BL G
Week 01:
Life 3.0 by Max Tegmark
Tegmark defines life simply as ′a process that can retain its complexity and replicate′
He looks at the bodies of all live creatures (including uni-celular bacteria) as our hardware, and their minds as software.
life 1.0 - where the software and hardware need to evolve rather than designed
Life 2.0 - hardware is evolved but still finite, but software is largely designed.
Life 3.0 - can evolve and design both it′s hardware and software
This book is a fantastic mixture of cosmology, physics, dystopian scenarios and ultimately the human condition. Tegmark tells the story of the universe from the point of view of the universe, involving us as merely loose specs of dust with big egos and god complexes.
The concept of a supercomputer and AGI (Artificial General Intelligence) is introduced as a natural progression in this epic story. Tegmark takes on the monumental task of explaining the concept of AI, neural networks, quantum computing. etc and does it in a remarkably logical and evocative way. His enthusiasm is catchy and the more you know the more you want to keep on reading!!
Past technicalities, Tegmark goes on to describe the ethics (or the lack of them) regarding the topic of AI. He takes you through different scenarios in which AGI could (and will) challenge human intelligence. From dystopian worlds to idyllic ones, your view on technology will be challenged with manifestos, short stories, and diagrams.
The subject of creativity and art are of course touched upon. Tegmark includes a fascinating argument todo with art and technology and the natural progression of them both. He talks about artists' works and computer vision work and combines the two in what he hopes will be next medium.
Life 3.0 is an exciting futuristic journey introducing the world of AI no matter what background or previous knowledge. It′s definitely one of my favourite books Ive read this year and one of the reasons I decided to take this course.
Snelting, Femke. 2006.
A fish can′t judge the water
How software choreographs life:
-
Software tells you to stop when the red man is standing and cross when the green man is walking
-
Software dictates whether my visa scans correctly
-
We all have to tap our cards to get on the train don′t we?
-
Our fingers can′t get enough of swiping right
Sure, the software is an extension of us, but just as much as a wooden spoon is. We love using technology, we′ve always strived to do so, even chimps use technology to eat termites. This, however, doesn‘t mean that the outcome is the tech itself. It‘s just another tool.
Since technology is just another tool It makes sense that it sometimes can have an agenda. However, most of the software that governs us today serves a purpose, mainly to organize, speed and enhance the way humans live.
Notes on the reading:
-
Insertion of procedure into human knowledge and social experience
-
We are not used to this, we are used to doing it ourselves, relying on our own intelligence
-
We have to be ok with the computational generation of knowledge of decisions, mostly because it is going to happen whether we like it or not.
Just like the horse and cart felt when the cars were invented and they were not needed anymore.
-
Our ego gets in the way, we think we know best
-
True, we do need to be aware of the meaning ′algorithm′, too many times this word is thrown around, half the population link it to their maths homework from GCSE.
An algorithm as a technical solution to a technical problem.
1. Algorithms are used to follow a model
You can measure the relationship between different elements of the algorithm. It is crazy the number of algorithms that we use day to day if you think about it.
Even before we had modern technology such as screens or keyboards we have been using algorithms to follow models to generate tasks, the best example is big assembly lines. With technology, this is growing exponentially.
It works on the basis of action-reaction, depending on the input, the process will be different hence output will be different.
2. Algorithms can be trained
Humans input parameters - it learns to pair queries, look for a yes or no answer, 1 or 0
To change the algorithm, it's not necessary to rewriting rather redesign the thresholds or values that certain parameters hold.
It′s usually the humans that are inputting the biased parameters, the algorithm is just following rules, it has no biased, it doesn't care.
For algorithm designers, the algorithm is the conceptual sequence of steps
2. Algorithms as committed to procedure
The more you think about it the more you are surrounded by algorithms, whether they are explicitly in front of you or hidden and camouflaged as something else.
Humans however as much as they complain and criticize algorithms, can't seem to get enough of them. Sure it is an unfair process, but it has already started.
Something that is not an algorithm can be something that relies on randomness?
Research an Algorithm:
For this task, I chose to investigate the calculations that govern traffic lights. I decided to interpret this in a B movie style.
Mr. Bucket worked in the assembly line until he got replaced by a mechanical arm. Poor Mr Bucket :(
Tay was an AI chatbot designed by Microsoft to learn through human interaction. Though great at first Tay quickly (in less than 24h) got corrupted by the dark influences of the internet and because racist and Xenophobic
Also: take a look at the time the intervals the tweets were written.
