Conventional AI vs. Generative AI

There are two relevant types of AI for an author/artist. One is often called traditional/conventional AI. Many programs use this type of AI to identify things like spelling errors, misplaced punctuation, and grammar issues. These programs use predetermined algorithms and rules to analyze against your writing. For example, they contain dictionaries and then compared your words to the spelling in this dictionary. Spell and grammar checks in common programs like Microsoft Word use this type of AI.

I am an advocate for people using these types of tools. They have been around for many years and can be very helpful. We all make typos! But also know they are not always correct. Check out my Software vs. Human Editor blog post for more information.

The second type of AI is called generative AI. This is the one causing so much controversy because many of these programs were created being trained on writing and art without the consent of the author/artist. This is a fact. CEOs and engineers have stated this openly. These programs absorb inputted information and can regurgitate it later. This is why plagiarism is such a concern when using it and why, at least in the US, AI-created projects cannot currently be copyrighted. It continues to absorb more information as users’ input prompts, pieces of their writing, and more content is generated.

Writing is a skill. It takes more than just imagination to come up with a good story. Writers need to learn how to write. They need to understand things like what head-hopping is and what point-of-view is and how to flesh out a character to make them well-rounded and not a flat caricature. Using generative AI not only relies on someone else’s previous work, it removes the creative elements and the earned skill. If AI is generating content for you, even with your own prompts, you are not writing it.

There are a host of issues with AI-generated content. Many editors will not work on such content because it often requires so much clean up, the project reaches the level of ghost writing. There is often repetition and plot holes. There is no true author voice. The words lack the human element that draws people in and allows them to connect with the characters and story on a deep level. Don’t sell yourself short by trying to cut corners and pushing out mediocre stories.

In line with this, I also do not support the use of generative AI for book covers, marketing content, or audiobooks. Now, I know all of this can be very expensive for indie authors. Believe me, I know. But if you’re willing to rely on AI because it’s cheap, how is that different from people using AI to create books?

However. Oh yes, few things are black/white! However, while using AI to create your content or even "fix" it can be problematic, there are things generative AI is typically considered good for. Want to come up with some relatively good numbers for a new planet you're creating? Generative AI is like a smart calculator that comes with its own formulas! Trying to find a word you can't quite remember? AI can work as a reverse dictionary. Looking for novels similar enough to your own to use in a query letter? AI may be able to pull up those with similar ideas. Get the idea?

The important thing here is to never use generative AI for things that have to be concrete and accurate. It can pull anything off the internet; doesn't mean it's correct. The numbers it gives you or the book suggestions may be off. Take the suggestions and look them up. But in this context of assistant, generative AI can help.

One last thing I do want to point out is the environmental toll of AI. The data centers that run these programs use a lot of electricity and water. We all use enough as it is, so maybe consider how much you need to use that AI or if some humans could lend a hand.

Overall, I don't recommend using generative AI for writing or art. It has it's uses, but let’s keep the humans in art. We’re the ones art is supposed to be for, after all.