Ok, this is terrible environmentally. We have found yet another way to drain away our natural resources. This is infuriating. Instead of putting so much and effort into a non-essential, harmful technology, how about we try to solve the problems that are impacting basic quality of life for millions rather than contributing to the problems.

What exactly are these workers doing to maintain these systems?

This really confused me. What is this trying to say? Why describe horrific scenarios?

This scary. As someone who wants go into journalism and writing, the idea that ideas and writing can be used without my consent by this type of technology is alarming. It is one thing to analyze that which is already written and collect data, but I do not like that programs are being trained to synthesize and replicate writing to create its own material.

This seems like the same style of model Voyant uses to detect word count frequency in a corpus. It's cool how it uses word choice frequency to predict sentence structure based off basic statistics. It is now starting to make sense why Voyant cares so much about measuring word choice frequency!

Umm...

Ok, this is pretty cool.

Ok. This makes sense. It identifies patterns and recreates patterns rather than copying the exact contents of something.

Go e are the days of trying to figure out what to eat for dinner and spending all day thinking about it..using data and AI hypothetically you could list the ingredients in your fridge and ask it to make a meal with a salad and main. How amazing! Has anyone given this a try before? I think I might have to.

Sure, the algorithm behind AI I suppose is basic in its intrinsic form but the actual development of AI tools and applications would indeed come with complexities no? I think the author wants us to see the basic forms of how AI was discovered but its integration with software and the various hurdles that brings can indeed be complex in its own right. For example, the development of AI tools like chatGPT obviously requires a talented team of engineering and software professionals to figure things out and probably utilize methods that are extremely complicated to manage.

When training a model, how do we establish what we are looking for to the AI? Sure we are giving it vast amounts of data to find specific patterns in the vector-space but how do we train it to stick to a particular path? There's this concept called objective functions which basically refers to the general optimization of an AI model why was this left out? https://www.larksuite.com/en_us/topics/ai-glossary/objective-function

Any time that people are afraid of something new or dismiss it for its problems they risk missing out on an opportunity to use a tool or allowing the tool to be used poorly. As soon as someone tries to scare you away from exploring something, that should be a signal to immediately learn everything you can I would think.

This is good. If this technology can help place data within the context of a much larger grouping of related data I think could be very beneficial for understanding large-scale themes and overarching relationships between disciplines.

The term definitely seems coined in order to instil fear, and terminology like this should always be treated with caution. It reminds me of when people talk about “big pharma” as a money hungry business out for everyone. It is not as scary as it seems though, and after a quick search big data simply refers to data sets that are too large or complex to be dealt with by traditional data-processing application software.

Distant reading is starting to make a lot more sense to me. Although, I feel to some extent the objective is to be able to read more without actually reading the context. Probably for some it means working smarter not harder. I wonder if this could potentially miss important pieces to the reading.

I totally agree that a lot of people that utilize language/text models or AI in their professional lives throw the word "big data" around quite frequently. Does it just refer to the general computational handling of large data sets analogous to distant reading? or could it imply something deeper; not only handling large amounts of work but also finding a specific pattern or objective within those vast datasets?

This is fascinating. I am very interested to learn about the use of words over time, changing connotations, and the contexts in which they were used. This is such a cool way of looking at how difference people and issues and people have been recognized by government over the years. Names are so important and I wonder if some words have seen an increase in use as marginalized groups reclaim previously derogatory terms for their own.

It is still so easy to get lost in a jumble of information. I have only just discovered Ctrl + f this academic year. I do not know how I managed before I found that shortcut. It can allowed me to quickly scan long, dry academic articles when researching for papers to see if the information is relevant to my topic. This has also saved me when I know I've read something I want to quote but cannot remember which of my 27 tabs it was in.

This is really interesting! Honestly, as a student I wasn't aware of this software prior to last semester. I feel as though I was living under a rock. A student was caught plagiarizing on their exam and their answers were clearly pulled from using this tool.

At first I thought Distant Reading included the concept of almost reading between the lines or having a view from afar, like reading something and having a complety different understanding. I'm interested to know how to better understand how to analyze larger amounts of data pulled from distant reading.

