0:00:00.470,0:00:02.919 As an architect, I often ask myself, 0:00:02.919,0:00:05.759 what is the origin of the forms that we design? 0:00:05.759,0:00:08.664 What kind of forms could we design 0:00:08.664,0:00:11.487 if we wouldn't work with references anymore? 0:00:11.487,0:00:14.872 If we had no bias, if we had no preconceptions, 0:00:14.872,0:00:17.079 what kind of forms could we design 0:00:17.079,0:00:18.679 if we could free ourselves from 0:00:18.679,0:00:20.815 our experience? 0:00:20.815,0:00:25.549 If we could free ourselves from our education? 0:00:25.549,0:00:28.576 What would these unseen forms look like? 0:00:28.576,0:00:32.287 Would they surprise us? Would they intrigue us? 0:00:32.287,0:00:34.702 Would they delight us? 0:00:34.702,0:00:38.623 If so, then how can we go about creating something that is truly new? 0:00:38.623,0:00:41.015 I propose we look to nature. 0:00:41.015,0:00:45.279 Nature has been called the greatest architect of forms. 0:00:45.279,0:00:48.867 And I'm not saying that we should copy nature, 0:00:48.867,0:00:51.247 I'm not saying we should mimic biology, 0:00:51.247,0:00:54.695 instead I propose that we can borrow nature's processes. 0:00:54.695,0:00:59.149 We can abstract them and to create something that is new. 0:00:59.149,0:01:03.027 Nature's main process of creation, morphogenesis, 0:01:03.027,0:01:06.856 is the splitting of one cell into two cells. 0:01:06.856,0:01:08.913 And these cells can either be identical, 0:01:08.913,0:01:11.088 or they can be distinct from each other 0:01:11.088,0:01:13.152 through asymmetric cell division. 0:01:13.152,0:01:16.951 If we abstract this process, and simplify it as much as possible, 0:01:16.951,0:01:19.184 then we could start with a single sheet of paper, 0:01:19.184,0:01:21.680 one surface, and we could make a fold 0:01:21.680,0:01:24.616 and divide the surface into two surfaces. 0:01:24.616,0:01:26.879 We're free to choose where we make the fold. 0:01:26.879,0:01:31.789 And by doing so, we can differentiate the surfaces. 0:01:31.789,0:01:33.683 Through this very simple process, 0:01:33.683,0:01:37.295 we can create an astounding variety of forms. 0:01:37.295,0:01:40.020 Now, we can take this form and use the same process 0:01:40.020,0:01:42.012 to generate three-dimensional structures, 0:01:42.012,0:01:44.420 but rather than folding things by hand, 0:01:44.420,0:01:47.005 we'll bring the structure into the computer, 0:01:47.005,0:01:49.980 and code it as an algorithm. 0:01:49.980,0:01:52.988 And in doing so, we can suddenly fold anything. 0:01:52.988,0:01:55.147 We can fold a million times faster, 0:01:55.147,0:01:58.308 we can fold in hundreds and hundreds of variations. 0:01:58.308,0:02:00.788 And as we're seeking to make something three-dimensional, 0:02:00.788,0:02:04.004 we start not with a single surface, but with a volume. 0:02:04.004,0:02:05.357 A simple volume, the cube. 0:02:05.357,0:02:07.077 If we take its surfaces and fold them 0:02:07.077,0:02:09.133 again and again and again and again, 0:02:09.133,0:02:11.996 then after 16 iterations, 16 steps, 0:02:11.996,0:02:15.845 we end up with 400,000 surfaces and a shape that looks, 0:02:15.845,0:02:18.380 for instance, like this. 0:02:18.380,0:02:21.092 And if we change where we make the folds, 0:02:21.092,0:02:22.733 if we change the folding ratio, 0:02:22.733,0:02:26.156 then this cube turns into this one. 0:02:26.156,0:02:30.295 We can change the folding ratio again to produce this shape, 0:02:30.295,0:02:32.122 or this shape. 