A computer can assess anything if it’s equipped with sets of definitions. For a computer, a definition is a set of incredibly specified characteristics. When it is applied to an input, if all of the definitional characteristics are met, an output results, confirming that the definition applies in that setting. If not all met, an output results, denying that the definition applies in that setting. People have definitions too, some of which are incredibly specific. For instance, in film analysis, the term chiascuro is sometimes applied to describe the way in which a director uses light and dark to provide a meaningful contrast in a part of a film. We have our own sets of definitions and examples of when these definitions prove true. In order to allow a computer to analyze a poem or movie, it just has to be provided a definition and several examples of when the definition is and isn’t true. However, when we attempt to program a computer to analyze some of the things that we analyze everyday, we find that our ‘definitions’ are more nebulous than previously imagined. For instance, how do we know when a section that has a particular contrast of light and dark in a movie is supposed to provide a significant meaning? When does an enjambment in a poem matter? It would take an incredibly long time to program all of these things. Also, the computer would have to be notified that sometimes a situation fits a definition even if it doesn’t fit all of the points in the definition, though there have to at least be some key features, three key features for example. A computer program that can interpret from somewhat nebulous, real world inputs already exists. There is a computer that can look at you in a doctor’s office and diagnose you. Many of the common symptoms of common illnesses have been programmed into this computer, and there’s no saying that that wouldn’t be possible for poems or movies as well. However, because poems and movies have the whole of history to reference, it would take a much longer time to program in many relevant things in order for the computer to make an assessment.

In our class discussion about humans versus computers, I kept making arguments that Professor Gil was able to counter on the basis that computers can be programmed to do almost anything. One of my arguments though, was that computers cannot live and learn through experiences like humans do, and although Professor Gil tried countering this, I still believe that there is no way computers can be programmed to interpret and make decisions based on past experiences.

In reviewing a book, a movie, or a song, the first thing I do is compare it to something else I’ve read, watched, or heard. Like this, I base most of my decisions off of prior experiences I’ve had. Maybe the song reminds me of someone I love or something I miss. Because of these connections we make based on what experiences we’ve had earlier in life, there are no two people in the world that will have the exact same opinions on certain books, movies, or songs. In this way, each and every one of us can bring something new to the table, something new and interesting that we can share with others, and like this, learn from each other and make different interpretations based on what new things we learn.

There is no way that computers could ever have this ability to respond and connect with this infinite variety of experiences it could be presented with. Like Professor Gil said, computers can be programmed to do many things in using assignments to follow a certain logic or pattern, but “connecting” to something means having a certain emotion in regard to it, and as mentioned in class, computers cannot be programmed to display emotion related to past experience. Computers are capable of doing a million things, but even with assignments, input/output, math/logic, and conditionals and recursions, a computer could not be successfully programmed to feel a certain emotion based on past experience as it listens and reviews a certain book, movie, or song.

When I interpret a movie, book, song, etc., I decode it based off of my personal experiences with other social artifacts. While each individual may share a common set of beliefs or interpretations with another individual or group of individuals, we all bring something unique to the analysis. No two individuals grew up in a completely identical situation. We are molded through our individual experiences and reactions to encounters with cultural objects. We then use those memories to make judgments and evaluations for the future. We interpret information with our own “cultural toolbox” and relate it to our social interactions. Therefore, I may base my interpretation of a scene in a movie, not only on the cultural cues and contexts, but also from my personal opinion on similar movies. I will establish whether or not I like a movie on how it compares to others I have seen previously and how I felt about those movies.

For computers, it is a different story. The system can be programmed to fulfill many functions and scan through a lot of data very quickly. While a computer can be taught to use assignments to determine a sequence or logic to determine aspects of aesthetics, the computer cannot be fully taught to interpret based on past experiences. There are so many differing interpretations in the real world. It would be hard to program a computer to be able to respond to all of the opinions. While it is possible to program all types of information, it is not the same as the feelings that go into the everyday functions of human beings. The computer could be programmed to study the make up of songs or movies or books and feedback the frequency of words or connections between words, but the true cultural significance may not carry the same weight. Though I am still learning what computers are capable of, I would assume that functions like assignments, input/output, math/logic, conditionals, and recursions could be used in combination to determine sequences, functions of different structures, word make up, trends over time, etc., the computer would just be missing the individual aspect.

Before one thinks about how a computer could be used to interpret a poem, a movie or a song, it should first be thought of how a human  interprets a poem, a movie or a song. When we do any form of interpretation, we do so by comparing the said medium with data already in our minds. For instance, if one wanted to make some sort of interpretation of something going on in a movie, then that person’s mind would take this instance that needs to be interpreted and do a few steps do make the interpretation. Your mind would call up your past experiences and information and then loop through them in order to make a valid comparison betweem what you already know and what you are seeing at the moment in order to make some sort of judgement call on what is happening on the screen. While this thought process may be human nature, the information in your mind had to have been pre-collected into your mind throughout your life before this interpretation can be made.

All of these steps in a thought process sound almost exactly like things that are used in programming: method calls, loops, conditionals, and database systems. For instance, if you watched the new Harold and Kumar movie, your mind will decide whether or not the current joke is funny by calling up whatever past experiences and choices you have with humor, seeing how they compare to the current joke, and executing a decision based on what comparisons were true. This all something that, while tedious, could be easily programmed using just basic structures of programming. While this process is an over-simplification of how the mind works and is not truly how thought process happens, it still is something that is capable of being programmed.

These days computers can do pretty much anything we want them to do, from simple calculations to a constant source of entertainment. Thus we find that the future is bright for computing challenges. This may be slowed down, however, by one of the biggest questions plaguing computer scientists: artificial intelligence. While advances in artificial intelligence have been strong, there are still fundamental differences, such as global knowledge, that have been holding it back. Global knowledge is the idea of using past experiences, cultural knowledge, and global events to form an intelligent opinion on a topic. It is this concept that has plagued AI developers for years, however there could be a way to write a generic code to encompass media, such as poems, music, and movies.

Except this is much more difficult than previously thought. The number of assignments would be enormous. For example, for a horror movie, there would need to be a variable for the number of times someone screams, the number of people killed, etc. with an almost unending list of variables to map out possible scenarios. This is so we can get a full review of the movie and not just a few who follow the limited variables available. The same basis can go with the input/output stage of computing. Let us say for poems the input would be the entire poem, just like a song would be the entire song, we could output almost anything. The number of lines, the rhyme scheme, the amount of times the song visits the chorus. This would be the main output of these two media, I think. The math portion is pretty straightforward since adding and subtracting are necessary for these processes. For a conditional, movies might play a better example to utilize. An example of a conditional could be: If person A says ‘string A’ and person B says ‘string B’ (we will assume something humorous), then denote this as a comedic phrase. However, we run into the same problem as before, the vast number of conditionals that would be necessary to encompass an entire comedy, or for that matter all genres of movies. The recursion, the last vital aspect of computing, might be a little simpler since it mainly involves repetition. For example, for a poem, we could write a recursion that begins at the first line and repeats through the entire poem line-by-line until the end for the same type of rhyme. This could then be repeated for the next line that hasn’t already been chosen as a rhyme to the first and you will eventually get a rhyme scheme. With these five vital aspects of computing applied to reviewing media, it seems it could be possible to create a code that would work. However, it would be so long and complex that it might not be the most viable option. When one tries to physically inject global knowledge into computers, the outcome could be messy.

I am attempting to become more familiar with computers this semester between this course, Intro to Programming, and Technosonics. While they have only covered the basics so far, I have gotten a taste of how computers process and interpret information. I have learned that you have to treat computers like they are completely  mindless and explain everything step-by-step. During java coding, I have observed how thought processes seeming like common sense to a human need to be broken down for a computer. Computers are fast and efficient with quantitative information, but lack in human subjective skills related to creativity, emotions, opinions, and conscience. While we have not dealt with complex material in my classes, I can speculate about the capabilities and limitations of using a computer to interpret poems, books, movies, or songs.

Computers and machines can use loops, arrays, and recursions to perform tasks like counting the occurence of specific words in these artforms. Conditionals allow the programmer to tell the computer “IF..” (you see this input) “THEN…” (perform this output). I suspect that there are further skills a computer can accomplish when analyzing such material beyond my programming skills and knowledge. However, I believe that the nature of how computers learn methods from a human doesn’t allow for the machine to comprehensively review a cultural object to the fullest. In songs, for example, a human can observe the piece as a whole and describe what moods and feelings it can provoke; meanwhile, a computer may simply have conditionals to label the song as a specific mood if certain words are included. The computer probably wouldn’t detect metaphors and other literary tools that humans can analyze. Humans can see beyond the literal words and reflect on deeper, hidden meanings in pieces.