Computer Science Subject Guide — Algorithm Studies, Ethics, and AI

Simply speaking, an algorithm is a set of instructions that are given to a computer in order to complete a given task.  Normally, these instructions include very specific calculations and conditionals that are given to the computer in order to calculate the output.   For example, IF X = 0, THEN (calculation), to receive the desired output.  These instructions can be represented visually via a flowchart as seen below:

Here is a video from TedEx, which explains what an algorithm is in a very simple way.

The video below from Wired is fun because David J. Malan, Professor of Computer Science at Harvard University, explains what an algorithm is at 5 different levels to 5 different people; a child, a teen, a college student, a grad student, and an expert.

At this point, you can go onto the next example of a natural language processing algorithm called Word2Vec.  If you want to learn more about algorithms from a computer science and computational thinking perspective, then look at the book linked below:

Word2Vec is a natural language processing algorithm that was created in 2013 . It stands for "word" plus two "vectors." What is a vector? A vector can have different meanings in mathematics versus computer science. The Oxford Dictionary of Computer Science defines a vector as,  "a one dimensional array."  An array is defined by the same dictionary as, " An ordered collection of elements of the same type, the number of elements being fixed unless the array is flexible." A one dimensional array is called a vector, and a two dimensional array is called a matrix.

In the Data Structures Wiki, there is a simple example of a vector using a snippet of the opening sentence from Charles Dickens' A Tale of Two Cities: "It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, "

Here is that phrase represented numerically as a vector:

Now that we have looked at vectors, let's examine a natural language processing algorithm called Word2Vec that emerged in 2013:

Algorithms can become problematic when they do not work equally for everyone in society.  For example, if a facial recognition tool only works effectively on Caucasian skin, there is a problem with the algorithm.  Subsequently, in examples of companies like Clearview AI, which is a tool used by some police departments for facial recognition, one can discuss, what are the ethics of scraping the images off the internet of people who perhaps did not consent for their data to be used to train AI.

If you click on the link below, you will see a video from 2009, which is a non-critical look at facial recognition technology.  How has our perception of facial recognition changed since this date?

Cite: Loh Kok Hong, & Ochre Pictures (Producers), & Lay, H. Y. (Director). (2009). Facial Recognition. [Video/DVD] Looking Glass International. Retrieved from https://video.alexanderstreet.com/watch/facial-recognition-2

Although artifical intelligence has existed since the 1940s with the emergence of new methods of computation by Alan Turing, it has never been more omnipresent and ubiquitous in our lives than today. Alongside the concept of AI, you may see words like machine learning, deep learning, natural language processing, and neural networks.  These terms all indicates methologies used with the large field of AI.  

Types of AI:

Here is a funny comic about machine learning:

to cite the above video: Neil Docherty, David Fanning, Five O'Clock Films, & Frontline Films (Producers), & Docherty, N. and Fanning, D. (Directors). (2019). In the Age of AI. [Video/DVD] Public Broadcasting Service. Retrieved from https://video.alexanderstreet.com/watch/in-the-age-of-ai

You are likely already familiar with generative AI tools like Chat GPT, and DALL-E, or AI face editing apps on your phone like Gradient and FaceApp. So what is generative AI? Generative AI uses large scale machine learning data sets to create new text, images, sounds, or video by responding to a prompt.  The University of Pittsburgh's Center for Teaching & Learning defines generative AI as follows:

Generative artificial intelligence (AI) tools use machine learning models trained on massive pools of information to learn patterns from data to create novel content like text, images, audio, or video in response to a prompt. Unlike internet searches, generative AI tools do not use algorithms to locate and curate existing sources. Instead, they create new content by predicting what word, sound, or pixel would come next in a pattern.

As you can imagine, these tools can be really useful, exciting, and interesting while also raising many questions, and sometimes concerns. 

Here is a talk by Dr. Mireille Plata of the University of Edinburgh, on the topic of generative AI:

There are new tools constantly emerging, which can help the academic research tool.  Here is a slide from Engineering Librarian Jason Cole of Kansas State University, which introduces some of the AI research tools:

Many questions of copyright have arisen with the emergence of generative AI tools.  If an AI tool learns from existing images to create a new image, then how do you cite it?  Who owns the copyright?  This and many other questions are being discussed with regards to copyright, citation, and AI tools.

News of the first patent application naming AIs as inventors has already made headlines around the world (15) The first artificially intelligent inventor was named ‘Dabus’ and, according to its inventors, it ‘relies upon a system of many neural networks generating new ideas by altering their interconnections.’ A second system of neural networks then comes in and selects potential ideas based on novelty and salience. With this mechanism, Dabus came up with two inventions: ‘a new type of drinks container based on fractal geometry, and a device based on a flickering light for attracting attention during search and rescue operations.’ (16) While the European Patent Office rejected the application, stating that the inventor has to be a human being, IP officials in South Africa disagreed, and made history in a landmark decision that awarded a patent which names an AI entity as the inventor (17).   Page 86, Chapter 5, Going Digital, A Cultural History of Copyright

Have you ever thought about the role of Google in your daily life? Do you know why you get the results you get in a search engine?  Do you understand the role of advertising and popularity in clicks of the search results?  How is a library academic database different from a search engine like Google?

Ada Lovelace, the 19th Century mathematician, considered to be the first computer programmer, always preferred the merging of the arts and technology, with terms she called the "poetical" versus the "analyst."  As we look to the future, it seems wise to keep in mind how the arts and humanities remind us of the importance of humanity in the face of technology.

There are more curriculum which are being developed to teach literacy about new technologies.  Below is one that was shared with me by colleagues.