Week 03:
Chapter 13: Figuring the Human in AI and Robotics
- Right of the bat, she talks about creating/researching machines that are human-like. She essentially wants to create AGI.
- Imaginaries (sociology): is the set of values, institutions, laws and symbols common to a particular social group and the corresponding society through which people imagine their social whole.
- The topic of figuration, how humans and language all have figures for communication
Donna Haraway: Really interesting aim of understanding science as culture, as a way of shifting the frame of research,
Claudia Castañeda: figuration of children
Euro-American imaginaries are one of the autonomous, rational, agency and projects of artificial intelligence reiterate that culturally specific imaginary.
Automata and Agency
Agency: The capacity of individuals to act independently and to make their own free choices. By contrast, structures in those factors of influence determine or limit an agent and their decisions. (For example, social class, religion, gender, ability etc)
Do ′things′ have agency as humans presuppose?
The Euro-American view is that: to be human is to possess ′agency′ and then proceeds with the question of to whom or to what such attributions should be extended-
AKA in much simpler English: according to Euro-American view if it has agency its alive.
It′s not impossible to establish the criteria of humanness but indeed it is very hard
Jessica Riskin: over the past three centuries human and machine development has served as a model for one another, AKA they have been growing, feeding and learning from each other to better themselves.
In 1700 with the invention of the Look we get one of our first examples of the hybrid human-machine partnership. Even though the loom still needed human operations, it relied on it′s ′computational power′ to carry out complex automation such a weaving.
OK GREAT! That happened 300 years ago. NOW machines and automation are more in tune than ever before.
What figures of the human are materialised in these technologies?
Embodiment
Humans are aware of their bodies, machines? well unless we tell them to.
Because they are machines they need to rely on different inputs and reinforcement learning. A machine locals ints importance in relation to a successful oration of mind or some form of instrumental cognition.
Can cognition and the knowledge that it presupposes be modelled separately from perception and motor control? Book′s answer is no…… Computer says No……..
Emotion
This is an interesting one because emotions are perhaps the tool humans use the most to identify themselves within society and with themselves. The recreation of emotions in a computer will never be achieved but one can get very close.
Does it matter that the computer doesn’t know emotions as long as the human receiver feels that the computer has them?
2001 is not up to date enough, a lot has happened since.
The emotional process, as well as a reason, are necessary for true intelligence - this means that intelligence remains the defining capacity of the human - yeah but computers now have intelligence and we still don’t believe they are human
′emotions were understood as processes in the general scheme of the body as a machine, thus emotions was a pattern written in the language of biology′
sure language of biology but it′s a language none the less, a language that can be mimicked because as Harraway mentioned earlier, language it's figural.
Essentially for us to be fooled by a machine it needs to read our emotions better than us and then display them back to us in the form of empathy or emotional awareness. It doesn’t matter if it's not feeling those emotions, because it never will.
Sociability
Funny that it's most associated with the term of Science Fiction
Cog and Kismet - example of back in 2001, AI has gotten much more advanced than that. MIT's debating machine
They upgrade themselves through interaction with humans. they build more complexity.
′study models of human intelligence by constructing them on a physical robot′
Are we obsessed with ourselves? is our ego that big?
I didn’t think the Kismet example of what that great to be honest. It was a bit outdated. Nowadays there are robots capable of much more. Boston Dynamics, Alpha Go, IBM Project Debater, even works of art like Female Figure by Jordan Wolfson would have been a bit more exciting.
The humanlike machine as a Fetishised object.
Sure, I guess we are all inspirited by nature
The parent circular trajectory: the implementation of humanoid robot caregivers,
I′m not too sure what she is refereeing too in this paragraph but I agree that the technology is there however knowing humans we have to be careful to align the Also goals with ours.
She’s worried not that the robots will be able to able to accomplish these tasks but that the goals and inspiration will be reduced and blinded.
What does she mean by If there is a disturbing circularity is an expectation for robotic simulations of human development
I think she is talking about two different things, the AI brain and the Steel Circuit body?
Chapter 14: Demystifications and Reenchantments of th human like machine
Again repeating the fact that humans, in many disciplines such as tech or anthropology have had a concern about the merging of machine and human.
Posthuman: interesting term. it also sounds like a pun on posthumous.
Hayles argument: the possibilities that computing affords for rethinking traditional conceptions of the human. She, however, thinks that humanists will be against this idea due to their beliefs.
