To add to this, Adobe's Firefly AI beta offers some nice interface enhancements for image generation: a browser for style, lighting, composition and more. Each represented by a chiclet containing the style name and thumbnail of what it looks (or does) like. Clicking on them added the chiclet to your prompt. It's a nice visual way to represent what are arguments in Midjourney. It's exciting to see a company embracing and experimenting with a visual approach to AI generation.
This certainly helps but I wonder if the solution to the interface challenge is really as easy as a bunch of multi-selects and dropdowns that add text to prompts 🤷🏼
To add to this, Adobe's Firefly AI beta offers some nice interface enhancements for image generation: a browser for style, lighting, composition and more. Each represented by a chiclet containing the style name and thumbnail of what it looks (or does) like. Clicking on them added the chiclet to your prompt. It's a nice visual way to represent what are arguments in Midjourney. It's exciting to see a company embracing and experimenting with a visual approach to AI generation.
This certainly helps but I wonder if the solution to the interface challenge is really as easy as a bunch of multi-selects and dropdowns that add text to prompts 🤷🏼
> I had Suleyman ask Pi a question that had shown up in the company’s user logs: “why does my husband want a divorce?”
Uh, should we be concerned that the CEO of a company that takes intimate, emotional requests from users is able to view those requests?
And, in addition to viewing these requests, is also willing to share said requests with a journalist?
Should have added that these queries had been anonymized, stripped of personal data, and in some cases re-written before he showed them to me