Disclaimer: I am not an AI expert, and do not know everything there is to know about AI. I am just a marketing writer wondering if generative AI is going to replace my job, a creative writer wondering how it will affect the publishing landscape, and a teaching assistant trying to help students navigate the challenges and opportunities presented by this tool.
The information below is what I presented to my class this month. I drew from several webinars and web conferences as well as thoughtful op-eds and newsletters, but I wished more than once that I knew of any existing resource that addressed all of the considerations listed here, from data sourcing to energy use. I am sharing my compiled links in case they offer a springboard for someone creating a similar resource.

Edited to add: All the sources cited are dated, so you might notice that some links are more recent than this post. As AI news continues to unfold, I have continued to update the Sources and Further Reading section so that I have all my references in one place. December 17, 2023.

---

Reflection questions

1. Based on what you currently have heard or experienced with AI tools, how do you feel about them? What seem to be the advantages or disadvantages, from your point of view? 
2. If you have used an AI tool such as Chat GPT to generate text, or if you think you would like to, what are the applications that most appeal to you? How do you see this tool fitting into your professional, academic, or creative writing?  
3. Are there aspects of using AI tools that you consider problematic? Which of these issues concern you most? 

What are we talking about when we talk about AI?

Artificial intelligence (AI) refers to technologies that process huge amounts of data to make complex calculations and perform functions associated with human intelligence. 

• Reactive machines respond to stimuli and can make predictions based on present conditions, but do not form memories or learn without specific input. Examples: Roombas, Netflix recommendations, computer chess. 
• Limited memory machines have the capacity to “learn” as they input more data; some make observations over a period of time, others must be fed additional data. Examples: Self-driving cars, large language models like Grammarly and ChatGPT.

(Two additional models of AI remain purely theoretical: theory of mind machines, which would be capable of processing how an individual’s thoughts and emotions affect others and are affected by others, and self-aware machines, which would have a conscious understanding of their own existence. These models do not exist outside of science fiction.) 

When we talk about AI use in the classroom and in creative fields, we are usually talking about generative AI: that is, large language models trained on massive amounts of data to learn human patterns of communication, which can respond to prompts in a natural way. Chat GPT is probably the most commonly used AI tool for generating text, although there are numerous alternatives both free (e.g. Google Bard) and subscription (Jasper.ai., CoPilot X). 

What to consider when using generative AI

Generative AI is a rapidly evolving technology, but as of this writing, there are several issues that pose obstacles or challenges to students and creatives using AI-generated text for content. 

• Inherent bias. AI is not capable of recognizing bias and may generate content that reflects racial bias, gender bias, disability bias, Western-centrism, etc. 
• Uncredited data sourcing. In most cases, creators are not notified, credited, or compensated when their work is used to train AI, or when AI incorporates their work into content it generates for users. 
• Lack of security. Particularly if you use open source AI like Chat GPT, anything you submit can be used as training data for AI and may appear in subsequent output for any other user–a potential risk for privacy and copyright. 
• Energy usage. According to one source, a generative AI challenge uses 5-8 times the amount of energy as a Google search. 
• Quality. AI is not capable of distinguishing between credible sources and unreliable sources, or between fact and fiction. For academics, it doesn’t handle research citations or qualitative analysis well. For marketers and publishing professionals, AI-generated text typically sounds too generic to stand out in a crowded field–unless you have an independent LLM trained on your brand voice. 

Across the board, industry professionals in publishing, marketing, and academia do not recommend publishing AI-generated content without human oversight. Individual organizations may require additional security measures to meet HIPAA or SOC 2 compliance standards.

Case studies from creative fields

AI as an academic tool

Under Penn’s Code of Academic Integrity, students may not use unauthorized assistance in their academic work. However, individual instructors and courses may authorize certain applications of AI. It is up to instructors to decide and for students to find out. The Center for Community Standards and Accountability has suggested that AI tools can help students with language practice, exam prep, paper outlines, or revision. (In this course, we do not recommend using Chat GPT for revising or proofreading your creative work; c.f. lack of security above.)
Read more:
Alien Minds, Immaculate Bullshit, Outstanding Questions (The Pennsylvania Gazette, April 23, 2023)

In CRWR 3000: Writing about Place, you cannot use AI to generate content for the class; your unique writing practice is both the objective and the method of this course. If you think AI will be a shortcut to help you understand an assignment or meet a deadline, we promise it will be more effective to contact us directly with questions or concerns! If you are curious about AI as a tool for writing inspiration, we have some other resources to help break writer’s block. 