I was curious as to why the word "Indians" was nearly removed from vocabulary. To find out, it was very similar to assimilation efforts completed in Canada. Teddy Roosevelt took land away from rightful owners and wanted American Natives to be otherwise known as "white man"

Analyzing large bodies of texts can indeed be useful in the digital humanities space. However, I think it limits us to the variety of works out there that may not be all text but rather audio. This video discusses how language models in the music realm can not only predict music recommendations based of lyrical content. It for sure uses text analysis and mining tools but it can also rely on factors like sample spectrum analysis to analyze large bodies of music and to recommend music with similar frequencies and amplitudes in a given beat. pretty cool stuff!

https://www.youtube.com/watch?v=PFAu93xxoGA&ab_channel=EuroPythonConference

This defiantly has potential to become an influential tool in the music space. If large language/text models can analyze musical works and understand lyrical patterns a given genre of music uses, this defiantly can be incorporated into out daily lives or the current apps we use. For example, apple music has a feature which can recommend new music based off previous styles listened to in the past. I wonder if "distant reading" style of algorithms is used in this context? perhaps its a combination of many different language models in one tool? I'm sure apple has the resources for this regardless!

People can get really invested in the gaming scene and this is a common thing to see. I think when the the last of us part 2 came out a lot of the voicing cast received death threats from fans due to how the plot was designed in the sequel.

Distant reading is not solely a product of the digital age and exists in many forms

I really enjoy how this was written. I think a lot of people in this world overestimate what they actually know and the knowledge they have accessible to themselves.

I personally find it interesting that people are gatekeeping this type of advancement. I understand that people love to be in competition with one another, but this technology has great capabilities to help us look through various information quickly. Why should we gatekeep us a thing, big data or not?

I am interested in seeing how far we can ( if there truly is a limit) push the combination of computation onto social sciences. Will we one day have such a deep understanding of society and culture that we can boil every interaction down to a singular algorithm? Or is life and society too complex and random to have anything defined so precise? Only time will tell.

the idea that history is written by the the victors also applies to the publishing world.

I was wondering when this word was going to crop up in this DH course.

I had to look up overfitting and it wasn't quite what I thought. It is about when a model fits too closely to theoretical data. This can happen because too few data points are used and can further limit the model's ability to consider other data/situations. I thought this was interesting as it reminded me about those videos talking about how algorithms can be biased and I had never heard the term overfitting before.

With the availability of the internet and modern technology, distant reading can or has already become much easier to do and more utilized in order to study intellectual historty.

It will be interesting what the near future brings with how fast technology is advancing.

Of course with innovation comes back lash, it is within human nature to want to not have/want change.

Ie. Technology such as the newest phones and older generations not wanting to learn/ not knowing how to understand then.

We tend to pit every new idea to an assortment of way/methods/things we already know rather than exploring them for what they were thought to be made for. I think seeing that there was an inward debate on the subject wasn't much of surprise but rather a given! Knowing of critics such as Stephen Marche and Stanly Fish, it is easy to see way it was the way it is.

I do however wonder which other disciplines we could better connect to? And whether it would be a better use of our time just understanding distant reading at its surface level or to keep a comparative with others too? (I would say comparing to others may help in the overall scheme of things).

This is an interesting point that reminded me of occam's razor. It is normal to overcomplicate things when sometimes it is the simplest answer or method that is correct/best. Just because something is simple does not mean it is not a good method to use.

Historical recordings are exploding in the time of digital humanities, making it far easier to explore trends throughout history and better keep track of our history as a society.

Seems to me distant reading is the computer form of having grad students doing a bunch of grunt work. I think as the article is going to try to get at, distant reading has its own computational history and humanities also has it's own history of big data analysis.

This method seems to resonate a lot with me as numbers are generally something that stick with me. Understanding the importance of these numbers and knowing what the results allow for a better interpretation of the population based on a given sample size. Although I would disagree partially as I think algorithms give greatly insight into what is or may be needed to be understood. As well as to help automize the process of otherwise very strenuous and labor intensive duties. (Radway's seems to run fine without it being automized but a larger sample size may warrant it)

I think this is an important point that is often overlooked. There are many things that are observed without actually looking at the background or history of it. It is normal to judge based on what you know, but oftentimes we do not know the full background on what we are judging. Everything has a backstory but often considering "how it came to be" is neglected.

