[Show GN: 100개의 가중치로 ‘투명한 지능’을 만들려다 실패했습니다.

100개 미만의 신경망 가중치로 투명한 지능을 진화시키는 실험을 진행했지만, 진짜 판별 행동은 나타나지 않았습니다. 표현 방식과 탐색 구조가 문제일 가능성이 높다는 결론에 도달했습니다.

https://news.hada.io/topic?id=27193

#neuralnetwork #ai #geneticalgorithm #transparency #experiment

Theo Jansen still at it with full force... (life goals!)

Strandbeest evolution 2025
https://youtube.com/watch?v=ANhA94ZqnEQ

...and obligatory re-sharing of some pics of our joint generative design workshop in 2013, incl. Theo's graveyard of older Strandbeests and mechanisms:

https://www.flickr.com/photos/toxi/albums/72157633305609130/

Also remembering that already back then he was very worried about the manufacturer discontinuing the specific pipes used as raw materials for these creatures and therefore bought several kilometers of piping as reserve in advance, to future proof his project and not having to potentially recalculate/recalibrate all the designs...

#TheoJansen #Strandbeest #KineticArt #GenerativeArt #GeneticAlgorithm

I wonder why genetic programming so often uses a Lisp-inspired paradigm instead of a Forth-inspired one. They are both extremely simple to implement, but Forth has the advantage of being expressible as a linear sequence, so standard genetic algorithm operators can be used on it. If you find Lisp syntax more readable, it would be easy to translate, since they are both inherently nested.

#GeneticProgramming
#GeneticAlgorithm
#Lisp
#Forth

Optimize your genetic algorithm parameters for better performance! Discover essential best practices that enhance your results and efficiency. #GeneticAlgorithm #AI #MachineLearning

https://isaacl.dev/gsm