I took a fiction writing course once during a hazy undergrad phase. One memorable story I recall, written by a classmate, featured a character named Awesome. That was before awesome was awesome. The character Awesome was ironically named. But the term, for me, took on a layered meaning from that point on.
This morning when I read Chanelle Henry’s “Is it too late to be awesome?” I wondered if awesome had jumped the shark. But this afternoon I saw what happened to Rafael Matsunaga’s Genetic Cars and I knew that awesome was just getting started.
Rafael is the personification of Awesome 2.0. A world-class Yo-Yo competitor, Rafael lives in Sao Paulo
His HTML5 work, done for the fun of it, brought genetic algos to life in a way never done before. Because Rafael’s car mutations use client-side compute (javascript), the next natural step was distributed, which is exactly what PubNub did with gencar.co and the race got really interesting. Now a collection of systems test mutations, dramatically accelerating the adaptation rate and fitness.
Genetic algos have been (in my view) the red-headed step-children of machine learning. But now thanks to the inspiration of Rafael and engineering of PubNub, real-world stuff may begin to actually benefit from novel machine-discovered designs.
Holland and Koza, take note: With the inspired work of Rafael, Awesome 2.0 has your DNA all over it.