Suchman, however, has a different view: She thinks humans will not was the merge between robots and AI because humans are still tied to the church ′these projects continue to restate the parochial and conservative forms of liberal humanism ′
Schuman is, therefore, going to find out what ′conceptions of intelligence and interaction are at play
Schuman is going got meet with Kog and Kismet (LOL)
Are robots computational artefacts for humans to use? Guess so
Kismet interacting with his creators
Mystification and Enchantments
I think it′s quite true that humans have always had a liking for inanimate objects. Humans have been using artefacts and giving the meaning for thousands of fo years. From magic and sorcery to toys or the Queen′s crown. Humans attach emotions to everything.
Shuman then starts talking about capitalism. She argues that one of the reason for robotic fetishism is the age of commodity capitalism.
′Will we be replicated′ vs ′in what socio-material arrangements are we differentially implicated and with what political and economic consequences′
Demystification
′bit rot′ the degradation of code int the absence of ongoing maintenance of its compatibility with continually changing software and hardware environments.
That′s what happened to poor Cog! He had been discontinued, Kismet was working, however, his interactions with Human and her colleagues were not successful. Probably because it was 2001.
Head was a little bit different (also 2003): better chatbot, could actually ask it questions and it had a large enough database to give coherent answers.
Nevertheless less on two locations, it showed signs of its machineless, on interrupting and on the wrong naming ′still lucy′
Difference between Randomness and Open-Endless
However Head was able to recite acceptable poetry. That means he had a good understanding of the human language and the relationship between its words.
One thing is certain, there has been a collaborative performance to create Head and a lot of the technology that was being pioneerd back then is now evolved tramatically.
How these figurations take place: machines and technology figure the body in differed ways. VR headsets look at body differently than an MRI scan or a Mocap Suit. They all envision the human body differently.
Suchman, Lucy. Human-machine reconfigurations
Week 04:
GANspection
Hammad A. Ayyubi
Department of Computer Science
University of California, San Diego
Generative Adversary Networks
I chose to investigate GANs because of the fascination I have with identity and the ethics behind who/what/where your identity comes from.
From deep fakes and facial recognition to social media and exposure to the internet, our face our faces and identities are now part of a huge dataset that is being used knowingly and unknowingly by an array of different companies, services, government agencies, and economics.
A random thought:
Animals look all the same - Humans don't. That is a great achievement in evolution.
Keep on reading
In order to tackle these questions firstly, I need to understand how GANs work. How the are trained, the dataset they use, the ethics behind it and the potential in their use for the future of technology (in the case future of my artistic practice, but we will leave it broad for now)
Ayyubi does a great job of not only explaining how a GAN learns and classifies data but also at answering a key question: do Gans learn a data probability distribution or do they memorize image/data.
Find the article below:
https://arxiv.org/abs/1910.09638
Swirling celebrity faces created by an AI program from NVIDIA.
Image: NVIDIA
Jeniffer Lawerence′s body being deep faked with Steve Buscemi face as she does a speech at the Golden GLobes
ARTICLE THOUGHTS
1.Intro
GANs allow a very special type of machine learning called unsupervised learning. They have become very popular lately because of the way they can process and catalog/label heaps and heaps of data
On one hand, we have deep supervised learning models - they require data that has already been cataloged (probably by humans). GANs seeks to bridge this gap by learning data distribution.
They do this by constantly sampling from the learned distribution and ensuring that the sample looks like real data.
2.Related Work
2.1 Generative Adversary Networks
The amazing thing about GANs is that they can generate realistic samples.
GANs proximate the data probability distribution implicitly and constantly draw samples from it. The ultimate objective s to make this drawn sample indistinguishable from the real samples.
The discrimination tried to distinguish real samples from the fake ones and the generator tire to oil the discrimination to pass the counterfeit ones as real.
Thus they work on the principles of game theory where one player tries to better the other one.
2.2 DCGAN
ReLu: In the context of AI Neural Networks, the rectifier is an activation function defined as the positive part of its argument. It was popularised in 2001 as a great activation function for neural network training. As of 2017 its been the most popular activation function for deep neural networks. Its a type of batch normalization.
2.3 PgGAN
pgGANs function is to generate high quality and high-resolution images. The key idea utilized is progressive gradual growing for the model complexity while training.
At the start, because both generator and discriminator only have a few layers the quality of the image they produce is poor. However, as the training progresses more layers are added. As such they start modeling finer detail and much higher resolution
3. Method/Approach
So far in this article, there are so many new terms and subjects that I am not very familiar with, however, the more I read on and investigate the more I understand how things work.