0:02:32.122,0:02:33.745 So we exert control over the form 0:02:33.745,0:02:37.249 by specifying the position of where we're making the fold, 0:02:37.249,0:02:41.955 but essentially you're looking at a folded cube. 0:02:41.955,0:02:43.019 And we can play with this. 0:02:43.019,0:02:45.635 We can apply different folding ratios to different parts 0:02:45.635,0:02:48.380 of the form to create local conditions. 0:02:48.380,0:02:50.323 We can begin to sculpt the form. 0:02:50.323,0:02:53.211 And because we're doing the folding on the computer, 0:02:53.211,0:02:56.907 we are completely free of any physical constraints. 0:02:56.907,0:02:59.858 So that means that surfaces can intersect themselves, 0:02:59.858,0:03:01.347 they can become impossibly small. 0:03:01.347,0:03:04.948 We can make folds that we otherwise could not make. 0:03:04.948,0:03:06.826 Surfaces can become porous. 0:03:06.826,0:03:09.555 They can stretch. They can tear. 0:03:09.555,0:03:14.190 And all of this expounds the scope of forms that we can produce. 0:03:14.190,0:03:16.940 But in each case, I didn't design the form. 0:03:16.940,0:03:21.541 I designed the process that generated the form. 0:03:21.541,0:03:26.029 In general, if we make a small change to the folding ratio, 0:03:26.029,0:03:27.624 which is what you're seeing here, 0:03:27.624,0:03:31.397 then the form changes correspondingly. 0:03:31.397,0:03:33.653 But that's only half of the story -- 0:03:33.653,0:03:37.976 99.9 percent of the folding ratios produce not this, 0:03:37.976,0:03:42.872 but this, the geometric equivalent of noise. 0:03:42.872,0:03:44.992 The forms that I showed before were made actually 0:03:44.992,0:03:46.640 through very long trial and error. 0:03:46.640,0:03:49.856 A far more effective way to create forms, I have found, 0:03:49.856,0:03:53.744 is to use information that is already contained in forms. 0:03:53.744,0:03:56.399 A very simple form such as this one actually contains 0:03:56.399,0:03:59.991 a lot of information that may not be visible to the human eye. 0:03:59.991,0:04:02.472 So, for instance, we can plot the length of the edges. 0:04:02.472,0:04:05.808 White surfaces have long edges, black ones have short ones. 0:04:05.808,0:04:09.385 We can plot the planarity of the surfaces, their curvature, 0:04:09.385,0:04:13.472 how radial they are -- all information that may not be 0:04:13.472,0:04:15.419 instantly visible to you, 0:04:15.419,0:04:17.844 but that we can bring out, that we can articulate, 0:04:17.844,0:04:21.220 and that we can use to control the folding. 0:04:21.220,0:04:23.179 So now I'm not specifying a single 0:04:23.179,0:04:25.355 ratio anymore to fold it, 0:04:25.355,0:04:27.907 but instead I'm establishing a rule, 0:04:27.907,0:04:30.343 I'm establishing a link between a property of a surface 0:04:30.343,0:04:33.316 and how that surface is folded. 0:04:33.316,0:04:36.267 And because I've designed the process and not the form, 0:04:36.267,0:04:38.589 I can run the process again and again and again 0:04:38.589,0:04:41.124 to produce a whole family of forms. 0:04:53.385,0:04:57.828 These forms look elaborate, but the process is a very minimal one. 0:04:57.828,0:04:58.957 There is a simple input, 0:04:58.957,0:05:00.909 it's always a cube that I start with, 0:05:00.909,0:05:04.485 and it's a very simple operation -- it's making a fold, 0:05:04.485,0:05:08.437 and doing this over and over again. 0:05:08.437,0:05:10.717 So let's bring this process to architecture. 0:05:10.717,0:05:12.317 How? And at what scale? 0:05:12.317,0:05:14.045 I chose to design a column. 0:05:14.045,0:05:17.421 Columns are architectural archetypes. 0:05:17.421,0:05:20.488 They've been used throughout history to express ideals 0:05:20.488,0:05:25.563 about beauty, about technology. 