-software development processes,
-the direction of this process,
-the impact of the future design understanding.
I think algorithms would become more and more complex and integrated to tackle various problems. Free software and open source data can play a critical role in making the tools and techniques available worldwide. This can accelerate problem solving processes in many directions. Hopefully, humans and nature will stay in the centre of attention and AI will serve it, but the opposite might also happen in case of long term mistakes in world policies by short sighted leaders.
I think generative design is an opportunity to achieve a better implementation of the concepts of:
- Computer-aided design (CAD) to Assisted & Augmented Design (AAD) ?
- Computer Aided Design and Drafting (CADD) to Assisted & Augmented Design and Drafting (AADD) ?
The future of generative design is firstly to augment capacity of creativity to generate quickly several versions of a proposal and secondly to help producers (artists or designers etc.) to manage lineage (like Data Lineage) of their works for themselves or others, even if there are physical or not. We still have example with Philippe Starck's A.I chair, who is introduced as the designer of the world’s first production chair created by artificial intelligence (AI) in collaboration with humans. https://redshift.autodesk.com/philippe-starck-designs/ and Autodek proposes others examples and https://redshift.autodesk.com/generative-design-examples/
There are opportunities too (maybe threats) to offer more autonomy or agency to machine or artifacts like robots or expressive objects. Have a best knowledge of the way humans are creative and understand how the intentional (or not when we integrate random events) creativity lineage work for each person (individual creativity) and between several persons (collective and shared creativity) will open the area of Imaginative Machine and Artificial Imagination (a component of AI which is missing).
For example, the artist Gregory Chatonsky, develops the idea of an Artificial Imagination (ImA). http://chatonsky.net/category/journal/ima/
And, the researcher Yann LeCun works on how to design a Self Supervised Learning model which will allow artificial imaginative things to explore latent space of Deep Larning and predict form, time and space aspects based on a kind of autobiographical memory (predict future from past or top from bottom or masked from the visible, etc.). https://youtu.be/SaJL4SLfrcY?t=2356
About autobiographical memory: autobiographical memory which includes memory, forgetting and remembering of past, present and future allows a part of living beiing to remember, retrieve, reconstruct, simulate, discover, reveal, deduce, simulate or imagine scenes, situation, solution or representations (spaces).
About Latent space: We can see the latent space of deep learning as the potentialy unlimited multi-dimensional (representation) space of what is hidden or what could be forgotten, remembered, discovered, revealed, deduced, simulated, reconstructed or imagined.
These deep Learning projects are example of reconstruction of shape from a part of objects.
Do 2D GANs Know 3D Shape? Unsupervised 3D Shape Reconstruction from 2D Image GANs
https://xingangpan.github.io/projects/GAN2Shape.html
Learning Deformable Tetrahedral Meshes for 3D Reconstruction
https://nv-tlabs.github.io/DefTet/
Facebook: PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization
https://shunsukesaito.github.io/PIFuHD/
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking