1. Utopian Speculative Future of Music Production

By the year 2051 technological advancement has taken music production to levels no one would have imagined in the 20th and early 21st centuries. It was all possible through the development of holographic computers, which are extremely powerful and interconnected. A worldwide agreement was made to guarantee that every human has access to a holographic computer, which is not dependent on physical material. The computer is a highly intelligent system that operates in conjunction with the user’s mind.

Digital Audio Workstations (DAW) have become very powerful, allowing music producers to create unprecedented materials. Its features include:

• Virtual session musicians. A function that enables users to have their musical ideas performed by expert real-life musicians whose performance abilities have been algorithmically translated into the program. This has allowed anyone to be able to create pieces of high-level performance, as well as creating an interconnected community in which those who have been trained to perform can offer their services to anyone around the world.

• AI sound processing. A function that helps users achieve perfect sound processing for mixing and audio alteration. As tracks are created an AI sound assistant ensures perfect balances in levels, equalization, compression, stereo imaging, reverbs, amongst others. These balances are defined through an interface in which users determine what genre of music they are working on. The program was developed and is constantly updated by sound engineers all around the world.

• The infinite library. All existing instruments have been virtualized and can be assigned to the virtual session musicians.

• Music knowledge download. This function allows users to download theoretical knowledge into their minds, giving them the instant capacity to compose, arrange, and create in any genre of music. This was possible through the creation of mental directories built on the knowledge of expert musicians from all around the world. When humans discovered that all individual minds are part of a collective human mind, and built a download system, the dynamics of learning were revolutionized.

Hence, music production has been released from many limitations and has become an activity that is more focused on creativity and innovation. Great music productions are created through the producers’ ability to intelligently combine the different functions of modern DAWs. Similarly, given the ubiquity of holographic computers, the virtual musicians function, and the music knowledge function, everybody has the capacity to produce music and express their unique selves through this medium. The culture of music and music production around the world has become one of appreciation of ideas and equity in which technology empowers human capacity. 

2. Dystopian Speculative Future of Music Production

By the year 2051 technological advancement has taken music production to levels no one would have imagined in the 20th and early 21st centuries. The advancements in Artificial Intelligence have transformed the way in which Digital Audio Workstations (DAW) operate, allowing all functions to be automated with minimal human intervention.

After all human skill was decoded to be applied using DAW’s virtual instruments, performing musicians are in extinction. This in turn changed the dynamics of the music industry, as humans are no longer at the forefront of live music performance. Instead, holographic avatars that connect to the decoded human skills system have become the idols of the entertainment business. Idolatry of non-human figures has led to many psychological, sociological, and philosophical issues, making people more alienated and unable to have empathy.

Traditional music producers are also becoming extinct as the Artificial Intelligence system of DAW is able to produce much more rapidly and commercially efficient. Powerful corporations use big data to understand sociocultural behaviors of consumption and incorporate that knowledge in DAW functionality to create pieces of monumental sales and subliminally promote their products. Hence, the music production industry is dominated by corporations that are able to use the algorithmic functions of the highly advanced DAWs alongside consumption data. This has made music a means for advertising and sales instead of channel for self-expression and artistic freedom. Even those individuals who are looking to use the DAW for creative and artistic purposes have been neglected, as the new industry dynamics have created a new type of “listener” in modern cultures, who is dominated by impulses and drives which are perfectly understood and manipulated by the corporations.

The total virtualization of musical instruments in music production software disrupted the industry and culture of musical instruments. Almost all companies that produced instruments have become extinct, as they were not able to compete with virtual instruments, which can be loaded in all devices such as phones and holographic computers, and performed through AI assisted programming. The small group of those interested in acquiring an instrument have to pay $50000 for an electric guitar or $450000 for a grand piano. Even when an instrument is acquired, the fast-paced dynamics of daily modern life makes it almost impossible for an individual to learn the instrument. Those who want to make a living in music production or music have no choice than surrender to the system and only those who are to produce for marketing and mind-control purposes succeed.

I thought the topic of this week was fantastic and really helped me to understand how algorithms function, their potential and dangers. The articles by O’Neil (2016, 2017) and the podcast by McRaney (n.d.) were particularly enlightening to me.

Rather than doing one microblog post using predictive text assistance, I was interested in trying out different text predictors to see what results I would find. I did the microblog post using two websites (Inferkit (Case 1) and Deep AI (Case 2) and my cellphone (Case 3), which I did in Spanish, given that it is the language set in my phone.

I was particularly surprised by the text created using Deep AI (Case 2) as it had a political tone that felt the most different from the way I would express myself. The post contains a quote that critically mentions Donald Trump. Anyone who knows me would know that I rarely speak about politics or politicians – it’s simply not an interest of mine. However, if I was to post this on my social media platforms it could easily be inferred that I have an inclination for political discussions and that I have a specific political position about a politician. In McRaney’s (n.d.) podcast it is mentioned that algorithms can have unintended consequences and we should think carefully about how they can affect people. The post is not dangerous in this explorative educational context but I do imagine how algorithms could lead to expressions that contain political, sexist, racist, and ethical implications that have nothing to do with our ways of thinking and create tension in our lives, as well as affect people.

The text created using Inferkit (Case 1) was also surprising, particularly because this platform didn’t create text on a word-to-word basis, but rather in sentences. The result is a very well-articulated and philosophical idea about the meaning of education. This made me think about Dr.Vallor’s (Santa Clara University, 2018) definition of AI as machine-augmented cognition for humans, because the experience of creating text through Inferkit felt like a more potent way of creating ideas compared to the word-to-word approach. I understand that Dr.Vallor’s  (Santa Clara University, 2018) refers more directly to using machines to perform extremely complex tasks in seconds, but I think it also applies in this case. This makes me wonder about some ethical issues about honesty in educational contexts. I actually went beyond the 250-word limit to see what else came up and the final result contained more complex well-articulated ideas. It even included quotations. I thought about how students could use this tool to cheat in schools, which is not only an ethical issue but also an educational one, as it represents an omission of doing tasks that are imperative for the development of skills (for example critical thinking or abstraction) that are needed to be successful in our society.

The text created using the cellphone (Case 3) provided insight into different aspects not found in the other two cases. I observed that, in this case, the algorithms were influenced by texts I’ve previously created from using my phone on a daily basis. I had a more clear experience of what was mentioned in McRaney’s (n.d.) podcasts about how algorithms predict the future based on the past. Some words such as “vida” are used a lot when I am communicating through instant messaging and it was interesting to see how it was suggested many ways while doing this task. I appreciate how algorithms can predict the future based on the past and have found that it is helpful when producing text in instant messaging. I understand now that this is one of the functions of algorithms, as they are precisely created to predict outcomes, yet I also see how, in other contexts, wrong ideas from the past (which could be racist, unethical, or sexist) could affect decisions we make in the future and create problematic feedback loops in algorithmic structures. This is the danger for which transparency and care from designers are urgently needed.

I was also left with the impression that the Spanish case was not as sophisticated, which makes me wonder about the technological hierarchies that exist between cultures, as in other cases such as MOOCs production.

Case 1: Inferkit

Text: Education is not about proving yourself, but about connecting with others and learning from them. Education is not about making you feel that you’re ready to perform or compete. It’s about engaging, being active, and getting excited about learning.

Case 2 : Deep AI

Text: Education is not about politics, it is about understanding the political world,” said Tom Reza, an assistant professor at Georgetown University’s Division of Criminal Justice. “All I’m concerned about is if President Trump believes he can achieve this or not he should be impeached if it does not happen.”

Case 3 : Cellphone:

Text: La educación no se trata de un problema de aprendizaje que se puede cambiar en la historia de la humanidad. El problema es que no se puede entender que es una forma de vida que no es fácil porque se trata de una forma de vida que nos permite vivir en una sociedad que se puede cambiar.

References:

McRaney, D. (n.d.). Machine Bias (rebroadcast). In You Are Not so Smart. Retrieved from https://soundcloud.com/youarenotsosmart/140-machine-bias-rebroadcast

O’Neil, C. (2016). Weapons of math destruction: How big data increases inequality and threatens democracy (First edition). New York: Crown.

O’Neil, C. (2017, July 16). How can we stop algorithms telling lies? The Observer. Retrieved from https://www.theguardian.com/technology/2017/jul/16/how-can-we-stop-algorithms-telling-lies

Santa Clara University (2018, November 6) Lessons form the AI mirror Shanon Vallor [Video] Youtube.com https://www.youtube.com/watch?v=40UbpSoYN4k&ab_channel=SantaClaraUniversity

Exploring the GUI was a fantastic experience to see more clearly the ‘dark patterns’ designers employ to lead the manipulate our behavior in digital environments. From the moment I opened the GUI I noticed I was in a very deceiving environment which purpose was to maximize the time I spent on the platform and the number of clicks. This made me think about Harris’ (2017) TED Talk about the persuasive techniques that tech companies use to control our minds. In effect, Harris (2017) mentioned that “the best way to get people’s attention is to know how someone’s mind works”, an inference we could make about those who designed this GUI. For example, we often feel stress by time pressure, a condition that was created with the GIU’s timer. Looking at it more deeply, the timer is instigating a sense of losing something (bringing up thoughts like “I’m running out of it!”) which created a condition of fear. Under such conditions we can lose concentration and clarity, which makes our behavior more erratic, meaning more susceptible to clicking the wrong buttons – which are the buttons they really want us to click. Fear could also lead to a paralyzed state in which the time we spend on the environment is maximized.

Another dark pattern I was able to quickly recognize based on my experience as a web and graphic designer was the intentional misuse of color to deviate our attention from the steps to complete the game. Often, the colors used in words that allowed progression in the DIU were similar to the background color. Similarly, the words were often small, not underlined, and positioned away from the DIU’s focal points. In this case, rather than an understanding of how our mind works, I would argue that some dark patterns are built on an understanding of how our sensory perception works to manipulate attention. I worked in the advertising industry for a couple of years, doing audio and multimedia productions for marketing campaigns. In the same way advertisers understand how visual perception works and this enables them to create visual objects (commercials, short videos for social media, posters, flyers, etc) that manipulate behaviour, they also create auditory objects for the same purposes. Something that often surprised me while producing the audio for commercials was how announcers and voiceovers were always directed to express in overenthusiastic ways. Many times, while recording, directors and clients would say “go higher”, “raise it”, or “do it with a bigger smile” to a point in which, in my view, sounded unrealistic (seriously, how excited can you be about buying a Pepsi?). I think that when a dark pattern is taken to unrealistic dimensions its deceiving intentions are more clearly exposed.

The concern I am left with is about the subtle dimensions of deception, which are much more difficult to scan. The GUI’s tricks and intentions were very easy to identify, but I am sure in other digital environments billions of people use daily (Facebook, Instagram, Google, etc) deception is way subtler. This would be like being approached by a thief whose whole body language and appearance clearly reveal his identity, and being approached by another thief who is disguised as an unharmful friend. For this reason, I think it is essential that we are educated about media deception and develop the necessary literacy skills that allow us to see through the lies. As mentioned by Tufekci (2017) the persuasion architecture of digital environments is much more sophisticated and potentially dangerous. It is important to understand also that this is not just a matter of selling products or manipulating behavior, but it also has the capacity to manipulate belief systems which are the foundation of an individual’s behavior.

References

Harris, T. (2017). How a handful of tech companies control billions of minds every day. Retrieved from https://www.ted.com/talks/tristan_harris_the_manipulative_tricks_tech_companies_use_to_capture_your_attention?language=en

Tufekci, Z. (2017). We’re building a dystopia just to make people click on ads. https://www.ted.com/talks/zeynep_tufekci_we_re_building_a_dystopia_just_to_make_people_click_on_ads?language=en

In this linking assignment, I’m going to make connections and analyze differences between the reflections my classmate, Tanya Groetchen, and I did for Task 9 – Network Assignment.

I found that Tanya did a good job at organizing the data to read it more clearly. She provided screenshots of how she moved around the nodes, separating the music tracks from the curators. This was a good way of separating the two different kinds of data from the graph. The size of the nods in curators doesn’t represent any information, however, for the music pieces, it represents how many times these were selected. Although simple, I think this was a very skillful approach and made the reading of data much easier. I didn’t use this strategy and see now how it could have saved me some time in data interpretation.

This makes me think about how important data organization is and how agency is needed for such a task. This could easily have been done with artificial intelligence but it seems like that function is not featured in Palladio. It’s interesting to contemplate how computers are crucial for the assembly and reading of data. Without such agency, how we read and find data today would be very difficult. Search engines and data analyzers precisely organize massive content in ways that make our life easier. However, as we’ve learned from this task, the computerized assembly of data can only go so far. This data has no way of providing information about what were the implications behind the patterns in the curation, and this is where we have to be extremely careful because we can make wrong assumptions. Does the fact that the Fifth Symphony by Beethoven was the most popular track mean that everyone likes classical music? Or does the fact that Johnny B.Goode was also among the most popular tracks means that most of us would enjoy ourselves in a rock n’ roll concert? Or maybe it shows that it is very likely that we’ve all seen Back to Future? Like this, we can ask lots of questions without a clear conclusion. I don’t think this is something negative – quite the contrary. Data can stimulate our minds to build assumptions through which we can question beliefs and expose bias. In this example that might not be of huge significance, but in other contexts, it could be a great way of learning about assumptions we make on race, gender, and culture. This exposure can help us to remove judgments, discrimination, and unethical practices.

Both Tanya and I explored the issue of the superficiality of the data, concluding that any reasoning on the motives behind the curation remains at a speculative level. To understand the deeper dimensions of why individuals choose certain things more data is required. I think that human behavior is so complex that to understand the mechanics of choice profoundly, we would probably need more data of a qualitative nature. I find it fascinating how, as we go deeper into the motives behind choices, at a certain point, there is most likely a shift from the individual to the collective, meaning that the thought which produced a choice might be only an impersonal thought that comes from a cultural heritage. In that sense, data can also provide us with insightful information about culture and groups of individuals. I appreciate doing this data analysis task, as it has helped me to understand the necessity of being careful while interpreting data, and also the capacity data has to storage individual and cultural patterns.

I am left with an intriguing question: is our personal makeup a complex set of data? Maybe this is why when we lack profundity in our ways of viewing people, we only see superficial data and make assumptions and wrong judgments. If we were to de-compose the data of an individual to the point that it is seen as a cultural construction, maybe we could see the essential ‘blank canvas’ of an individual (essence) while understanding that we are the same. This might be the way to undo isolation, alienation, hatred, fear – and find love.

I am glad this visualization has been provided, as I was curious to know what songs my peers had selected for last week’s assignment, and what implications could be made about our collective mind.

This visualization displays the popularity each song had for the curation. Pieces like the Fifth Symphony, Jaat Kahan Ho, Percussion, Johnny B.Goode, and Melancholy Blues were very popular. I wonder how much this popularity is biased. Could it be that these pieces were selected because of the Fifth Symphony’s high status and recognition in music, the current trends of interest in Indian and Eastern culture (Jaat Kahan Ho), or the familiarity and cultural resonance they evoked in the listeners? (Johnny B.Goode, and Melancholy Blues). I actually selected all of these pieces for my curation (apparently) based on other reasons, but this visualization makes me reflect on how my decisions were most likely biased by my cultural background. Of course, this is only one perspective and way of interpreting the data. We could also argue that these pieces have an essential element that evokes universal acclaim, meaning that they are truly special. I believe it can be seen in both ways and there is truth in both scenarios. This makes me think about ‘best songs of all time’ rankings in which often at the top we’ll find songs like “Imagine” by John Lennon, “Like a Rolling Stone” by Bob Dylan, or “Hey Jude” by The Beatles. It seems like their position is influenced by both biased cultural factors and true specialness.

The visualizations of communities of individuals with similar responses were useful to play with assumptions we could make about how these individuals would agree with one another. While looking at Community 4 (see image below) we can see how these 5 individuals would agree unanimously in several cases. However, it’s interesting to observe how some of them have some choices that don’t meet with any of the rest. Would these be points of disagreement? Could it represent potential conflict in some way? I don’t think this represents an issue in this case, as this curation is a playful exercise. But how about in other contexts such as dating or other kinds of relationship services? Also, It is interesting to think about how these connections are a digital representation of connections that exist in our daily relationships. I can imagine how the individuals who are close to us (particularly friends) represent numerous connections of ideals and preferences. Similarly, the difficulty we encounter in relationships could be seen as isolated (or conflicting) nods.

Of course, the graph only represents one level of agreement, which is the selection of the pieces. However, the rationales for the curation could be quite different, even conflicting. One individual could have selected the pieces based on emotional content, while another could have selected them based on cultural diversity. On the surface, the decision is the same, yet the motive is different. This could create some issues and misinterpretations of data. For example, let’s imagine that two individuals decided to have a celery drink instead of a Coca-Cola. We could make the assumption that both prefer the taste of celery drinks and are in agreement in that area. Yet, on further investigation, one individual might have chosen the celery drink only because of its health benefits and, in fact, doesn’t like the taste of celery at all, while the other individual loves the taste of celery and didn’t consider the health benefits so relevant. I believe this shows how much information can be erroneously interpreted and the potential danger of algorithms, particularly in ethical dimensions.

Image: Community 4
Green: unanimous agreement
Red: representations of possible disagreements