As we progress Ayyubi describes the learning process of the DCGANs and how they employ the following methods to investigate it
We are going to explore Circular interpolation, Extrapolation, and Vector Arithmetic
This bit of the article was very technical so I thought it would be better to share extracts of it
As you go through each approach, it becomes more and more apparent how complex and logical machine learning can be.
uiñ
4 EXPERIMENT
The two datasets that where used I would say came from a reliable source. However how much data, accurate data is needed to create a successful GAN model?
It was so interesting watching the difference between images created through circular interpolation vs images created through Vector Arithmetic.
The visuals displayed after different times and models are truly fascinating.
Also the idea of BIG DATA and the TIME it takes to train models is something that at least for me was very underestimated. For the first 50% of your training time, the algorithm will not give you what you want. This gradual process of learning is fascinating to me because you can see it growing by itself.
5 Conclusion
The article concluded that GANs to indeed learn meaningful latent representations and encode semantically relevant information. Through the experimentation there where different methods applied to the learning techniques of the DCGAAN such as linear interpolation/extrapolation and circular interpolation fo latent vectors.
After training the data proved that there is a meaningful latent space representation that was indeed being learned rather than image memorization. Moreover, the process in which the article described and tested its data is also a very good way of understanding how GANs learn and the quality of their learning.
A GAN is not just a copy and pasting machine, it has rules and regulations that dictate and steer the algorithm into the user's outcome. It is true, the user has limited control and at the end of the day it all comes down to the database and the rules you give it that determines how well or WHATit will learn.
Nevertheless, in terms of image generation GANs are like anything I have ever seen before and after reading the article I look forward to exploring them more.
————————————————————————————
Class Work
Pair up with someone who has the same interest topic as you do and engage in a discussion about your interests.
how your article relates to your research
- it looks at machine learning through neural networks
- more specifically Generative Adversary Networks
- really interested in generating data from humans, this is the way to do it
- understanding how machines learn to generate high-quality images in different ways
3 KEY Words
cloning, consciousness, dystopian
EXPLORE FURTHER
cloning, identity. Who owes your image. Can you generate an identity?
WHY
Our relationship with technology is getting blurred by the second, especially when it comes to our images and the internet.
I am really interested in the ethics behind human-AI cloning and how this will drive humans AGI. To what extent do public figures own their faces? Do we have enough information to create a clone of someone, if so will this clone be more human than the person?
Week 05:
Chapter 13: Figuring the Human in AI and Robotics
Unrolling the Learning Curve
Sofia Audry
Have an alternative approach to Machine learning and AI for artistic creation
using the training face of machine learning and the generation process as the art itself
to create the unexpected. She is going to be using LSTM recurrent neural network applied to text generation.
Intro
Artist seeks to create the unexpected. Machine learning can provide that because of algorithms. However more often than more, it just serves as a means to copy and transfer the artistic styles. It just needs data.
“indeed machine learning is designed to recognize regular patterns, and when employed for generative purposes is attuned to reproducing things that already exist:”
I think this is a key difference between ML for artist purpose because artists (like said before) see to create the unexpected.
Context
ML originates in cybernetics (need to learn more about the origins of cybernetics)
I also love the fact that Audry refers to it as a disruptive science. Its true biology, neurology, anthropology all need/ rely on cybernetics to thrive.
When Audry goes on to talking about Agents is when this paper starts to get very interesting. She refers to them as:
“A system of agents that use feedback from their environment to adapt over time by trial and error.”
Just as Max Tegmark described before, the is very similar to life 1.0 How curious that this structure happens tone at the core of deep learning. After a period of time, these agents are force-fed gigabytes of data, resulting after several iterations in the foie gras of the deep learning revolution.
Agents are different to generative art. Not only this but adaptive behavioral agents are distinguished from non-adaptive because their behavior evolves over time.
How a shape of behavior emerges from randomness:
randomness ( morphogenesis) transforms over time (metamorphosis), or remains stable (morphostasis).
In a nutshell the gigabytes of data, you feed the agent start off random and finish stable.
Approach
Throughout the paper its clear that Audry wants to distance herself from generative art, non-adaptive agents and algorithmic art. She uses LSTM recurrent neural network to train a series of artworks. These agents are both predictive and generative.
They are used for their predictability in translation for example.
“A unique feature of deep-learning systems is that they improve iteratively.”
They start fro nothing, and as they get exposed more and more to data they become better at prediction.
generation and adaptation
The ability of an agent to be both a reader and a writer.
Audry is also concerned about the physicality of her book and how it lends materiality to the agent and confers an identity beyond its abstract virtual existence.
