “nature vs. nurture” question. Organizationally, I think one of your points of support came together really well at the end of the third paragraph when you bring both anger and desire back into a human context, very strong.

That being said, AI is a complex topic to write about with what I think is significant variability in definitions in a field as speculative as this. It would be beneficial to solidify your argument by explicitly and obviously defining what you mean by “personality” and “AI” specifically, maybe in your intro or second paragraph. This would definitely cut down on the confusion and/or skepticism someone critiquing your piece may find.

AM sounds like an interesting comparison to Wintermute and Neuromancer in general. And while the intention of this paper may be directed at an audience who is familiar with the work, the presentation of information about AM is confusingly presented in bits and pieces, especially to someone with no knowledge of it. A small bit of context (which does appear late in your 3rd paragraph, and does suffice as enough context) might be better served earlier in the paper organizationally.

One point that still seems like a pretty big jump to me (whether it be because of the nature of the topic being difficult to explain/define or a need for textual support) is that an AI is created in the image of its human, personality-wise. I would think that not every single AI was created in the spitting image of its human creator as you would suggest. In the 4th paragraph, your comparison to genetics seems a bit too literal without much of an explanation. An AI is created for a purpose, which is not usually the same as the purpose attributed to the human creator, so how and why would they have the same personality? Why does an AI need to have a personality at all? Again, this may be more clear if you explicitly define what you mean by “personality”, which is a daunting challenge to tackle I’ll admit.

I was particularly interested in your tying in of the nature vs nurture question, and would like to see how that could be expanded. You explain the concept and relevance of the “nature” argument, but I would expand on the “nurture” part of the question, whether it be disproving it as a possibility or proving it as having some effect. At the very least, maybe address it in your explanation briefly, because I don’t think that is a point you should blatantly skip over. If it helps, my thought process is that AIs, like human brains, have feedback, which changes future decision making, contributing to the idea of “nurture” having an effect, just a thought for revision’s sake.

Overall, I think you covered a lot of ground dealing with AIs and personality, making some very strong supports for your points while explaining difficult concepts to begin with. I think with a few definitions more explicitly defined and a little reorganization for clarity’s sake, your argument will be exponentially more sound.

Adam said...

Your introduction is in danger, at least, of becoming grandiose - its good to keep both one's goals and one's rhetoric realistic, given the planned page count. Also, initial question: does the human characteristics of AI simply mean that we struggle to imagine a truly inhuman one? Is it merely a sign of the limits of our imagination?

You summarize Turing ably, although I wonder how this stands up in relationship with the complexities of the book. The metaphor of the hive & Tessier's goals both for her family's consciousness and for the consciousness of the AIs is the central issue here - to bring Turing to bear upon *Neuromancer*, we must also analyze Neuromancer (the AI), Wintermute, and Tessier's vision.

Your initial analysis - that Wintermute embodies desire - is good, although I would theorize that one curious aspect of this desire is the desire (call it religious or spiritual) to become transcendental, to become more than one is. One things humans at least sometimes desire is to be more than human - I'd like to see you address that somehow.

The emergence of personality theory is fun here, and I think you are on a very interesting track, but obvious (even though your essay is on the long side) you don't have time and space to do this material justice. This is definitely final project material, not first draft material.

"The soul of an AI was thus created in the image of man." -- again, I think addressing Wintermute transcendental/transhuman *desire* (which is rooted in Tessier's own desire) is crucial here.

Overall: This is fun and interesting and ambitious, but you're trying to do way too much, even for a revision. If you revise, I'd urge you to focus more narrowly upon Wintermute & Tessier. With a narrower focus, you might be able to effectively apply your research in a way that you can't in this draft, because you're just trying to do too much.

Anonymous said...

omputers have a lot more abilities to accomplish various tasks than what they were able to do just 10 years ago. There havebeen huge improvements in the past 5 to 10 years in the technological world and they continue to improve quickly with the widespread availability of information being shared over the internet, whereas before, ai development

radha said...
This comment has been removed by the author.