0:05:25.563,0:05:27.485 A challenge to me was how we could express 0:05:27.485,0:05:30.781 this new algorithmic order in a column. 0:05:30.781,0:05:33.765 I started using four cylinders. 0:05:33.765,0:05:37.541 Through a lot of experimentation, these cylinders 0:05:37.541,0:05:40.676 eventually evolved into this. 0:05:40.676,0:05:45.292 And these columns, they have information at very many scales. 0:05:45.292,0:05:47.810 We can begin to zoom into them. 0:05:47.810,0:05:51.365 The closer one gets, the more new features one discovers. 0:05:51.365,0:05:55.037 Some formations are almost at the threshold of human visibility. 0:05:55.037,0:05:56.989 And unlike traditional architecture, 0:05:56.989,0:05:59.773 it's a single process that creates both the overall form 0:05:59.773,0:06:05.118 and the microscopic surface detail. 0:06:05.118,0:06:07.809 These forms are undrawable. 0:06:07.809,0:06:11.185 An architect who's drawing them with a pen and a paper 0:06:11.185,0:06:12.977 would probably take months, 0:06:12.977,0:06:15.361 or it would take even a year to draw all the sections, 0:06:15.361,0:06:17.665 all of the elevations, you can only create something like this 0:06:17.665,0:06:19.801 through an algorithm. 0:06:19.801,0:06:21.833 The more interesting question, perhaps, is, 0:06:21.833,0:06:24.462 are these forms imaginable? 0:06:24.462,0:06:27.097 Usually, an architect can somehow envision the end state 0:06:27.097,0:06:28.961 of what he is designing. 0:06:28.961,0:06:31.786 In this case, the process is deterministic. 0:06:31.786,0:06:34.145 There's no randomness involved at all, 0:06:34.145,0:06:36.152 but it's not entirely predictable. 0:06:36.152,0:06:37.728 There's too many surfaces, 0:06:37.728,0:06:41.338 there's too much detail, one can't see the end state. 0:06:41.338,0:06:44.552 So this leads to a new role for the architect. 0:06:44.552,0:06:48.163 One needs a new method to explore all of the possibilities 0:06:48.163,0:06:49.939 that are out there. 0:06:49.939,0:06:53.171 For one thing, one can design many variants of a form, 0:06:53.171,0:06:55.385 in parallel, and one can cultivate them. 0:06:55.385,0:06:57.633 And to go back to the analogy with nature, 0:06:57.633,0:07:00.129 one can begin to think in terms of populations, 0:07:00.129,0:07:03.657 one can talk about permutations, about generations, 0:07:03.657,0:07:08.520 about crossing and breeding to come up with a design. 0:07:08.520,0:07:10.905 And the architect is really, he moves into the position 0:07:10.905,0:07:14.337 of being an orchestrator of all of these processes. 0:07:14.337,0:07:16.681 But enough of the theory. 0:07:16.681,0:07:19.208 At one point I simply wanted to jump inside 0:07:19.208,0:07:23.037 this image, so to say, I bought these red and blue 0:07:23.037,0:07:25.973 3D glasses, got up very close to the screen, 0:07:25.973,0:07:28.045 but still that wasn't the same as being able to 0:07:28.045,0:07:30.397 walk around and touch things. 0:07:30.397,0:07:32.300 So there was only one possibility -- 0:07:32.300,0:07:35.397 to bring the column out of the computer. 0:07:35.397,0:07:38.317 There's been a lot of talk now about 3D printing. 0:07:38.317,0:07:41.276 For me, or for my purpose at this moment, 0:07:41.276,0:07:44.358 there's still too much of an unfavorable tradeoff 0:07:44.358,0:07:51.141 between scale, on the one hand, and resolution and speed, on the other. 0:07:51.141,0:07:53.398 So instead, we decided to take the column, 0:07:53.398,0:07:55.837 and we decided to build it as a layered model, 0:07:55.837,0:07:59.965 made out of very many slices, thinly stacked over each other. 