Tag: Generative Art
-
Spline-Based Snakes
A lovely name for a useful image analysis technique. Kazushi Mukaiyama seems to be using it in Shizuka now. -
Latent Semantic Analysis
Via Planet Lisp. This looks interesting. I’ve wanted to apply Markov Chains to bitmaps and PostScript for ages, this looks even better. -
Hackers And Painters
Hackers and Painters is out. It looks good. I’m just waiting for Amazon.uk to get my copy past the tachyon projectors.Sample chapter, “hackers and painters” (pdf).
See the book site for more details and reviews.
More Paul Graham essays can be found on his web site.
Now if he can just get on with Arc… ๐
-
Harold Cohen Tate Talk Now Online
The RealMedia archive of Harold Cohen’s talk at the Tate is now online at:http://www.tate.org.uk/onlineevents/archive/harold_cohen/
Click on the “play” link to watch.
I’m sat next to Dave, who asks how AARON decides when to finish a work during the Q&A session at the end. I was too involved in looking at the projected images of AARON’s latest work to ask a question…
-
Algorithmic Aesthetics
I finished reading “Algorithmic Aesthetics” today. It’s a tantalising glimpse into how entirely procedural, rule-based description, evaluation and/or generation of artworks might function. I’m not convinced that length of input divided by length of output tells us very much about an object’s aesthetic interest, but Gips & Stiny cite more than one historical example of just such an aesthetic measure, and the complexity of execution versus the complexity of effect of a work can be a convincing measure of one dimension of an artwork’s interest.
I think that the examples Gips & Stiny construct deal with style and pattern rather than with higher-order aesthetics. But style is certainly part of aesthetics, and the authors’ separate work on describing specific styles of architecture and design using Shape Grammars shows that their ideas could be effective even with 1970s technology.
The book contains the first reference to “generative” art, well, “generative techniques”, that I’m aware of. Well worth a read if you can find a copy, or you can browse the complete text on the web. -
Things AARON Doesn’t Do
What AARON does is fascinating, but AARON doesn’t do is interesting as well.AARON doesn’t:
- Remember anything between drawings.
- React to the outside world.
- Request extra information or abilities.
- Write its own programs.
These capabilities are handled by another module called “Harold Cohen”. ๐
-
Computer Arts Society
The CAS is restarting. See the homepage at the BCA, or contact christos@logothetis.co.uk (the name given out on flyers at the Harold Cohen talk). -
Harold Cohen Talk Last Night At The Tate
Harold Cohen gave a good talk, and a good Q&A session afterwards (including fielding questions from Huw Jones from the Landsdowne Centre For Electronic Arts and Dave Bausola from Ixi). I managed to meet Harold and briefly say hello beforehand.I can’t do justice to the talk in a weblog entry, so it’s fortunate that a video of the talk will soon be online here. The talk ranged over Harold Cohen’s introduction to computing, the potential of technology, the artistic demands of expertise and technological innovation in an “idiot-proof” age, the struggle to explain colour to AARON, how AARON’s latest work came about, and more.
AARON’s latest work is painting-based with a wider range of objects, shadows, and more expressive virtual “brushwork” as well as a new palette made possible by its new archival quality large-scale inkjet printer. It’s strange seeing AARON’s plants and other objects without their traditional black outlines, but AARON is now truly painting rather than colouring drawings. I do wonder about “simulating” the layering of opaque colour to be rendered flat by an inkjet printer: this feels different from the physical “proofs” AARON has produced previously, but the results certainly look good and the effect could be achieved with real opaque paint anyway. I think I missed several minutes of the talk just looking at the slides of the new images as they were shown. I’ll need to check that archive myself to see what I missed. ๐
-
Generative Philosophy
Looking at Bayesian-filter-busting spam email, I had an idea. Train a genetic algorithm to make Markov chains. Train the chain on Project Gutenberg’s philosophical texts. Train a spam filter on the same texts. Then evolve philosophical texts, evaluating them by keeping the ones that the filter catches.
Generative philosophy…