Artificial intelligence experiments
This section shows my experiments with artificial intelligence (AI). 
As every project is unique and has never been done before, the first thing to design is the design process.

Therefore, I must explore ways of generating design methods, form, and inspiration, as one approach for a project rarely fits another. 

These experiments show how AI can assist as a creative assistant to speed up and inform the design process. It is important to explore the possibilities of AI, only then, not to be left behind the rapid technology development. 

2019

AI car generator
When Czinger designed their latest performance car, 21C, they took great inspiration from racecars from the ’60s. What if AI could assist this process, and learn design features from a wide range of performance cars?  
Hundreds of performance cars were inputted and combined with the Czinger 21C to create new inspiration for my design process.

Using Playform.io
Porsche design characteristic spotting
What if AI learned the design characteristics of Porsche and combined them into new forms for further inspiration? The AI builds up the imagery from zero, as a “reversed deconstruction”. It tries to find “what is most Porsche first”. How early in the calculation can you spot “it is a Porsche”? What makes them so recognizable?

Using Playform.io
Sketch renderer
What if AI could render out sketches? Using my sketching style as a guide, an AI was trained to render raw line drawings in various similar rendering styles. 
I wanted to see if AI could assist my process of designing by presenting new inspiration through its way of rendering/representing form.

Using Playform.io
Style transfer 
I wanted to introduce natural renewable materials to a train design. 
To prime my ideation, I trained an AI to quickly visualize the potential of this aesthetic by morphing moodboards of wood materials onto existing trains.

Using Playform.io
Human learnings via machine learning
Using AI to combine existing trains into a mass of extracted design characteristics. 
This made me able to spot recurring form elements to use or avoid in my further design process. 

Using Playform.io 
Write your design and hit enter
The designer of the future is the one who best can describe what you are looking for. 
Ask the right questions with a beautiful (form) language. It is poetic! 

Midjourney & DALL-E Mini Boris Daymar et al.
An alarm clock that donates to charity every time you hit snooze, product design
An alarm clock that donates to charity every time you hit snooze, product design
An alarm clock that donates to charity every time you hit snooze, product design
An alarm clock that donates to charity every time you hit snooze, product design
Sleek, white, and cute drone, product design
Sleek, white, and cute drone, product design
Sleek, white, and cute drone, product design
Sleek, white, and cute drone, product design
Generative Design and Topology Optimization
​​​​​​​To learn Generative Design, I explored how the generated forms can match constraints using minimal material and parts while sparking inspiration for new forms. 

Using Autodesk Fusion 360
GPT-3
An insightful prep talk before project start.

Using OpenAI GTP-3
GPT-3 Simply
Experimenting with writing assisted AI.

Using Simply by Chris Lu  
Aston Martin design generator
What if I could condense and extract design features from the Aston Martin heritage, and merge them into new combinations? Using artificial intelligence to comprehend Aston Martin's form language and generate serendipity.

Using Playform.io