I think paragraph goes on to essentially say that the questions being asked were not well suited to this tool. There's a connection to earlier when they talk about the fact that computational methods could be introduced to literary studies with the view of filling a gap in knowledge. So it makes sense that initially these computer based methods weren't seen as effective because scholars didn't know the right questions to ask.

I think this is something people misinterpret about big data, as this article talks about in the beginning. It's the classic quantity vs quality argument. Again, about Paul Schacht's research he talks about analyzing Dicken's writings and using a big data method let him do that but analyzing more and more novels past the ones Dickens had wrote wouldn't have made his research better.

It's intriguing, I think this paragraph has summed up digital humanities in a way all those articles from week 1 weren't able to do for me. DH seems to be a constant balancing act (and then analysis) between the digitization of research and presentation and the flexible way humans have of thinking. Maybe this has been really obvious to everyone else but the topics we have covered in the last few weeks (digitizing physical literature and no big data research) have really highlighted that aspect of DH.

I have to admit, it is probably because I come form a science background, but I like this way of doing research, using numbers and data wherever possible. I don't have fond memories of trying to analyze books in high school for metaphors or other literary tools.

It's really interesting to read about the kind of questions being asked by these scholars. They are all quite creative questions and it makes me think about what I may what to do for my final project.

I think the Paul Schacht video for this week will have a lot of similar opinions to this article, just based on this first bit of reading. Like he says in the video, the distant reading analysis helps him to answer his question but he still had to use his intuition to come up with the question and he has to interpret the results. The term Big Data can make it seem like a human element is lost but I think that is very much not the case.

this analogy, although seeming kinda silly actually helped me understand this bit

Is a video game with no video still a video game? How are digital humanities defined and who gets to define them?

These readings with links embedded within them provide great distractions :)

Is google also using distant reading, when you look up a word and google has a little graph of it's usage over time? Edit: oops, the paragraph below answers this question

I wonder if certain key historical figures who had a massive societal impact used specific strings of words, connections to other words and specific language that we can correlate to each other. If such connections exist, the applications could be endless in our society. Imagine having the cheat sheet to grow a social movement of any sort. This idea both excites and terrifies me at the same time.

It is increasingly more common to find these mediums of societal information available only in the digital space. To me, this means that there is not only an opportunity but rather a necessity for distant reading for those interested in pursuing the humanities of our own day and age as well as into the future of our society.

On more of a personal note, I find that the information to be garnered from distant reading can be much more interesting and insightful to more impactful subjects overall. One could read a singular novel and infer a few details about the author and what they are trying to convey or analyze several thousand novels from a specific time or culture and infer interesting notions such as the values of the time/culture and how they preferred to communicate.

Seeing the comparative of both older artist and newer artist is amazing but the red text highlighting that some rap artists actually surpassed Shakespeare in terms of unique words used in their first 35,000 lyrics was astonishing! A big one that did stick with me was how Nikki Minaj had more unique words said than Drake and seeing an overall trend of older artist having use more unique words more frequently than new artists do now. (Of course with some outliners, here and there). I really enjoyed seeing this example and how diverse the scope of Voyants uses can be!

Text mining seems to have a large role in todays society, from Risk management (Such as financial/insurance) to Cybercrime Prevention to better and more enhanced Customer Services. Services such as chatbots capturing data and extrapolating it to give results relevant to one's problems and help fasten up the processes. Looking forward to using Voyant Tools!

This could also be a way in which artists could strategize the use of specific words in songs to attract a larger audience. They could look at the similarities between top hits and find certain words that were used in all of them and then include it when advertising the music. For example, they could use it as a tag when posting on instagram or twitter and it may attract more attention.

Looking at the frequency in which words were used may be a good indicator of how history shaped literature. Certain words that were used more frequently may show what most of the population was feeling at that time or even what was going on. Words that fell out of use could also represent the end of certain period.