Yesterday, I launched monochrome.blog
It’s an experiment in using AI to augment my photos. You may have noticed that when I publish photos here on Paolo.blog I rarely add a legend, and typically limit the title to one word. I even use the same word multiple times for series like Focus, Work, Wait, or Ride (or even Walk).
For monochrome.blog I created a Custom GPT, a small AI assistant based on ChatGPT, that takes one image as input, and produces:
- An ALT text describing the image. This is useful to increase accessibility to special reading devices.
- A reflective legend capturing the essence of the street photo.
- Three suggestions for titles.
Agent Photo, the assistant in question, is open to use for everyone. It only works well with black and white street photos, as that’s what it expects as input. If you use it with other types of photos your mileage may vary.
On monochrome.blog, all photos are mine, all the text is produced by Agent Photo.
Comments are open on the site, but please know that the answers will be generated by Agent Monochrome 😉
Leave a Reply