With Clone Vision, your clone can now understand images. You can manage whether or not Vision is enabled in Clone Studio Settings.

​
Training

In the train tab, you can now upload individual images to train the clone on visual experiences in your life. Some examples:

  • Upload a picture of yourself so your clone knows what you (or anyone else) looks like
  • Upload an image of a meal, so that your clone knows what a high quality meal looks like
  • Upload an image of a specific chart, so you can give your clone graphical analysis skills
  • Upload an image of your favorite workout, so your clone knows the proper form.

A picture on its own will provide information on the contents of the picture itslef, but perhaps you want to provide some additional context.

For example, if you’re a tennis coach and you uploaded a picture of someone’s backhand, you may want to add additional context on why the form of that backhand is correct.

To do this, you can click on the image you’d like to add context for in the Content table. Then, you can add context so the clone will know what to say about similar images.

​
Conversations

Users can now send your clone images in conversation. They can ask you questions about the images, or just send you pictures to get your feedback. The use case varies depending on your expertise & what you’re known for.

If you expect to get a lot of image data in your clone conversations, we recommend uploading training data in the form of:

  • Images (as explained above), especially if you want the clone to use images as examples in its conversations
  • Manual Data Entry (for example, you can upload a manual entry that says “whenever someone uploads a picture of their face, tell them they look beautiful”)