Prepossessig
Wither Heights contains a few more than 600,000 characters. Which is small compared to the modern language modeling datasets that contain several million characters. (THIS IS CRAZY)
I had no idea making the letters lower case could help an agent. Its also best to remove low-frequency characters like parenthesis or slash, witch confused the agent.
Training
From randomness to 150 iterations.
A probability distribution is represented by a function that produces one probability value for each possible characters
It uses probability to generate sequences and then learns of these sequences by iterating them again and again.
This is called the Markovian process and its very common in natural language processing. Audry goes into a lot of maths at this point that I think I need to read again to fully understand but I managed to follow the process of probability and its limitations.
It was really interesting to find out about the limitations of an LSTM agents and to be honest, it makes sense.
One of its limitations is that it makes the assumption that the closes elements in the past are the most important for predicting the future, which is an imperfect premise.
This an issue because the neural network fails to grasp long term dependencies between sentences aka it's going to still sound/feel like a computer.
Temperature Adjustment
Like every algorithm, you have to adjust some parameters. In this case Audry had to adjust some values to make the probability distribution more “peaky”.
Shortlist
After learning step by step how the process of using LSTM neural networks works, Ive noticed there is a lot of adjusting and testing. Find this super interesting because it just confirms how you can never be sure how the network and agents are going to behave.
Transition between Models
Interpolation is such a big part of ML, I was reading a paper about GANs and they use different types of interpolation to learn off datasets of images. Im super excited to learn it's used in the text as well.
Results
Audry is gain got to describe the progress of the agent as it runs through the reading in terms of time. She also explains show ′time′ is understood in terms of the character position.
Morphogenesis
My new favorite word.
It sort of means the concept of an agent becoming more and more familiar and comfortable with the world it lives in aka as Audry puts it - the text it reads.
It is fascinating to see and read in the following pages how these random letters slowly become interpreted and distributed according to rules.
The reason why it became obsesses with white spaces and frequent characters (such as e, t, a)is because of the probabilistic approach: these letters simply have a higher probability.
Conclusion/My thoughts
I really enjoyed this paper and it made me think about process and machine learning for the first time. It was fascinating to see how the model developed, made mistakes and evolved. This paper has gotten my enthusiasm for running.
However, there was one thing about Audry's conclusion that made me second guess the article. I understand that the artworks are the process of the LSTMtrying to imitate a famous English writer, and NOT to successfully make an AI produce a novel.
This is all right I guess for the purpose of research however I disagree with the aim because the bar has been set kind of low. Thought the conclusion fell flat. Why not attempt to make and ai that DOES recreate novels, funny novels, abstract novels.
It's just a thought.
Week 06:
Winnie Soon
Talk at Goldsmiths
Machine Learing, Flobbers and Weibo.
The talk by Winnie Soon was a fantastic dive into the world of Machine Learning as both an artistic medium and a research tool.
I was really inspired by her thought process on how to explore the world of machine learning and give it a creative purpose. I personally am very intrigued bt ML and admire all the people that take part in its development, so as an artist Winnie Soon definitely inspired me.
It was also great to see (as always) a powerful female voice in the world of Artificial intelligence and how artists are positioning themselves with hit new medium.
In particular, her recent prototype system which translated Weibo platform was very intriguing. I found it fascinating firstly because Chinese characters are so weird to me. Moreover just to think about the extent and ambition of the task to translate . or interpret a whole website is very outstanding.
When she passed around her latest build it was really cool to see that progression and semantic transition from character to letter as the pages progressed. So far I have been very and almost only interested in the visual/humanistic side of ML and learning algorithms. so Winnie's work inspired me to look a bit beyond that and think about what other datasets are there, What about sounds or emotions?
Although Soon's process was fascinating and a work of art on its own, I thought her outcomes were perhaps a bit underwhelming. In a way I was more interested in what went behind the scenes than what was the finished product, it was a bit too serious for my taste.
This aside I thought it was very interesting when she started talking about the Flobber and its philosophical significance. I thought it was really clever and such an interesting and abstract way of looking at the time and how it develops. In a way, we are constantly surrounded by flobbers.
Overall this talk was super inspiring, I am looking forward to implementing what I learnt today on my end of the term research project. Looking forward to exploring different ML can be used as a medium to create not only art but a research practice.
Week 08:
Computation in the Aesthetic Practice of
Zach Blas
Even before the talk, I was pretty sure I was going to like Zack Blas. His artwork contains themes that I am very intrigued by such as dataism, social inequality, and the concept of being human in the age of technological revolution.
