Discussion about this post

User's avatar
Timothy Burke's avatar

People aren't going to acquire deep knowledge if AI is used in the way that its chief boosters promote. You wouldn't know if it had edited your manuscript to your satisfaction if you hadn't edited and written work substantially on your own before. Generative AI is going to kill the developmental loops that most people who've worked with it recognize are essential to using it as a tool. Since it's not essential--you could have done all the work you describe doing yourself or with human help before generative AI existed--this is a pretty strong argument against using generative AI. We could do without it--the constraint on you in this case was that you didn't have the money to pay humans to do that work and you didn't want to take the time it would take for it to get done. That's arguably a problem either with our economy and the way we make funding available to start small businesses or a problem of allocation within your own household, e.g., you aren't seriously invested in the business you are starting. (Which, if so, might turn out to be a much bigger problem for all of us, if we are all of us starting businesses so easily and without initial risk that we swamp the economic niches open for those businesses.)

I don't see anybody promoting substantial AI use who is even slightly addressing the problem of its near-total choking off of the paths that lead people to develop the expertise that is supposedly the core value that we have to offer our workplaces or our clients.

Jennifer Gruden's avatar

Just a note - I'm sure you've thought this through for yourself but some agents and publishers are requiring human authorship warranties, so if anyone's writing a book and planning to publish it traditionally, it's a good idea to do a check on the market expectations for AI content and use. It may be particularly important to have used premium versions where your text does not go into the training data.

4 more comments...

No posts

Ready for more?