To bring the 2025 series of Closer to a fruitful end, bestowed upon us is the opportunity to revisit its first editions. A newly release GPT-5 Prompting Guide breathes new life into our previously tried tests.

GPT-5 is out and underwhelming. Our mini six-week experiment has come to an end. There's only five more regular editions of this newsletter before we wrap up with some preliminary conclusions in edition #25.

GPT-5 is out! That's just in time for the last edition of our 6-week experiment using the same image. I'm curious to find out of this new model brings any improvements.

The second-to-last edition of this experiment within an experiment, where AI gets to not only describe the image, and then generate an image based on said description, but also determine how to describe the image. Succes, finally?

After this week, there's two more editions of the same image being described in different ways. Thankfully, because it's not bringing us anywhere. But that's interesting to note in itself too.

We are sixteen weeks into this experiment, and I'm unsure if this is still providing value. This is the third week we let AI decide how to prompt and the results are... well, you be the judge.

Week two of our 'experiment within an ongoing experiment' where AI is not only doing the image description and generation, but is also deciding the prompting to get to that description.

Hey, so, thought: you know how we're already generating AI-images, based on AI-descriptions of real life images, right? What if... we also asked ChatGPT how to properly ask for those descriptions? Would that make it better?

I'm reconsidering my approach, but I'm not sure I'm willing to change it. Not right now. These are still interesting times we live in.

Imagine a cat behind a closed door, only paws and nose peeking from under the small crack between the door and the floor. Now imagine it again. How different are those results? Quite a lot, as it turns out.

We all know things can get weird in AI image land. But what if you give it a weird image? Will we get twice the weirdness or will they balance each other out?

SHOCKER: Asking AI for a direct replication of your work will give you a direct replication of your work. But does it really?

Learning that it's the imperfections of our actions causing things to feel real is one of this week's biggest take-aways. And that's something I think we should ponder on for some time.

Eight weeks in. Time to evaluate early results and recalibrate accordingly. What other directions should we take this experiment in?

Not to ruffle someone's feathers but a little birdie told me to continue this newsletter until the pecking order between us, beautifully flawed and inherently creative humans, and AI, the ever logical mimicker of life, has been determined. Who comes out on top this week?

By popular demand, ChatGPT gets pushed aside to give Midjourney its chance to shine. This is also the first time I'm using a self portrait as the base image. What do I look like according to AI?