Archive for December, 2008|Monthly archive page
Another Evolutionary Painting
Leave a CommentBlog posts are like buses…when they come they come in clusters of the same kind or number.
Hot on the heels of computer-generated algorithm to design – here’s yet another manifestation – this time the challenge was to paint as close to Mona Lisa (or for that matter, any other image) as possible simply by using no more than 50 semi-transparent polygons.
As described:
We start from random 50 polygons that are invisible. In each optimization step we randomly modify one parameter (like color components or polygon vertices) and check whether such new variant looks more like the original image. If it is, we keep it, and continue to mutate this one instead.
You can see that by iteration 904314, it’s very close to the original (or probably about as close as it can get for just 50 overlapping polygons).
Go on, play with it yourself here and see how it is done (you can use your own pictures too!), and the explanation (source code coming soon too) is here.
Car Design by Evolution
Back in school I used to have a professor who taught us about design+genetics (and called it Genometrics). The central thesis was for designers to move away from the role of designing the object to designing the paramaters/rules in which the object can exist. In a one-semester exposure this was nothing much more than programming parametric CAD software to churn out hundreds of designs based on a series of randomly varying dimensions (within reasonable bounds).
So we were supposed to find an object, program a range for a core set of dimensions, and let it be randomized within these bounds. Due to the ‘law’ within the programming, the outcome is bound to be varied and yet have identifiable ‘genes’. For instance, here are some stool designs (not necessarily valid) that were executed by the computer:
I remember the majority of the class balked at the idea. Some of the reasons include:
“So what does that make me? I’m here to learn design – if the computer does everything, then what’s the point?” – the same was said for a lot of other things that are taken-for-granted design tools for designers nowadays too.
“How can the computer make good designs – it has no brains/intelligence?” – Well maybe not in 100 iterations – but what about in 1000? 1 million? 1 billion?
I was somewhat sceptical too, but the idea of ‘genes’ captured my imagination. The idea that you can boil a cacophonic, complex external object (or even systems), and distill it into its essence with just a few variables. However, the shortcomings of the above exercise lied in the fact that at the end of the day, the judgement for ‘good design’ is subjective and human. This readily makes the computer seem incompetent.
A contrasting case-in-point:
This is a Flash program by Matthew where a primitive car design is iterated by computer. The (objective) aim of the car (that defines whether it’s a good design or not) is the length of treacherous terrain it can go pass before crashing. The variables are the size and initial positions of the 4 circles, the length, spring constant and damping of the 8 springs.
If you let it run, you’d see that as it crashes, it reboots and tries to refine the design again, and through time, the design gets improved without further manual input.
Here is the difference – with a quantifiable, objective feedback to the success of a design, computers can automate and rapidly refine designs (very likely) better than a human can. If the evaluation is subjective, however, the process becomes ineffective or slows down by orders of magnitude.
In a landscape where we are increasingly talking about user-generated content, democratic design and increased semantics intelligence for computers, this may become more relevant. There are already web-advertisements that modify its own designs (font size, colors, images, etc.) on-the-fly based on real time feedback on click-through-rates.
How/where else can this be applied?
NASA: A case for the Moon
NASA’s internal Powerpoint presentation style was critiqued by Edward Tufte who had even apportioned some responsibility to poorly structured/presented information in Powerpoint for the Columbia crash. Powerpoints like those are certainly not easy (at least for the layman) to read and understand – perhaps those decks really needed rocket scientists to decipher.
A rare gem coming off NASA is this presentation pressing for Moon exploration, though:
In the 100+ slides, “Why the Moon?” compares lunar exploration to the expedition to Antartica last century, drawing a case for continued exploration on the moon. Quite an interesting and inspiring read – PDF though.
Reactogon – New Music Visualizer
Hot on the heels of music and new ways of visualizing them – here’s reacTogon, a multi-touch music arpeggiator. Watch the video to find out how it works!
Auditorium – the Music Game
Have a go at Auditorium – like many web-based Flash games it is based on a simple concept and features slickly-executed graphics. The concept that sets it apart from other Flash games is the focus on the audio sensory: you are supposed to solve puzzles by guiding streams of particles to the goals – in the process making music.
