I don’t really know how to start this article, or where it will take me - because I’m going to try to tackle a huge topic in the creative field right now: AI
So where to start… I could ask ChatGPT how to start, let’s do that:
Prompt: “Give me an outline for an article I'm writing about AI”
ChatGPT:
Introduction
1. What is Artificial Intelligence?
2. Key Developments in AI
3. Applications of AI Across Industries
4. Benefits of AI
5. Challenges and Ethical Considerations
6. Future of AI
Conclusion
Call to Action
There you have it, an outline for the most boring, predictable article I could have possibly thought of about AI, which brings me to the central premise of this article:
AI is fundamentally not creative.
Before we dive in let me say that I do actually use AI from several different companies. I use ChatGPT, Gemini, I’ve played around with DALL.E and Midjourney, I use AI to transcribe and close caption videos, and I use AI in lightroom and photoshop. I’m far from an AI hater, but as with all Lathley articles I will try my best to present a nuanced, in-depth look at a topic that most articles just skim the surface of.
Most importantly I’m not here to try to fight the future. If you try to fight the future you will lose, and AI is going to be a big part of the future (and in many ways is already a big part of the present). But humans are terrible at predicting the future. In the not so distant past it was thought that we would have flying cars by now, and the idea of being able talk face to face with someone through a screen that wasn’t plugged into a wall was too far fetched even for science fiction. So anyone claiming to know what the future of AI looks like is, let’s say, naive.
AI is a tool like any other, and in a lot of ways current AI is really just expanded versions of tools we’ve had for years. The AI that Adobe has plugged into Photoshop and Lightroom is really just enhanced versions of tools like healing brush and content aware fill - tools those programs have had for years. In a lot of ways ChatGPT and Gemini are more akin to expanded versions of autocomplete and autocorrect than actual intelligence. AI is a tool, and a great tool, but I believe the issue lies in AI developers and users fundamentally misunderstanding the tool they have. It feels like AI is a screwdriver, but everyone is trying to use it like a hammer. Let me explain.
Recently Apple released the iPhone 16 that’s “Built with AI in mind” - in the release video they showcase the power of AI by showing it rewriting a dinner party invite for friends from bullet points to sentences. To me this is using a screwdriver as a hammer - using AI for human to human interaction, especially among friends. Why would anyone have trouble writing a few sentences to their friends inviting them for dinner? Why would you need AI for that? To me that’s just as bizarre a use case for AI as having it write a heartfelt letter to your grandma.
It’s not that AI can’t write things for you, but having AI write something for you is one step lower than having “Dictated but not read” at the end of something you write, since you didn’t even dictate it. So for human to human interactions, if you wouldn’t feel comfortable putting “Dictated but not read” at the bottom, don’t use AI to write it.
AI imagery is another thing that creatives and people are trying to grasp the use case for. People have been manipulating photos for as long as there has been photography, and AI is just the next step. It’s a tool, and every photographer (and client) will have to decide how much they want to use this (slightly) new tool. It’s always a choice how far to push an image, and that’s been the case for decades now. Ikea used 75% computer generated images in their catalog way back in 2014!
Photography is always changing and evolving and this is just the latest tool in a long history of new tools. I highly doubt it will be the photographic reckoning some suggest though. After all, some of the most prolific painters like Picasso, Monet, Van Gogh, Dalí, Pollock, Warhol, Kusama, Basquiat, etc. created the entirety of their work after the invention of photography. Photography might have changed certain parts of painting but it didn’t kill the entire industry. Far from it, unless you’ve actually seen an original in real life the only reason you know what a Picasso painting looks like is because of photography. As with anything, anyone who claims to know how a new technology will affect a centuries old practice is almost certainly wrong. But I can say this: I can’t imagine anyone hiring a prompt engineer instead of a wedding photographer.
Let’s talk about another use case for AI - research. Tools like ChatGPT can quickly research and condense material into key terms and bullet points, which can be fantastic for research. The issue is that without being able to cite sources, there is no way to know where the information is coming from, whether it’s false info, out-of-date, or just completely made up by the AI itself. This means that usually any research done through AI has to be researched and fact-checked again manually, which negates a lot of the time savings in the first place.
Okay so we’ve bashed AI a lot, so how the heck can you actually use AI effectively? Well I can’t answer that, but I can tell you how I tend to use AI. I’ll talk about text AI since that’s the most common - but I use it like it’s a very cheap writing assistant, and you get what you pay for. That is to say if you want an extremely mediocre first draft of something jammed out in 10 seconds, AI works great. Or if you want the cliff note research on something done so that you can really dive into the research later - AI. Or if you’re a little stuck in writing and you want to brainstorm topics and ideas with a dull intern just to get the brain juices going - AI. I’ll give you an example:
Let’s say I have a company that is just coming out with a new pair of pants specifically designed for rock climbing and I’m writing the product description for the product page on the website. I want to appeal to true rock climbers, not just a bland description for everyone, so I ask ChatGPT:
“Write a product description for a new pair of climbing pants using slang and lingo that only true rock climbers would get” - not the best prompt I know since it doesn’t have any details about the product but that’s beside the point.
ChatGPT gave me this:
Product Name: Crux Crushers Climbing Pants
Product Description:
Send it in style with the Crux Crushers Climbing Pants—made for climbers who thrive on sketchy beta and sandbagged routes. Crafted from bomber stretch fabric, these pants handle everything from crimpy gastons to gnarly knee bars. The gusseted crotch and articulated knees give you all the mobility for high steps and heel hooks, while the reinforced seat keeps things durable through chossy chimneys and epic whippers.
The low-profile waistband fits comfortably under your harness, and deep, secure pockets hold your chalk brush, topo, or a snack when you're running out a sketchy pitch. Moisture-wicking tech keeps you cool when Elvis leg sets in on the crux.
From projecting in the gym to full-on send trains at the crag, the Crux Crushers have your back. Rack up, chalk up, and crush your project—because these pants aren’t tapping out before you do.
This reads like it’s written by a lunatic. It’s bizarre, awkward, and clearly written by someone (in this case something) that has no idea what it’s saying and obviously doesn’t actually rock climb. Terms like beta, sandbagged, crimpy, knee bars, chossy, chimneys, whippers, Elvis leg, crux - these are all actual rock climbing slang terms but they have been assembled in a way that clearly shows they were written by a total gumby (Gumby: A beginner or inexperienced climber who lacks knowledge or skills. "That Gumby doesn’t know how to tie a proper knot.")
All that said it’s not useless, there are some interesting nuggets in there. The line: “The low-profile waistband fits comfortably under your harness, and deep, secure pockets hold your chalk brush, topo, or a snack when you're running out a sketchy pitch.” That's actually a pretty good line and one rock climbers could resonate with. Even the product name it came up with: Crux Crushers is pretty cool, but it’s not original - surely just scraped from the internet.
Overall it’s an okay place to start, if an actual human re-worked and rewrote this it could be really good. The best part is that you didn’t start with a blinking cursor on a blank page which can be a really difficult place to start. You have a bit of something to go on, rework, and rebuild, and it took a free program 5 seconds to make.
It’s like a very mediocre writing assistant, or a very good autofill - not entirely useless but not nearly as handy as you’d like.
So there you have it, the kind of rambling, meandering, somewhat useful article that could have only been written by a human ;)
