A confession…

…I use Artificial Intelligence.

I’m not sure what other people’s opinions are on the whole thing, but as a lifelong early-adopter (I’m running the iOS 27 beta on my phone even as we speak), it is part of my brand to try new things. I like experimenting on my computer and attempting to do things that other people wouldn’t necessarily do. Computers are a tool that enable people to do activities that they wouldn’t normally be able to do.

I’ve run the gamut with AI and discovered quite a few things about what it is capable of doing and what it is NOT capable of doing. First, let’s start with the writing, something that is part of my bread and butter as a human being.

I have come to the conclusion that artificial intelligence is NOT good at writing. I say that with a few caveats, one of which is that AI is a grammatically good writer. Sentences that it creates are well-formed and it keeps you in line in terms of the mechanics of writing. It will sit back and whip you into shape in terms of what you are writing about. It can also produce emails so much faster than I can. But that’s the mechanical, soulless writing that would be better off not in our lives.

Writing, however, is not about mechanics. Writing is about style and creativity, and both of those things are missing from every program that I’ve used. Artificial Intelligence is completely unable to come up with something original. That’s not in its wheelhouse, unless pushed and prodded, but even then, it is simply basing things on other writing that has been done in the past. The thing that makes us human; the off-kilter, voice-based, thoughtful things that, right or wrong, make us who we are…those things can’t be reproduced with AI. It can fake it, but it can’t do it.

Not that any of that matters in this homogenized world that we live in. I have a tendency to gravitate towards the flawed nowadays. I am starting to like things with problems, because those problems, good or bad, are what make it real. To be able to produce something that is flawless…well…I’ve never really enjoyed those types of things.

But I do use AI with writing. I am currently using Claude to help me edit my science fiction book.

Uh-oh, people say.

I am an independent writer. I don’t have a publishing house to back me up. I don’t have an editor to go over my work with a fine-toothed comb. (God bless my wife, who has read everything that I’ve done and offered her opinions, a job that is paid only with the joy of being married to an awesome man!). I have published books with weird typos, awful sentences, and bloated paragraphs. I’ve published really bad books sometimes because I’ve never had someone push back and tell me what my strengths are and what my weaknesses might be. I have no one in my life sitting back and going, Hey, Randall, let me spend countless hours editing your book and giving you feedback, all for free. Sorry, folks, but no one has offered to do such a thing. Most people don’t even read what I’ve written after they buy it (if they buy it at all).

Claude does that.

I have given Claude very specific parameters with my writing as we edit. I have told it from the beginning that I am asking it to look for grammatical errors and story inconsistencies. I’m writing a gigantic, sprawling sci-fi epic that deals with new worlds, new technologies, and grand ideas. And having written that book over the course of a year and a half, all while holding down a full-time job and a part-time job at the same time, I can sometimes get lost. Names might change spellings. Abilities might morph as the novel progresses. Character traits might jump from one person to the next. These are hard things to manage.

Claude is helping me through that process. And Claude can be quite mean when doing so. I have told it to be.

So in the editing process, I will feed Claude a couple of pages a day. Here’s an example of what Claude will write back to me:

What's working:

The gap between what the studio knows (Buttondown Protocol, King confirmed dead) and what they're broadcasting (missing, uncertain) is good dramatic irony that also functions as political commentary — media choosing political expediency over truth is a recognizable real-world pattern that doesn't need to be spelled out. You're trusting the reader again, which is the right call.

"KING MISSING; UNIVERSE IN DISARRAY" as the chyron is funny and apt.

What needs work:

"While we can only speculate in regard to what is happening in the Parcival Region at the moment, as there has been a communications blockade" — this sentence starts with "while" and then doesn't complete its grammatical structure cleanly. It reads as if a clause got dropped. Worth a quick fix for clarity.

"To try to make sense of all of this chaos, we turn to our regular political commentator" — "to try to make sense of all of this chaos" is a slightly limp construction for a broadcaster as polished as Tamlin. He'd phrase it more crisply.

One continuity flag: in chapter three, Kartoo Blowback said "our king could very well be dead at this very moment" and appeared genuinely distressed — "his head drooping towards his chest." Now he's pivoting quickly to Carthoid political positioning. That's not necessarily inconsistent — political commentators can grieve and advocate simultaneously — but the shift from performative mourning to "the monarchy was fragile to begin with" and "the King was weak" is notable. Is Kartoo's earlier grief being revealed as performance, or has he simply moved from emotional reaction to political analysis? Worth being clear in your own head, because Kartoo is a recurring presence and his actual position on the King will matter if he becomes more active in the plot later.

