Artificial Intelligence Introduction
Artificial intelligence (AI) refers to a branch of computer science concerned with building machines and programs capable of performing tasks that mimic human activity and intelligence. Newer forms of AI are capable of learning the patterns and structure of training data (such as text and images) and then generating new data that has similar characteristics (Wikipedia). Already in wide use in areas such as visual arts, programming, information technology, help desks, and other business sectors, AI is certain to become a natural part of the workforce and daily living, and thus cannot be ignored. Naturally, there are many concerns about students using AI, with cheating and plagiarism among them. However, AI can be a useful tool that can be used as an assistant for many tasks once its capabilities, limitations, and programming are understood.
You should never use artificial intelligence technology to complete assignments without the explicit direction or full knowledge of your professor.
If you are required to use AI for any course, or if you choose to use it for any reason, it is your responsibility to use it ethically and with full knowledge of its advantages and limitations. This page will help you understand and work appropriately with AI.
Page Contents:
How Does Artificial Intelligence Work?
The developers of AI tell us that tools such as ChatGPT and Bard, also known as "large language models" or LLMs, work by predicting the flow of language. AI uses publicly available sources to "recognize the relationships that most commonly exist between individual units of meaning (including full or partial words, phrases and sentences)" (Hoffman, 2023). According to Chat GPT itself (2023, August), this type of artificial intelligence
...is designed to generate human-like text based on the input it receives. For this to happen, the model is trained on a massive amount of text data from the internet. It learns to predict the next word in a sentence, which helps it capture grammar, syntax, and some level of world knowledge. The result is a 'pre-trained' model with knowledge of language patterns and facts. After pre-training, the model is fine-tuned on specific tasks using more narrow and focused datasets. This process tailors the model's behavior to perform certain tasks, like language translation, text completion, or question answering. The GPT architecture employs attention mechanisms that allow it to weigh the importance of different words in a sentence and generate coherent and contextually relevant responses. The large number of parameters (weights) in the model, often in the tens or hundreds of billions, enables it to generalize well to various tasks and generate text that is often remarkably coherent and contextually appropriate.
ChatGPT was trained on a large body of material; as it was developed, humans interacted with the model, ranking answers so that it could learn which were better responses (Northwestern University, 2023).
What Artificial Intelligence CAN Do
Artificial Intelligence has the potential to help you get inspired and improve your work, so long as you do not use it as a replacement for you own thoughts, ideas, and writing. Some ethical ways that artificial intelligence can be used to inspire your work include:
- Suggestions for project topics.
- Keywords for library database searching.
- Editing your work for spelling and grammar.
- Synthesizing and arranging knowledge, so that basic facts can be learned and checked in one place instead of many.
- Coding, providing code snippets, explanations of programming concepts, and debugging assistance.
Naturally, any material generated by any form of AI should be proofread, personalized, improved, and appropriately cited.
If you want to record meetings, there are AI tools that can take notes and provide summaries, such as Zoom's AI Assistant and Microsoft's Co-Pilot. That these tools may not be available unless you are the meeting leader; if you would like summaries of class sessions, you should ask your professor if recording is possible.
What Artificial Intelligence CANNOT Do
Despite its amazing capabilities, there are things that AI cannot do. Understanding its limitations may help you work confidently with artificial intelligence. Here are examples of what AI cannot do:
- Access Paid Content: Just like Google, AI finds content that is popular and paid by advertising first, and it generally cannot access material behind firewalls, thus excluding premium library material. Artificial intelligence can search abstracts and summaries on Google Scholar and publisher's websites, but that type of analysis is shallow and misses critical parts of the material.
- Mine the Most Current Internet Content: At this time, AI is a few years behind in the content it can mine. Therefore it is unable to access the most current scholarship on any topic.
- Thoroughly Analyze Newer or Under-Represented Topics: Topics or subjects with minimal content on the Internet will not produce good AI output. Additionally, newer concepts and theories typically do not appear in peer-reviewed academic publications for a year or more, due to the time it takes to write and publish scholarly material.
