AI is here to stay, whether we like it or not.
Some forms of AI, like voice recognition software, can be used responsibly and ethically. For example, when I use AI to pull the transcript from one of my videos so I can repurpose the words I said, there are no ethical problems.
Issues with AI arise, however, when it is used to manipulate, repurpose, or recreate someone else's words, voice, image, or style.
AI and consent
Unless you've been hiding under a rock with cotton wool stuffed in your ears, you know that consent in all sorts of areas of public and personal life is now seen as very important. Discussions I've seen lately about consent range from talking to teens about sexual consent in an attempt to combat rape culture to marketers who only sell with consent.
If you followed the recent actors' and writers' strikes in Hollywood, you'll have noticed a lot of talk about consent (and the right to withhold consent) around using AI and things like
- an actor's image or likeness
- an actor's voice
- a writer's style
Hopefully you all see that actors and writers must protect such things because that's how they make their livelihood. And what is true for the artists and creators in Hollywood is true for artists and creators everywhere.
When you create something, you have the right to expect to retain your intellectual property rights.
It's starting to hit the courts
If you've had a conversation with me about AI, you've likely heard me say that all of these issues will soon be thrashed out in the courts. That's starting to happen. For example, in January of this year, George Carlin's estate sued the makers of an AI generated comedy special using Carlin's image and style.
As you'll see in the article from The Guardian that I linked to above, the lawyers for Carlin's estate are alleging "violations of Carlin’s right of publicity and copyright." Since I'm a writing coach, I've never had any need to learn much about the right of publicity, so I won't attempt to comment on that. I do know a bit about copyright, though.
Based on the Guardian article, seems that the argument that Carlin's copyright has been infringed comes from the fact that the company who made the AI special trained the AI on recordings of Carlin's performances. Then, the AI tried to produce something similar.
If Carlin's estate wins, it won't be long before the producers of other materials used to train generative AI (everything from recordings, works of art, and books to websites, blogs, and online courses) start lodging similar complaints in the courts. This is just one of many reasons why I don't think it's a good idea to use this kind of generative AI to create anything for your business.
AI and truth
AI doesn't always tell the truth. If you follow American politics (especially the events surrounding the various court cases involving a certain former president and his team), you'll have heard about this. As discussed in an NPR article, Michael Cohen "says he unwittingly passed along to his attorney bogus artificial intelligence-generated legal case citations he got online before they were submitted to a judge".
In other words, he asked Google Bard (a generative AI platform, like Chat-GPT) for evidence to support a claim he was making in court. The AI couldn't find any, so it created some.
There have been other, lower profile, cases of generative AI making stuff up when it doesn't have actual answers to queries it is given.
For your business, the risks of using made up stuff from AI may not be as high as sending made up citations to a judge, but it's not going to make you look good either. My advice would be just don't use it.
AI and your writing
As you'll have gathered, I'm not a fan of writers using AI to create their content - this stance holds across all the fields I work in from academic writing, to blogging, and nonfiction books.
Whatever you write, if your reader wants to know what a bot thinks, they'll ask one. They're reading your piece because they want to know what you think. When you use AI to create content (or even to generate ideas) you're doing a disservice to you, your clients, and your readers. Let's look at how.
You
When you write your book, article, course, or blog post you crystallise a lot of your thinking about your specialist topic. It doesn't make sense to let a machine take that away from you. If you did, you wouldn't get any of the benefits of having written it, like increased clarity and confidence.
Taking the time to put your thoughts and processes into words forces you to think consciously about things that otherwise seem to "happen naturally". To grow as a person, writer, and professional, you need to take the time to do this.
Your Clients
When your current and future clients (or for any academics reading this, your students and colleagues) see that you've published a book, article, course, or post they rightly expect that to mean that you know everything that's in it. That you took the time to consider how to present that information. And that you know where the information came from -- and no, they don't expect it to have come from a bot.
Your Readers
Your readers have many of the same expectations as your clients. When people read your book, article, post, or course materials they want to hear from you. If they wanted to know what a bot "thought", they'd ask one.
Help for struggling human writers
As I hope the LinkedIn bots some day learn, I'll never help train their AI to write better. I will, however, help you lovely humans make your writing easier.
It seems to me that when writers turn to generative AI it's usually out of fear and/or frustration. Both can be offset with a solid plan, like the one you'll find in "How do you decide what to write?". Give it a go. If you still need help, click the "Let's talk about writing" button in the menu on the right side of your screen (on desktop; for mobile and tablet, scroll past the end of this post to find it) to book a chat with me about your writing problems.