Structural note: this is the shortest chapter in the book, and it's functioning essentially as a news interlude — a beat that updates the galaxy's official position before we move to the next active scene. That's a legitimate structural choice, and the chapter length fits the function. Short chapters in a multi-POV book can work well as palette cleansers between longer, more intense sequences. Just make sure chapter eleven picks up momentum rather than giving us another short interlude — two consecutive news broadcasts in a row would slow the book's forward motion noticeably.

That’s the typical feedback that I get from Claude. What it does is make me think through what I’m writing a bit more and it forces me to fix things that I’m too ignorant/too lazy to fix myself. It’s great feedback (and this is the nicer feedback…Claude can be brutal at other times). I’ve told Claude to give it to me straight, without allowing me to cut corners. I will fix the issues above, then copy and paste it back into Claude to see what it thinks. Then the cycle begins anew.

Some of you may already have put me in a box in terms of how AI works, but to be quite frank, this is helpful. Claude is functioning as a tool that enables a person with a lot of work on their plate to get things done in a way that I wouldn’t be able to do otherwise. Claude is the kind of editor that I have needed my whole life, but I’ve never had the opportunity to work with. And to be honest, Claude might actually be better than a human being, because it remembers things that human beings can’t. The other day, it flagged a character change from the first chapter of the book that I added on page fifty-six. Most people wouldn’t remember that.

So, yeah…I’m using AI to help me edit my work. It’s a very, very useful tool and is allowing the best of my writing to emerge from a vast soup of words that would otherwise simply sit on my computer, unread by anyone.

But as you might surmise, I am a creative person, so I wanted to see what else I could do with AI. AI video is something that I’ve been confused about. What good is it? What does it do? What is the use?

I see lots of things that talk about AI-generated movies and I don’t see a lot of future in that. There’s something soulless about seeing AI-generated actors reading lines. Everything might look right, but it’s not quite there.

CGI is the biggest use-case for AI-generated video. Seeing Spider-Man last week and knowing that the majority of it was CGI makes me wonder, None of this is real. Why is AI any different?

So, I decided to make a music video to test what I could do. I picked a U2 song and came up with a basic (if on-the-nose) concept and I made a video to see what I could do with it. I made this over the course of 24 hours and nothing in this video is real. All AI-generated.

Everything in that video was from my brain. While I did not make the images that are up there, those are images that were in my head and that I told Runway to create. So, where is the line between being a creative person and having a computer create things for you. This was such a fun process that I’m excited to do it again. It’s like a whole world of opportunity opened up for me in such a short period of time.

It was an interesting experiment and gave me some real insights into what this could be used for. Everything in the video is based upon an AI agent in Runway and every shot (save one) is the first shot it spit out, all based upon my prompts.

Now, what’s the difference between this and CGI, aside from the millions of dollars that this would have cost if it was made by an effects studio? Not a lot.

I can easily see this being the future of visual effects. Now, if a filmmaker can dream up something, then they can make it (which creates a whole lot of problems on its own), without the budget-busting issues that it would have had in the past. Will this take away jobs? Absolutely. Will this democratize the process of filmmaking? Absolutely. If I had the money to pay actual artists to create these things, would I? Absolutely. But I don’t. I’m a high school film teacher. Without AI, I could not make what you saw up there.

With this in mind, I now know how to make a film that has long been gestating in my head. I know exactly how to do it and I plan to do just that. It’s going to be one of my next projects.

There are so many issues around this. I know them all. I know the job losses. I know the environmental impact. I know the financial repercussions of the whole thing. I know these things.

But now, I know my enemy. We can’t put our heads in the sand and pretend that none of this is happening. Sure, you can stand on principal that you will never use AI and you can live your life like that, but there are good uses for this. And it’s not going away. One needs to know how to use it and what the impact of using it might be. Not using it at all is a problem just as much as using it is a problem. I want to know what it can do and how that is going to impact the future.

Which leads me to my last perspective on AI, which is coding. I’ve been using AI to code all summer long and the things that I have built are glorious and something that couldn’t have been done two years ago, let alone two months ago.

But I’ll save that for next week…

Send me your hate mail and let me know what you think…