The Weaponization Suite is defiantly one of his most iconic works, it encompasses an array of manifestos that are becoming more and more real as the years go by. The level of facial recognition whether it's consensual (unlock your phone) or unavoidable (china AI cameras) is getting higher and higher.
This concept paired with the ′queerness′ aspect of how the mask was made adds a second layer of complexity to the project. It is through this visualization that the audience is being captivated and intrigued.
If I have to be honest, I was very surprised when I first saw the mask and then when the real meaning was being explained of the scanning o different queer men's faces. I thought it was such a brilliant way to add to the manifesto of technology.
My background is in 3D animation and one of the things that I was most intrigued about and we are actually exploring in our current research project is how data is being quantified. I found it really interesting the way he chose to quantify his data in these masks. Of course, the face is the primary way we can recognize people, through the use of generative processes every mask that he makes can be potentially different.
Moreover, when it comes to political repression, a mask is basically the most symbolic thing I can think of. Masks have been used for political manifestos since the start of time and they have such heavy baggage attached to them.
Week 11:
Hertz, Garnet "Disobedient Electronics"
The Case of the Strangerationist: Re-interpreting Critical Technical Practice.
I was truly fascinated with what a disobedient object was. I had never heard the term before and or the first time in a really long time it made me think long and hard about my own practice and how I've been wanting to explore disobedient objects without even knowing they existed.
I focus a lot on my artistic practice in feminist issues in Latin America. Most of my pieces trick the audience into believing they are going to experience something fun, Latin and colorful, whereas inside of the color lies a very serious message.
After reading Hertz my pupils just dilated as I explored artwork after artwork each more impressive and meaningful than the next. I was really inspired by the Abortion Drone and Transparent Grenade. Each object, very different from each other showed defiance and power in a way I had not seen before. It's gotten my cogs rolling and thinking about how I can make my own disobedient object for my final piece.
When asked to come up with a definition for a disobedient object, we defined it as:
"An object who′s purpose has been shifted to carry out a manifesto"
We thought it was important to include two words: purpose and manifesto. In the examples we read about, all the objects had been intervened or evolved to encompass a and deliver a specific manifesto. We thought that was what made the object so valuable and created the feeling of disobedience.
Later on, in the group project, we were asked to think of our own disobedient object and expand on the thought process. The subject that we chose was bullying.
We thought to create a device that a child could press and alert the teacher that a child was being bullied. Soon after we presented we came to the realization that our object was not very disobedient. Therefore here are a few disobedient iterations:
When the child presses the object:
- it disables bullies social media for 1 hour.
- the bully gets salad for lunch
- the bully gets extra homework to do with bullyng.
Week 12:
Some thoughts on Black Feminism from the perspective of a Latino Feminist.
plus
Awesome lecture/talk by Clareese Hill
I found this week's reading quite interesting and political in terms of identity and nationality. This subject is one of intense debate and discourse and I am very happy that it has been mentioned and talked about in class.
In the paper by Safiya Umoja Noble "A future for intersectional Black Feminist Technology Studies" she immediately makes clear that the issue is very much rooted in the "struggle to recognize multiple interlocking systems of oppression" that have been going on for roughly 40 years. She also mentions and urges women to take action and participate in the #BlackLivesMatter which was "founded by three queer women" (such a cool fact!)
One of the things I found super eye-opening was her mention of the internet and how it's vastly influenced by white western internet. For years I have been using this tool and venerating it. I have watched its adds, commercial, gimmicky videos and fed on everything it had to show me. After reading this paper my notion of the internet has completely changed and I am now more aware than ever on how we act as unwilling consumers to an agenda that has nothing to do with my beliefs, race or values.
"[T]he Western Internet, as a social structure, represents and maintains White, masculine, bourgeois, heterosexual and Christian culture through its content." This is something that I had never really thought about and it struck me that never had. What Brock mentions all of these users interpreting racial dynamics through the electronic medium of the internet and by doing so "redistributes cultural resources along racial lines"
-
Black women receive 65% of new AIDS diagnoses
-
Single African American women have a median wealth of $100
-
African American women with children have zero median wealth
-
The poverty rate of African American lesbian couples is 21.1 percent versus 4.3 percent for White lesbian couples
-
African American women are three times more likely than white women to be incarcerated. According to the American Civil Liberties Union, Latinas and African American women are disproportionately affected both by crime (since they are more likely to be victimized) and by incarceration, especially those who are primary caregivers for their children.[16]
I cannot finish his reflection without mentioning the above stats by the Center of American Progress about Black women live in the US. We call ourselves progressive, we call ourselves kind but there is no doubt an underlying problem in the fact that as a society directly or indirectly we have not let go of the racist gene. I have read these stats over and over again getting shocked each time.
Even though this paper is focused a lot on black American women, I cannot help but think about my own situation coming from a Latin American country. Apart from the obvious fact that we are also underrepresented in the media Latino women to have been an internet exoticization and even subject of mockery in American TVs such as Family Guy or South Park. The worst part is that a lot of the times the women who are being used as stereotypes will never be aware of this because 1) they have no access to American TV and 2) they have very little access to smart technology and the internet.
This paper has really helped me trigger different feelings that I have always thought but never really seen represented or talked about in an academic paper. I am looking forward to using technology to explore these themes further and tie them to my own cultural background.
This is why it was so inspiring to be in the talk with Clareese Hill.
It was really interesting to hear her talking about her work and her approach to VR involving black feminist discourse. Coming from a VR background we get a lot of VR briefs that ask for the technology for the technology's sake. A lot of the times we end up creating ’cool’ looking work that has no meaning.
This is one of the reasons I got really excited when Clareese started talking about the use of data and how necessary and powerful it is for underrepresented communities. She is right when she says that technology is oppen to everyone it is oppen source.
Week 13:
Digital Ethnography readings
A chat with Pete and his worlds and characters
I was looking forward to this session because of past experiences in the participation of digital ethnography. I used to freelance for a company called Brand Genetics where they contacted me because of their interest in exploring 360 videos as a tool to gather information about single couples.
Their aim was to use 360 and VR for the researches and the subjects in order to gather more realistic and organic research. The particular subject that I was looking into was single people living in London and what they kept in their pantry. We gave each participant a 360 theta camera which they kept for one week and gave us back video blogs of themselves commenting on their diets, how much food they bought, where they bought it from and how they deal with waste.
I was in charge of the technological side in briefing everybody with the cameras, how to operate and save and best practices for 360 filmings. It was really fun to be a part of this research and being asked to use this technology to help extend the possibilities.
This, however, was almost three years ago and a lot has changed since then. For one, 360 has almost died out and people are already using VR less and less commercially. I know at least two companies that have struggled with the shift, one of them having to close.
After having read Deborah Levitt′s "Understanding the implications of radically new experiences" it was pretty clear to me that there is a lot more to VR than just watching pretty pictures or being told stories.
She makes very valid points in each of her sections. I particularly liked the part where she explains why VR could never ben an empathy machine, I have heard this discourse many times and have had debates with colleagues about it. This aspect tied to her explanation of “Homuncular flexibility” really created a rounded up argument to really question to what extent can we rely on VR to give us accurate representations of anything!!
Her mention of the virtual lobster experiment made me giggle in awe.
Flash forward to the next day. I was really excited to hear Petes talk as I had already done a bit of digging around about his work. I immediately felt identified with his use of color and how that projects on to his own persona.
While he was talking he had the pleasure to show us some of his work. In particular his VR Game/Experience. I really enjoyed listening to him talk about technical and artistic aspects in his work and, at the same time, how he used the VR world very much as a medium to blend our own realities.
VR definitely has the power to immerse itself even though it's lacking in the empathy department. Pete definitely took advantage of that and showed us an intense world with lots to see and explore. While he was explaining the gameplay he started to talk about different characters that he appropriated from all over the internet and gaming world.
Right from the start, he was very clear on how he had approached different gaming groups and immerse himself in that culture. It was really important for him to learn about these groups to accurately personify these characters and further enhance the world he was creating.
When talking about these characters he mentioned the issue of male and female character s and how he wanted to distance himself from that. He spoke a lot about the nonbinary aspect of them and how he had been inspired by ancient Japanese dress and culture and used this to influence his work.
This really resonated with me as a Latino because we too have a lot of history in our clothing and different attires mean different things. There is definitely a lot more to talk about just on the aspect of characters in VR and their clothing. I think I'm going to do a little bit more of introspective thinking about exactly how Latino dress attire has shifter during the years and the hypersexualization of the Latino woman and the way of dressing and expressing herself.
Week 14:
Digital Narrative and Witnessing Artworks:
Clouds over Sidra by Gabo Arora and Chris Milk in partnership with the United Nations and Samsung, 2015.
I have been following Chris Milks's work since my BA days back in 2017. At this time 360 filmmaking was at its peak and mass-produced 360 content started to make it mainstream. Milks' way of portraying a narrative where the viewer was in control of the view was especially interesting.