0:07:59.965,0:08:01.862 What you're looking at here is an X-ray 0:08:01.862,0:08:04.637 of the column that you just saw, viewed from the top. 0:08:04.637,0:08:06.637 Unbeknownst to me at the time, 0:08:06.637,0:08:09.013 because we had only seen the outside, 0:08:09.013,0:08:11.173 the surfaces were continuing to fold themselves, 0:08:11.173,0:08:13.437 to grow on the inside of the column, 0:08:13.437,0:08:15.983 which was quite a surprising discovery. 0:08:15.983,0:08:19.597 From this shape, we calculated a cutting line, 0:08:19.597,0:08:22.621 and then we gave this cutting line to a laser cutter 0:08:22.621,0:08:26.435 to produce -- and you're seeing a segment of it here -- 0:08:26.435,0:08:31.326 very many thin slices, individually cut, on top of each other. 0:08:33.480,0:08:36.293 And this is a photo now, it's not a rendering, 0:08:36.293,0:08:38.093 and the column that we ended up with 0:08:38.093,0:08:41.140 after a lot of work, ended up looking remarkably like the one 0:08:41.140,0:08:44.850 that we had designed in the computer. 0:08:44.850,0:08:46.971 Almost all of the details, almost all of the 0:08:46.971,0:08:50.011 surface intricacies were preserved. 0:08:52.626,0:08:54.972 But it was very labor intensive. 0:08:54.972,0:08:57.373 There's a huge disconnect at the moment still 0:08:57.373,0:09:00.133 between the virtual and the physical. 0:09:00.133,0:09:02.259 It took me several months to design the column, 0:09:02.259,0:09:04.977 but ultimately it takes the computer about 30 seconds 0:09:04.977,0:09:07.825 to calculate all of the 16 million faces. 0:09:07.825,0:09:09.786 The physical model, on the other hand, 0:09:09.786,0:09:13.994 is 2,700 layers, one millimeter thick, 0:09:13.994,0:09:18.141 it weighs 700 kilos, it's made of sheet that can cover 0:09:18.141,0:09:20.279 this entire auditorium. 0:09:20.279,0:09:22.367 And the cutting path that the laser followed 0:09:22.367,0:09:27.484 goes from here to the airport and back again. 0:09:27.484,0:09:29.233 But it is increasingly possible. 0:09:29.233,0:09:31.844 Machines are getting faster, it's getting less expensive, 0:09:31.844,0:09:34.604 and there's some promising technological developments 0:09:34.604,0:09:36.387 just on the horizon. 0:09:36.387,0:09:39.459 These are images from the Gwangju Biennale. 0:09:39.459,0:09:42.983 And in this case, I used ABS plastic to produce the columns, 0:09:42.983,0:09:44.837 we used the bigger, faster machine, 0:09:44.837,0:09:47.894 and they have a steel core inside, so they're structural, 0:09:47.894,0:09:50.870 they can bear loads for once. 0:09:50.870,0:09:52.885 Each column is effectively a hybrid of two columns. 0:09:52.885,0:09:56.294 You can see a different column in the mirror, 0:09:56.294,0:09:58.344 if there's a mirror behind the column 0:09:58.344,0:10:01.416 that creates a sort of an optical illusion. 0:10:01.431,0:10:03.262 So where does this leave us? 0:10:03.262,0:10:07.730 I think this project gives us a glimpse of the unseen objects that await us 0:10:07.730,0:10:11.613 if we as architects begin to think about designing not the object, 0:10:11.613,0:10:15.006 but a process to generate objects. 0:10:15.006,0:10:18.279 I've shown one simple process that was inspired by nature; 0:10:18.279,0:10:21.127 there's countless other ones. 0:10:21.127,0:10:24.574 In short, we have no constraints. 0:10:24.574,0:10:28.144 Instead, we have processes in our hands right now 0:10:28.144,0:10:32.725 that allow us to create structures at all scales 0:10:32.725,0:10:35.757 that we couldn't even have dreamt up. 0:10:35.757,0:10:40.576 And, if I may add, at one point we will build them. 0:10:40.576,0:10:47.199 Thank you. (Applause)