The use of generative AI in writing is a topic that comes up a lot in DIY MFA courses and programs. Let’s face it, this is probably the biggest technological revolution we’ve had in the writing space since movable type. As I’ve shared in the past, I am “AI-cautious.” I don’t like how many LLMs (large language models) use unlicensed work to train their models, nor do I like what these large computer farms are doing to our environment.
At the same time, I think putting our fingers in our ears and saying “La-la-la-la-la I’m not listening” with respect to AI is about as practical as being anti-internet in the 1990s. Unless you plan to live in a cabin in the woods next to Walden Pond and isolate yourself from all society, you have to accept living in a world where generative AI is part of our reality.
How do we use AI ethically?
I’m neither a lawyer nor a philosopher, so I’m not about to try parsing out the ethical nuances of AI use. Instead, I’ll point you to a source I trust—The Author’s Guild. This organization has been instrumental in helping writers navigate the process of getting fair compensation for their unlicensed work being used to train LLMs. It has an excellent online guide on the ethical use of AI.
After reading this guide, I feel cautiously optimistic that big corporations may eventually do the right thing and start licensing the work they use to train their models. And if they don’t, at the very least, artists and organizations like the Author’s Guild will take to the courts and maybe force these companies to do it.
Speaking of cases, many people have asked me how I feel about the Anthropic settlement, especially since my book is one of the thousands listed in that database. My response is always the same: It’s not great, but it’s better than nothing. I think authors and artists should be compensated far more than what the settlement has determined. (Let’s be real, $1.5 billion divided among however many thousands of books is not going to yield very much for each writer, nor do I think it’s punitive enough to prevent this kind of nonsense from happening again.)
Still that settlement sets a legal precedent, which is important because it makes it less of an uphill climb for other, similar cases to be decided in the artists’ favor. In my opinion, the legal precedent is in many ways more important than the pecuniary outcome.
I’m also gravely concerned about what AI is doing to people’s brains and to education as a whole. As the founder and instigator of DIY MFA, I’m a firm believer in self-motivation, especially when it comes to learning. But AI is now baked into so many online tools (including search engines) that it’s virtually impossible to do even the most basic research or creative work without coming across some AI generated result.
I suppose this would be tolerable if AI’s accuracy were not so hit or miss, but the way things are right now, it’s garbage in, garbage out. There’s no quality assurance department—aside from our own brains—to help us determine the accuracy of something generated by AI. This means if we’re not careful, we could be using AI generated content and inaccurate information without even realizing it.
What can we do about this?
First and foremost, get informed. AI isn’t going anywhere. Unless we want to live in the dark ages while everyone else is zipping into the future, we need to understand what AI is about and be aware of any controversy that surrounds it. Arm yourself with the right resources. I already mentioned the Author’s Guild, but I also highly recommend Jane Friedman’s paid newsletter The Bottom Line, where she reports on all things publishing-related, including generative AI (when relevant). This is one of those go-to publishing resources I recommend to any author who’s serious about getting published and understanding this industry.
Second, be cautious about what material you feed into AI tools. Make sure you understand all the terms of service. Read the fine print. I know we’ve been conditioned to click “accept all terms” when signing up for a new app or piece of technology. Do not do this with AI. Make sure you understand exactly how that AI tool uses and stores your data. I especially caution you against entering your entire manuscript into an AI chatbot or software. Unless you are absolutely certain that you retain all ownership and that it will not be used to train the model, do not feed it into the AI.
You also need to be aware of who owns the output. It does you no good to input your manuscript into an AI tool if you don’t own whatever material it spits out. Do your due diligence. If you choose to use AI, make sure you understand exactly what happens to both the input and the output.
Third, recognize that AI is deeply flawed and often inaccurate. If you’re going to use it, take the resulting output with a grain of salt and always supplement it with your own research and brainpower. Do not take AI at its word. Make sure to verify all results.
AI’s Impact on Education
Where things have gotten fascinating for me is the impact AI has had on education. My kids are in 6th and 8th grades, and it’s interesting to see how their schools handle the use of technology and are responding to generative AI.
My daughter (the 6th-grader) goes to a school where they’ve gone back to paper-and-pencil exams (remember the good ol’ days of blue books?) because they don’t want the students using AI during tests. My understanding is that they are opting for forms of evaluation where the students do most of the work in class, like presentations and in-class projects. This is all in an effort to prevent the use of AI in cheating.
My son (the 8th-grader) goes to a school that has taken things even a step further. In addition to monitoring the students’ use of technology, they have baked AI awareness into their curriculum. They have an entire subject called Media Literacy, where they focus on identifying generative AI outputs and also learn how to navigate the landscape of social media and the internet responsibly.
If generative AI challenges us to pause and take a step back to more basic forms of learning and testing, that might not be such a bad thing. In fact, I’m not opposed to teachers and administrators limiting the use of screens in the classroom. It’s hard enough to limit screen time at home when all the homework requires logging into a school portal. If more of those screen-related boundaries can be set in the classroom, so much the better. If avoiding AI forces classrooms to go “back to the basics” and my kids end up writing more with pencil in hand, I am all for it.
To AI or Not to AI… Where Do You Stand?
Now I’m curious. Where do you stand on generative AI? Are you opposed to its use for writing but feel it’s okay for things like checking errors? Do you find it useful for brainstorming or marketing purposes? Are you completely against it in all forms (aside from maybe spellcheck)?
I’d like to get a bit of a poll going and if we get some interesting results, I may share it in a subsequent newsletter. (Don’t worry, we’ll only share the aggregate, not individual, results.) Click the link below that best represents your take on AI.
I’m an AI user.
I use AI regularly in many contexts. It saves me tons of time doing tasks I dislike, which allows more time for writing.
I’m AI-cautious.
It may turn out to be useful, but before I can make up my mind about AI, I need more information. I want to wait and see how things develop.
I’m AI-avoidant.
I dislike what AI is doing to our world and our culture. I refuse to use it in any form. If it’s anything more than spellcheck, I avoid it.
Until next time, keep writing and keep being awesome!
P.S. For more info on Gabriela Pereira, the founder and instigator of DIY MFA, check out her profile page.
