Spreadsheet of raw outputs from IBM Watson Tone Analyzer analysis of each blog post from my blog writing in August 2012 at Princeton Theological Seminary. Watson breaks down writing into three attributes: emotion, language style, and social tendency. Expand the table to view the entire output.
“No one has ever written, painted, sculpted, modeled, built, or invented except literally to get out of hell.”
- Antonin Artaud
So far in the proceedings of this Medium page, I have written out what was troubling me regarding the end of my relationship with my wife and how the recent fragmentation with the relationship with Yana brought back many of those old memories. I have avoided thinking about “being productive” and “moving on” because I wanted to learn what was perhaps a source of why these relationships did not work out. For a couple months, I’d write poems in the place I was at and try to capture how that environment made me feel with words as well as place into language the sentiments and feelings that came with that place. The Artaud quote above is a favorite, as it simply states the reasons why we create anything as people. What I have been driven to in conversation after conversation recently is how much in common we all have with each other. We all suffer, and what binds us together is the language we use to relate. Language broadly defined. What you say and write, as well as the clothes you wear, the body language you use, and even the perfume or deodorant you put on.
How we attract or repel each other is very similar to the way that atoms make and/or break bonds. As a chemist, I see it that way. For eight years, I have wandered in my life. Mourning the loss of the ability to do academic chemistry at a high level, I entered years of rollercoaster ups and downs. All the while in this walk though I was learning something. I gathered a lot of first hand experiences about how the “real world” works outside the confines of elite academics and learned a great deal about how people like to engage each other and higher and lower things. A common theme that appears over and over is that art and creativity, individuality and personal expression are very important to most people. I started finding a lot of hope and guidance in poetry and music. Visual arts and dance too, all formed central ways that I started to relate to people in new ways. I started a fairly regular habit of blogging several times a week and that eventually became a place where I would share poems and other ideas that came to me.
About a year ago, I started to pay attention to the release of commercial artificial intelligence software (AI). I imagined that it might be an interesting tool to use to merge the two sides of my personality, artist and scientist. I started wondering if I could see trends in the nature of my writing from my blog during particular phases and times. In a broader sense, I wonder if AI as applied to language analysis can be used by people to detect trends in mood and mental state that might go undetected due to lack of resources for a formal therapist or psychoanalyst. In just the brief analysis of my first month’s posts from my blog in seminary I noticed that anger and fear are the predominate emotions in the writing. It is a deep interest of mine to continue to analyze the posts of my blog and compare the outputs on IBM Watson to the writings of other poets and writers. I’d be curious to see if there are trends in writers who have died via suicide to those who lived full lives to an old age. In this investigation, the confessional poets of the 50’s and 60’s are of great interest. Furthermore, I am interested in comparisons to artists of other racial groups and cultures. Trap music lyrics immediately come to mind.
I am thinking of turning this Medium into a data science and poetry blog as a prelude to making tools, apps and websites, for teaching people how to write in a therapeutic way and eventually invite them into a conversation. It’s an “interesting” (terrifying) brave new world that we face where AI’s might be used to psychologically profile people electronically. Many people at places like MIT seem to love the idea, see this recent Technology Review article. A friend from Atlanta who has a brother with schizophrenia recently started NeuroLex.
His startup uses voice samples to “diagnose” states such as psychosis. He is on the right track. However, my approach is different than his. He desires to install machine learning/AI voice analysis devices into mental health care facilities. My desire is to give people the tools to avoid entering into these mental health care facilities in the first place. My greatest fear of AI is not of the technology itself, it is of it’s ability to oppress and control people who suffer from conditions that might appear to an AI as “mentally ill.” I “came out of the closet” pretty much immediately after I was diagnosed with bipolar disorder I in 2010 at MIT and have learned a great deal about how people with mental illnesses are viewed, and feared. Jim is doing a good thing, but it’s opposite in the way it should be done. Where is his brother in all this, in the video why is not his brother there? Anyhow, next post will be data visualizations of the data in the spreadsheet along with more analysis of poems from me and others.
Gotta go! Peace 🙂
Just for fun, an IBM Watson Tone Analyzer of the text of this post:
Look, JOY! 🙂 ❤ Clearly, I’m onto something in terms of a good path in life with this. It’s fun to me!
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.