Quite interesting!
The art of Keys Arrangement
If you’ve ever wondered how they transited from rotary dialing to button-press on the telephone, here’s an interesting background story. Before it became a standard that every phone now follows, human factor specialists (or the equivalent in those time) actually tested 18 different possible key layouts:
The participants were asked to key in a bunch of numbers and timed for it. Other factors like aesthetics and error rates were also computed. The five finalists were as follows:
Some pretty mixed results there actually. The familiar layout of the predecessor (the 4th one – ‘Telephone’) scored the best on timing, most likely due to the inherent familiarity. To be honest though I don’t quite know how they chose the (3×3)+1 arrangement that we now have though. From the data it seems it could’ve really gone any way. Perhaps because it’s a less polarizing option?
Now, someone needs to explain why the arrangement is different on calculator numberpads.
And while we’re on the topic of key arrangements – here’s a different but similarly intriguing story about the placement of the arrow keys: how did they come to the arrangement that is standard across all keyboards today? It turns out that there were testing and studies too:
Check it out here.
[via mental floss]
A milestone…
Would just like to take a break from regular program to mark this milestone – where the blog has crossed 1 million page views. Big sites may get these stats in a day, a week, or a month, but Gems Sty grows organically and slowly – it’s been just about 27 months from the first post.
Many thanks to loyal readers, occassional browsers and all – I hope Gems Sty has brought something to you. And if you have any suggestions or wish to see more/less of some stuff, do shout out in the comments too!
MUJI Award 03
The MUJI Award 03 results have been released – 1986 entries vying for the prize within the theme of “Found”:
For the coming MUJI Award 3, we want you to find “Found MUJI” from your viewpoint. Learn from the wisdom accumulated by our predecessors all over the world, find good points in such long-established merchandise, and convert them into a design that fits our modern life. We expect to see your “yes, of course” products that are also great in the modern age.
Here’s the Gold Prize winner – Straw straw:
That’s exactly what it is – a straw made of straw. Straws have long been replaced in manufacturing by extruded plastic. Straw Straw asks the question “why emulate nature with artificial plastics when nature’s own solution has already been there all along?”, returning back to the natural simplicity of a straw Straw – a reunion of natural material and form.
As judge Masaaki Kanai points out, his reaction was “That’s it – it’s exactly what it is?!”, and probably many are thinking that there is nothing particularly new nor inventive (straw straws were used before plastic straws were used). One might be expecting a stunningly clever new contraption that does something like never before, and be disappointed/confused by the choice for the Gold Prize.
For me it’s symptomatic of the society’s (or at the very least, MUJI and MUJI’s designer-judges) march towards not creating more things. And aptly for this year’s theme of “Found” – seeking the things that have been forgotten, buried, became niche…and restoring it to the consciousness of the mainstream. In that respect, I thought this was a deserving entry.
Silver Prize winner is Trash pack for outdoors – taking advantage of the natural structural properties of a pyramid (inspired by the shape of milk packaging commonly found in Japan) for stability, eliminating the need for additional layers of structure particularly for outdoor activities like picnics or camping.
For this concept, I do wonder a bit about disposal though. As the paper bag gets filled, how does one seal it? Paper bags or plastic bags have the natural handles that turn into tying mechanisms – how about this? It doesn’t seem to show through in the award images.
The Bronze awards are: Tachia Mat (hand-weaved straw that can be used as bedsheet); Grandpa’s Nail Hook (nails that have been designed to allow you to pound at it at a specific angle for hanging); Camelia washing-up Powder (byproduct powder from oil-production used as natural washing detergent); Second Skin (double layered cloth: one side soft and the other side waterproof, for flexible usage as towels, bags, poncho, etc.) and Precise Staper (a stapler that helps you align your paper for consistently accurate stapling).
Any comments on the winning entries?
And, I’m seeing if any of our readers participated in the MUJI design but didn’t win – do send me the concepts: I’d collect and publish them here so you can share your design with more people, and maybe get some nice feedback on them! Send them over to sunboar@gmail.com.
