For the record, I am not a member of the Authors Guild, but I am a member of the Dramatists Guild of America, which has endorsed a recent petition by the Authors Guild to protect intellectual property from AI theft.
The Authors Guild is asking for some basic parameters around AI generation. The first one is consent. If I don't want my words used by an AI generator, they shouldn't be used. The second issue is compensation. Writers are entitled to compensation when our words are used, and that includes when they are used by an AI generator. The third major issue is transparency, which has been noticeably lacking in companies using AI generation.
How could we establish these basic principles? Well, for one, we could require that when people's names or the titles of their works are used in prompts, individuals and copyright holders get a share of any profits made by an AI generator. If a student, for instance, uses an AI generator to write a response paragraph about Arthur Miller's The Crucible, and the AI generator is making ad revenue off of this, at the very least, the Miller estate should share in that revenue. (Of course, the student probably shouldn't be using the AI generator to cheat anyway, but that's a different issue.)
Another thing we could do is require anyone who uses AI-generated content to label it as such. If a work contains more than, say, 10 percent AI-generated content, anyone using it (such as a website) needs to say so. Not only do we all lose out when students use AI to cheat on assignments, but we are all going to suffer when journalists, artists, and others use AI to cut corners and make deadlines.
These are some very basic suggestions, but they are doable, and they would be an improvement over the status quo. Silicon Valley needs to take note, and so, for that matter, does Washington, D.C.