- Analyze Certain Sensitive Topics: Some AI models like ChatGPT are programmed to avoid discussion of what they classify as "harmful ideologies." Questions that ask AI to consider problematic concepts, no matter how carefully they are phrased, may be met with a message to the effect of, "I'm sorry, but I cannot engage in discussions that promote or glorify harmful ideologies." When asked "What harmful ideologies will you refuse to discuss?" ChatGPT (2023, August) responded that topics it will not engage include but are not limited to:
- Hate Speech
- Extremism
- Misinformation
- Harassment and Bullying
- Illegal Activities
- Self Harm or Suicide
- Adult Content
If you do not search the library's databases when working on a research project, your final product will be lacking in critical information that is not available on the open web. This could lead your professor to suspect that you did not complete your own work.
Problems with Artificial Intelligence
New technology brings with it concerns about accuracy and ethics. Some information fed to AI as training material was poorly written, inaccurate, biased, or even violent. While companies have tried to cleanse their AI models of the worst content, they have not been able to remove it all. In addition, like any technology, artificial intelligence is subject to inconsistencies, misinformation, and technical issues including:
- Hallucinations: An overload of input or a lack of data to mine can result in AI generating nonsensical, incorrect, or made up information.
- Being Flat-Out Wrong: AI can re-generate whatever bad information it mines, or conflate closely related ideas that are frequently mentioned together.
- Theft of Intellectual Property: AI programs mine material taken from online sources not located behind a paywall without the permission of the authors/creators and thus the output, particularly when unattributed, may constitute intellectual property theft. Further, a number of well known authors and other entertainers have filed lawsuits against OpenAI (the creator of ChatGPT) for violating copyright by using their protected works to train ChatGPT, which in turn allows the creation of “ 'derivative works' that can mimic and summarize the authors’ books, potentially harming the market for authors’ work" without asking for permission or compensating the creators (New York Times, September 20).
- Propaganda: AI output is subject to be tainted by misinformation, malinformation, and propaganda, such as doctored images and "deepfake" videos. For example, until at least 2018, there was a website with "martinlutherking" as part of the URL that was managed by neo-Nazi propagandists. There is currently an active organization that promotes "conversion therapy" (the practice of trying to turn a homosexual person into a heterosexual person), and though this might seem easy to identify as harmful pseudo-science, its website and documentation follow the standards laid out by the American Psychological Association, making it extremely difficult for a novice to distinguish their output from results of studies based on true scientific method. Mistaken, nefarious, and harmful information such as that in these examples can be mined and repeated by AI in a way that sounds authoritative.
- Lack of Transparency: Because the developers of AI do not disclose their algorithms, it is impossible for researchers to analyze output for inherent bias and accuracy of source material. ChatGPT (2023, August) had this to say about its own transparency: "ChatGPT does not have access to its training data, and it cannot provide specific details about the sources used during its training. The model was trained on a mixture of licensed data, data created by human trainers, and publicly available data. OpenAI, the organization behind ChatGPT, has not publicly disclosed the specifics of the training duration, the individual datasets used, or the proportion of the data that comes from different sources. Therefore, ChatGPT cannot provide transparency regarding its sources in terms of specific documents or data sets."
- Reliability Checks: AI currently offers no reliability predictors or crowdsourced fact checking such as is the case with Wikipedia.
- Security: Like any other technology, artificial intelligence can be susceptible to malware, spam, phishing, and other forms of privacy breaches and data theft. For more see Artificial Intelligence and Personal Security.
Beyond that, AI has created other knowledge divides that contribute to the phenomenon of information privilege:
- Equity, or who can use these tools most effectively when use is appropriate, or without being detected when use is in appropriate or disallowed. Who can afford to pay for the tools when they become monetized?
- Data Quality and Accuracy: AI will provide data that it can find by scouring the free Internet; where one can find data sets and statistics that are incomplete, incorrect, or subject to misinterpretation, and these errors may be repeated in an AI inquiry. Per ChatGPT: "Inequities in data collection can lead to underrepresentation or misrepresentation of certain groups, making it difficult for the models to provide fair and unbiased information or support for those groups" (ChatGPT, September 2023).
- Algorithmic Bias, which favors patterns of the user’s own searches, paid content, and inherent bias in the content it mines.
- Language and Cultural Bias: AI tends to be biased toward the dominant languages and cultures that are prevalent in the data used to train the program. This can result in less accurate and less comprehensive responses for users from marginalized linguistic or cultural backgrounds (ChatGPT, August 2023).