By the use of sounds, lighting, and action Clouds over Sidra transports you to a conflict in which you have to take part in as soon as you decided to put on the headset. Through the use of smooth cuts and gentle trajectories, it is inevitable to feel mesmerized and captivated by your surroundings. a
You have now been shifted from a spectator to a witness. This change is one of the reasons I think 360 narratives can create emotions. There is something very interesting about POV action and the feeling of ’being’ in a digital environment which, in most cases, is vastly different from where you probably are physical.
As a witness, the experience that the digital narrative has on you is very different because now, you have no choice but to be there. With human rights topis especially to do with the Syrian Refugee crisis I understand why a medium like VR could be seen as effective.
"We are shown the emaciated child crying at the pain of hunger, or the harsh realities of life in a desert refugee camp, to provoke a response from us at an emotional level"
Nishat Awan: Digital Narratives and Witnessing The Ethics of Engaging with Places at a Distance
Nevertheless, it is very interesting to acknowledge that this technology may also be acting with a digital savior complex. It is very easy to come to a convention in Switzerland as part of a wealthy corporation, put on a VR headset, experience how dreadfull life is for poor Sydra and then hop back on your private plane home.
There is something very much not OK with that. Does it matter that that corporation is then going to help?
it is Sydra "who has to show us her destitution and her will in the face of it; she has to perform it."
Liquid Traces - Left to Die, Directed by Charles Heller and Lorenzo Pezzani, 2014.
Liquid Traces is a piece that left me with tears in my eyes. It was so so sad and so emotional to hear about the political disruption and protest driven conflict. The fact that NATO had so much data, about tracking systems, radar information, satellite imagery, and electromagnetic waves.
This piece showed shows just how shameful humans can be. The more I watched it the more I felt disgusted. I felt like they were describing all of the worst qualities in humans. The NATO countries took no responsibility as countries of asylum and as a consequence, only 9 people survived.
It shows how politics can murder.
"We can talk as much as we want about human rights and the importance of complying with international obligations, but if at the same time we just leave people to die – perhaps because we don't know their identity or because they come from Africa – it exposes how meaningless those words are," said Strik, a Dutch member of the council's committee on migration, refugees and displaced persons, and the special rapporteur charged with investigating the case.
Week 15:
Computational Art and Multispecies Storytelling.
"sf is a sign for science fiction, speculative feminism, science fantasy,
speculative fabulation, science fact, and also, string figures"
The topic of Multispecies Storytelling was very interesting for me. Again, just like the disobedience object. it was the first time that I had encountered this term, I felt it was almost philosophical.
The video that went through the thought experiment of looking at a cow as technology really enlightened my mind. I thought the most interesting thing about this was to look at a biological process as a technological one.
This ties very nicely with String Figures and the role of pigeons in Terrrapolis. I found it baffling and almost surreal how Harroway dove into the subject of pigeons as this almost otherwordly source of technology and tied it so majestically to so many aspects of human life. From being seen as vermin to being the most adaptable bird on this planet, pigeons have traveled with humans and adapted everywhere in the world.
At one point I remember thinking it was so interesting how these birds alongside the cow have essentially changed the way humans interact, the way laws are constructed and how our cities and systems work.
I'm not too sure if multispecies is something that I am massively interested in at the moment. I still have my heart set on exploring more Machine Learning and the human aspect of it.
I still think that multispecies is a very interesting subject more so because I feel as humans, especially in computational arts and research we get too caught up in solely the human aspect of technology. It was a real breath of fresh air when I was reading about how broad technology can be to the point that it can include multispecies analysis.
Finally, I found this really interesting paper by S. EBEN KIRKSEY from City University of New York Graduate Center and STEFAN HELMREICH from MIT about multispecies ethnography and " how multispecies ethnography centers on how a multitude of organisms’ livelihoods shape and are shaped by political, economic, and cultural forces"
This again blew my mind away because it is such an interesting way to look a system where that is cultural, political or artistic and in a way take away the human ego and put humans and other species at the same level fo study.
"The work of Donna Haraway provides one key starting point for the “species turn” in anthropology: “If we appreciate the foolishness of human exceptionalism,” she writes in When Species Meet, “then we know that becoming is always becoming within a contact zone where the outcome, where who is in the world, is at stake” (2008:244)."
Article can be found at: https://anthropology.mit.edu/sites/default/files/documents/helmreich_multispecies_ethnography.pdf
