mitchellens.ink

For the past 25 weeks I’ve been running a simple experiment. Every Wednesday, I’d take one of my photographs, ask AI to describe it, and then feed that description back into the machine to see what image it would generate in return. No tweaking, no retries. One human image in, one synthetic image out. The idea was straightforward: measure how “close” AI could get to the way a photographer sees.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #024. Prompts are in grey boxes. Descriptions are in italics. Continuing in last week's spirit of sharing images from personal projects I like a lot, this final regular entry into Closer is shot during my first long(er) distance walk: the Grebbelinie. This photograph is taken while standing on the Liniedijk, just outside Leusden.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #023. Prompts are in grey boxes. Descriptions are in italics. Back in 2020 I did my first long(er) term photography project that wasn't a 365 project. It was my first attempt to discover a voice of my own. The project revolved around three 360 degrees panorama photos, that I mounted on the inside of steel frames.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #021. Prompts are in grey boxes. Descriptions are in italics. It's like OpenAI is aware we're wrapping up with Closer for now and decided to keep things interesting for the final few weeks.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #021. Prompts are in grey boxes. Descriptions are in italics. We have four more weeks left. I needed a ChatGPT Plus subscription, so that's a change from previous weeks. Let's see if it's any better at image-generation. We're reusing the prompt from last week, by the way. Perhaps Plus is better at interpreting the description, maybe it won't.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #020. Prompts are in grey boxes. Descriptions are in italics. Now that our unfruitful mini-experiment has come to an end, we can conclude we shouldn't ask ChatGPT to write prompts for us, that are then used to describe input images. Unless you want underwhelming results.

Hello my patient friends, In the spirit of keeping things as short as I can. Let’s dive into this month’s thoughts, and progresses. Slow progress As promised, we’re focusing this newsletter on my longterm photography project Covering again. Which, by the way, has a new button added to the project page that links to a folder in my online photography archive, housing every picture I’ve ever taken in Amersfoort that has a GPS location embedded in the file.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #021. Prompts are in grey boxes. Descriptions are in italics. It's like OpenAI is aware we're wrapping up with Closer for now and decided to keep things interesting for the final few weeks.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #020. Prompts are in grey boxes. Descriptions are in italics. Now that our unfruitful mini-experiment has come to an end, we can conclude we shouldn't ask ChatGPT to write prompts for us, that are then used to describe input images. Unless you want underwhelming results.

Welcome to Closer, A weekly experiment in AI interpretation: I share a photo, ask AI to describe it, then generate an image from that description. This is edition #019. Prompts are in grey boxes. Descriptions are in italics. Chat GPT-4o gave me six different Prompt Styles and we've tried them all on the same image over the past six weeks. The first week was hilariously bad, the second week was uneventful, the third week made me question the continued value of this newsletter. The fourth week?