How can designers and creators really put AI into their everyday workflow?
The problem with current AI tools, conclusion from 30 interviews
After talking with several designers and creators, our product team has summarized 3 key problems for people to really put AI into their everyday workflow. Before starting the discussion, let’s first define who are the “designers” and “creators” we’re talking about.
Designers & creators
This includes a whole bunch of people in various disciplines indeed. In our definition, this includes anyone who creates stuff, such as games, videos, audios, images, texts, physical products, services(service design) or even strategies (consultants?).
https://a16z.com/2022/11/17/the-generative-ai-revolution-in-games/
You might question how?
Film producers and marketing copy writers have totally different workflows. Interior designers design in a completely different way from fashion designers although they are both categorized as designers in general.
Although they all seem to have completely different needs, the 3 key problems for these designers & creators to really put AI into the everyday workflow are similar, as long as we’re excluding developers & code-writing.
Problem 1, the user interface
Why?
Designers and creators all have similar problems because when we use AI, we all need to compare and contrast the AI outputs, and iterate on them.
That brings us to the first problem:
the user interface of current AI tools is terrible.
Let’s take designers as an example. When we’re in the ideation (brainstorming & research) and iteration process, we need a place to see everything at once.
We need a place where we can see everything, so we can truly organize all the ideas and graphics.
However, with the interface of current AI tools, we’ll need download images, scroll up to find previously produced images, download images again, copy paste prompts, edit prompts somewhere, generate new images and download again, and finally, we still need to upload all images to somewhere else.
With current tools, brainstorming and iteration become a true nightmare for designers & creators.
Answer: an infinite canvas + database
an Infinite Canvas to put everything on the same place, a database for you to record numbers and iterate on different prompts.
We believe that canvas + database is indeed the better workspace for designers and creators to truly put AI into the everyday workflow. That’s why we’ve built Fabrie.
https://www.youtube.com/watch?v=5kAHSvAh4U8
Problem 2, more AI features
We need lots more AI features, not just generating graphics and texts
- AI that helps do rendering? turn sketches into 3D model?
- AI that searches and finds similar style/artist reference from given graphics? brainstorms on design ideas/techniques for us?
- AI that can create mood boards for given furnitures and materials?
- AI that can produce animation? apply different effects shots from different angles from given videos?
- AI that turns pop music into funk style or jazz?
Problem 3, train our own AI
Ultimately, we’ll all need our specific AI for real work. An AI where we can input resources, have it trained on these inputs and outputs for us.
I’d really like to input a whole website like Pottery Barn (a house decor brand), and have AI create 3D room rendering based on this input for me.
Ending
Problem 2 and problem 3 are still something Fabrie is still working on, especially the technical setup and the user interface. Any feedbacks or ideas is welcomed.
