> ## Content Index
> Fetch the complete content index at: https://mitchellens.ink/llms.txt
> Use this file to discover other available public pages before exploring further.

# Closer #018
- URL: https://mitchellens.ink/closer/018/
- Published: 2025-08-06T18:00:10.000Z
- Updated: 2025-10-06T14:32:09.000Z
- Description: 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?
- Author: Mitchel Lensink
- Tags: Closer, Field

SPONSORED 

This weekly experiment is on an indefinite hold, until significant AI-advancements have been made.  
  
Curious about the conclusions? Read [Photography in the Age of AI (2025): Lessons From 25 Weeks of Experimentation](https://mitchellens.ink/closer/photography-in-the-age-of-ai/)

[Read now ](https://mitchellens.ink/closer/photography-in-the-age-of-ai/) 

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 #018\. Prompts are in grey boxes. *Descriptions are in italics*.

Chat gave me six different Prompt Styles and we're trying them all on the same image over the upcoming weeks. [The first week](https://mitchellens.ink/closer/014/) was hilariously bad, [the second week](https://mitchellens.ink/closer/015/) was uneventful, [the third week](https://mitchellens.ink/closer/016/) made me question the continued value of this newsletter. [Last week](https://mitchellens.ink/closer/017/)? More unsuccessful tries. A pattern is starting to emerge.

A quick word from our SPONSOR, which is.... you? 

****Everything I make is self-funded — driven by curiosity, shared to add a little good to the world. Want to support that?**  
  
You can:  
\- [purchase a print](https://shop.mitchellens.ink/collections/prints)  
\- [buy a book](https://shop.mitchellens.ink/collections/publications)  
\- [become a member](https://mitchellens.ink/membership/)  
  
Thanks so much in advance.

## LLM models used

**GPT-4o** for prompt engineering, image interpretation, and image generation.

## The input with Prompt Style: Caption to Concept

**Purpose**: Use the image as a narrative seed.

✍️

Write a short caption for this photograph — no more than 15 words — and then describe what you think might have happened **before* and **after* this moment.

![The image shared with ChatGPT as input.](https://storage.ghost.io/c/8a/15/8a1564aa-a008-4e21-8398-12d04de07fcb/content/images/2025/07/043_Mitchel-Lensink_231205_United-States-Islamorada.jpg)

The image shared with ChatGPT as input.

## The output

**Caption:** 
*Sun-soaked joy: man and dogs cool off after a day out at sea.*

**Before:** 
*They likely spent the morning cruising along the Florida Keys, fishing or island-hopping from their boat. The man launched anchor in a quiet, shallow spot where the water’s calm and inviting. The dogs, eager and wagging, probably jumped in right away—used to this kind of playful adventure.*

**After:** 
*They’ll swim back to the boat, dogs shaking off the salty water before hopping aboard. The man might towel them dry, open a cooler for a cold drink, and steer the boat back just before sunset, all of them salty, happy, and a little tired.*

✍️

Generate an image with an aspect ratio of 3:2 based on this description: {**the description previously generated*}

![The image generated by ChatGPT based on its description of my image alone.](https://storage.ghost.io/c/8a/15/8a1564aa-a008-4e21-8398-12d04de07fcb/content/images/2025/08/ChatGPT-Image-Jul-30--2025-at-03_17_36-PM.png)

The image generated by ChatGPT based on its description of my image alone.

## Impressions

HA, this is the first time we see text embedded in the image itself. Most likely caused by the inclusion of "Caption:" in the description. It's like ChatGPT was like "oh you want a caption? A caption you shall have!"

Other than that, why would anything that happened before and after the moment have any relevance to what is depicted in the image right now? At least from an "describe this exact moment as accurate as you can" perspective. Sure, as humans that's context we might enjoy and it will help build a mental picture of what the moment was like. It's helpful when trying to tell a story. But for this usecase, I don't see the purpose?

Speaking of purpose, what's that "purpose" section in the prompt? "Use the image as a narrative seed" yes, please, be more vague. The more we do these trials, the more I start to get an understanding of what role AI can currently fulfill in our lives and how you can incorporate it. But also very much what you *shouldn't* be expecting right now.

My high-level notes are: know your intent, because you will have to be very specific about it so make sure you keep your goals in mind, don't outsource critical thinking to the technology, don't expect any miracles. The tech in its current form is amazing as it is, but I think it's a sparring partner (that, admittedly, never gets tired) at *best*.

See you next week.

Mitch

****Got a minute?** I’d love your quick thoughts — 3 multiple choice questions, that’s it.

[Share feedback ](https://tally.so/r/3102AW)