One of the things I love about building Gumnut is that people keep coming up with new ways to use it, that I personally would have never imagined. For example, yesterday I was chatting with Ayo at the office, and I gave him a demo of the product’s current capabilities.
We started with a pretty basic prompt: “Show me photos of the mountains with my kids in them". Claude and Gumnut did a fine job, but this prompt doesn’t really explore what agents are capable of when paired with Gumnut. To respond to this prompt, Claude invoked a single tool to search my photo library, then displayed and described the results:
“Show me photos of the mountains with my kids in them”
Ayo’s next request was much more challenging: “Create a portrait of me that is composed of photos of people in my photo library”.
For this prompt, Claude ran a series of queries against my Gumnut photo library for people and faces. and then built an interactive web app:
And here’s what I get when I run that app:
“Create a portrait of me that is composed of photos of people in my photo library”
These are the kinds of creative use cases that get me really excited about the potential of Gumnut. Use cases that enable people to get more joy out of the memories they already capture. Use cases that in the past would be too time consuming or tedious, but are easy with agents and Gumnut.
Feel free to send me additional ideas you’d like to see me try! I’m out with my family on a trip next week, but I’ll be back the week after, and I’m planning to start opening up Gumnut to more people soon after that!
