The place of AI in publishing
27/09/24 12:54 Filed in: General
The role of artificial intelligence (AI) in our personal and professional lives is a topic of constant discussion, and I fear it will remain so for a long time to come.
On the one hand, in context of the work place, it represents an opportunity to automate and streamline practices, and to rethink the way we approach both individual tasks and the need for things that we can all-too easily take for granted as being necessary.
Take this website, for example. A few months, it became painfully, embarrassingly obvious that I was way behind on adding official documents such as the Privacy Policy and Affiliate Disclosure, among others. On top of that, I was frantically busy, and in desperate need to Do Something, and Fast.
In the olden days, I would have searched the internet for examples of privacy policies by similar companies and organisations, read them several times and then tried to cobbled together my own version, all the while wondering fearfully if I had missed something.
Not this time. I did what most people would do nowadays: I took myself off to ChatGPT and asked it to provide me, in this case, with a privacy policy for a small independent publisher. I waited just a few seconds, and it gave me a perfectly reasonable template that I could adapt to the requirements of P-Wave Press.
The reality, of course, is that ChatGPT merely did exactly the same thing as I would have done, by finding and internalising several examples of a privacy policy from comparable companies, then generating a workable synthesis. The difference was that the AI algorithm did it much faster than I could have done, and did not suffer any self-doubt in the process.
So far so good.
The problem with AI is not that it can save an overburdened and time-poor small outfit a few hours by churning out something as everyday as a sample privacy policy. The problem, as discussed last year by author L.A. Davenport (and revisited in a recent column), is the expectation by its human users that it can be relied upon to perform tasks for which it is, so far, singularly ill-equipped, and then be seen as a justification for getting rid of a whole swathe of the publishing workforce.
From a creative standpoint, tools such as ChatGPT can be useful for generating ideas or as starting points for assessing manuscripts. Think images for a book cover mood board, or a first-pass analysis to look for instances of plagiarism.
It is our view, however, that AI should not at all be used for the actual creative process. We are making books for humans, not robots, and it is crucial, essential even, that the process of writing, developing and designing original works remain in the hands of real, not virtual, people.
So, I wanted to make a pledge, right here, right now, that P-Wave Press will never rely on AI tools to perform more than the most basic tasks that we are not able to carry out ourselves, with the skillsets we have available to us. All the rest will done by humans, no matter how much longer it takes.
As an artist once observed to me: we need to see the touch of the hand. And it should remain ever thus.
P-Wave Press
On the one hand, in context of the work place, it represents an opportunity to automate and streamline practices, and to rethink the way we approach both individual tasks and the need for things that we can all-too easily take for granted as being necessary.
Take this website, for example. A few months, it became painfully, embarrassingly obvious that I was way behind on adding official documents such as the Privacy Policy and Affiliate Disclosure, among others. On top of that, I was frantically busy, and in desperate need to Do Something, and Fast.
In the olden days, I would have searched the internet for examples of privacy policies by similar companies and organisations, read them several times and then tried to cobbled together my own version, all the while wondering fearfully if I had missed something.
Not this time. I did what most people would do nowadays: I took myself off to ChatGPT and asked it to provide me, in this case, with a privacy policy for a small independent publisher. I waited just a few seconds, and it gave me a perfectly reasonable template that I could adapt to the requirements of P-Wave Press.
The reality, of course, is that ChatGPT merely did exactly the same thing as I would have done, by finding and internalising several examples of a privacy policy from comparable companies, then generating a workable synthesis. The difference was that the AI algorithm did it much faster than I could have done, and did not suffer any self-doubt in the process.
So far so good.
The problem with AI is not that it can save an overburdened and time-poor small outfit a few hours by churning out something as everyday as a sample privacy policy. The problem, as discussed last year by author L.A. Davenport (and revisited in a recent column), is the expectation by its human users that it can be relied upon to perform tasks for which it is, so far, singularly ill-equipped, and then be seen as a justification for getting rid of a whole swathe of the publishing workforce.
From a creative standpoint, tools such as ChatGPT can be useful for generating ideas or as starting points for assessing manuscripts. Think images for a book cover mood board, or a first-pass analysis to look for instances of plagiarism.
It is our view, however, that AI should not at all be used for the actual creative process. We are making books for humans, not robots, and it is crucial, essential even, that the process of writing, developing and designing original works remain in the hands of real, not virtual, people.
So, I wanted to make a pledge, right here, right now, that P-Wave Press will never rely on AI tools to perform more than the most basic tasks that we are not able to carry out ourselves, with the skillsets we have available to us. All the rest will done by humans, no matter how much longer it takes.
As an artist once observed to me: we need to see the touch of the hand. And it should remain ever thus.
P-Wave Press