AI as a stopgap measure in creative industries

Meghan Anderson is the head of marketing for Jasper, a generative AI tool specifically for marketing. She uses Jasper to check grammar and to adapt existing content (such as a blog post) into additional marketing content (including social media and ads). Kelley Peters, marketing director of Tomorrow.io, uses generative AI for content creation, campaign optimization, personalization, and data analysis. Both marketers caution against using AI for original thought, strategy, or anything that requires lived experience. They don’t publish anything AI-generated without human review. They also don’t use OpenAI, concerned that feeding their original content to open source AI would dilute their carefully cultivated brand voices.
(Source: Can A Robot Run Your Marketing?: AI in Climate Tech Marketing)

Keith Riegert, publisher of Ulysses Press, has his team work with AI for an hour each day to get familiar with what it can and cannot do well. They have used it for press releases and content fill for book covers that aren’t the right aspect ratio. Cliff Guren, digital media strategist at Forbes Books, uses AI to summarize manuscripts (although he rewrites the summaries); however, he worries that if editorial assistants do this, they won’t get the fundamental skills they need to advance as editors. 
Bill Kasdorf, an accessibility expert at Kasdorf & Associates, sees positives and negatives for AI use in book production. Using AI to generate audiobooks for a publisher’s backlist could potentially make old books available to more users. However, AI-generated image descriptions and alt-text tend to hallucinate details–not great for accessibility.
(Source: Artificial Intelligence: Revolution and Opportunity in Trade Publishing)

AI as a writing tool

Sean Michaels, author of Do You Remember Being Born?, argues that a creator should be able to use any tools available to make beautiful art. His recent book imagines a collaboration between an aging poet and AI, and he used a specialized AI tool trained on Marianne Moore’s poetry to create verse for the novel.
Read more:
Marianne Moore and AI Helped Make Sean Michaels’ New Novel (Publisher’s Weekly, October 20, 2023)
Literature Machines (Boston Review, October 23, 2023)

Different publishers may have their own rules regarding whether AI-generated text is eligible for publication. For example, award-winning science fiction magazine Clarkesworld does not permit AI-generated short story submissions; however, they are constantly flooded with AI-generated fiction and have closed their submissions portal twice to manage the submissions queue. 
Read more:
It Continues…. (Clarkesworld, May 18, 2023)

AI as a competitor

After a 148-day strike, the Writers Guild of America reached an agreement with the Alliance of Motion Picture and Television Producers (AMPTP) which included language intended as a guardrail against replacing the labor of screenwriters with AI. The agreement specifies that AI-generated material will not be considered source material (and therefore AI-generated material cannot undermine a writer’s credit), and while a writer can choose to use AI, they cannot be required to use AI. 
Read more:
Summary of the 2023 WGA MBA (WGA Contract 2023, September 25, 2023)

The Authors Guild and 17 authors filed a class-action suit against OpenAI for copyright infringement of their works of fiction on behalf of a class of fiction writers whose works have been used to train GPT without their knowledge or permission. The complaint draws attention to the fact that the plaintiffs’ books were downloaded from pirate ebook repositories and then copied into the fabric of GPT 3.5 and GPT 4 which power ChatGPT–making it possible for AI tools to generate soundalike content that users may attempt to pass off as human-generated, profiting off of a known author’s existing reputation. For example, books have appeared on Amazon with Jane Friedman’s byline, although she did not author them or profit from them.
Read more:
The Authors Guild, John Grisham, Jodi Picoult, David Baldacci, George R.R. Martin, and 13 Other Authors File Class-Action Suit Against OpenAI (The Author’s Guild, September 20, 2023)

Sources and further reading  

Artificial Intelligence: Revolution and Opportunity in Trade Publishing. Publishers Weekly. September 27, 2023. Web conference. 

Can A Robot Run Your Marketing?: AI in Climate Tech Marketing. Alder & Co. September 7, 2023. Webinar. 

Understanding the four types of AI, from reactive robots to self-aware beings (The Conversation, November 13, 2016)

Criticism

The AI Boom Could Use a Shocking Amount of Energy (Scientific American, October 13, 2023)

AI machines aren’t ‘hallucinating’. But their makers are by Naomi Klein (The Guardian, May 8, 2023)

Artificial intelligence technology behind ChatGPT was built in Iowa — with a lot of water (AP News, September 9, 2023)

Chat GPT is a Blurry Jpg of the Web by Ted Chiang (New Yorker, February 9, 2023)

ChatGPT generates fake data set to support scientific hypothesis (Nature, November 22, 2023)

ChatGPT Replicates Gender Bias in Recommendation Letters (Scientific American, November 22, 2023)

Google’s AI could soon consume as much electricity as Ireland, study finds (The Next Web, October 11, 2023)

How much water does AI consume? The public deserves to know (OECD.AI Policy Observatory, November 30, 2024)

The Exploited Labor Behind Artificial Intelligence (Noema, October 13, 2022)

The Green Dilemma: Can AI Fulfill Its Potential Without Harming the Environment? (Earth.org, July 18, 2023)

Most readers want publishers to label AI-generated articles — but trust outlets less when they do (Nieman Labs, December 5, 2023

These Women Tried to Warn Us About AI (Rolling Stone, August 12, 2023)

You won’t believe how much ChatGPT costs to operate (Digital Trends, April 20, 2023)

Tools and resources

500+ Useful Jasper Prompts [note: these prompts have not been tested on ChatGPT]

FACT SHEET: President Biden Issues Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence (The White House, October 30, 2023)

How to Make Productive Use of ChatGPT (Jane Friedman, April 19, 2023)

How to stop Dropbox from sharing your personal files with OpenAI (CNBC, December 14, 202#)

One Useful Thing [newsletter] by Ethan Molick, Associate Professor at the Wharton School 

The Muse [newsletter] by Birgitte Rasine, founder of Lucita 

Women in AI Ethics™