He Let People Draw on a Computer, and Created a Whole New World | This Is How Our Lives Are Designed

The cases of CAD misuse suggest that when the hands and the brain are separated, it is the brain that suffers.

In 1960, a 22-year-old named Ivan Sutherland, then a doctoral student in electrical engineering at MIT, badly wanted to draw something on the huge computer in his lab.

He turned the idea into his doctoral thesis and wrote the world's first drawing program, Sketchpad. At the time he had no idea that this thesis and this program would go on to shape the structure of our houses, the shape of our cars, the interfaces on our computers, even the way furniture and smartphones are made today. Without his invention and his theory, Steve Jobs might have needed far longer to work out the human-computer interaction of the Macintosh, and we might never have watched the animation Pixar makes.

After Sketchpad appeared, the design profession — which perhaps owes part of its standardization as a profession to it — entered the age of Computer Assisted Design, the age of CAD tools.

How did that happen?

After World War II, in the second half of the 1950s, MIT received large sums from the U.S. Air Force to develop military computer systems. MIT's Lincoln Laboratory installed the TX-2, then the most advanced computer in the world and the one with the largest storage. It had a 9-inch screen that could be worked with a light pen, taking input and giving output as images.

Without the TX-2, Sutherland would never have had the chance to design Sketchpad, because ordinary computing machines at the time still needed a stack of punch cards for any exchange between human and machine.

Ivan Sutherland was born into a middle-class family in Hastings, Nebraska. His father was a civil engineer. By one account, Sutherland once said in an interview that before third grade he had dyslexia and could not spell correctly, and that this shaped his strong interest in drawing and in visual expression. The source of that account cannot be verified. But Sutherland has on many public occasions described himself as a "visual thinker."

In high school, Sutherland met a computer for the first time: Simon, one of the world's earliest personal computers. He and his brother used punch cards to write a few simple programs. At 19 and again at 21, he published papers on computer programming in professional journals.

After a master's degree in electrical engineering at the University of California, he went to MIT as a doctoral student, where he was advised by the mathematician Steve Coons and by Claude Shannon, the creator of information theory. He joined the computer-aided design project MIT was then developing, which aimed to simplify the way people and machines deal with each other.

But what really pushed Sutherland to design a drawing program was watching his colleagues play Spacewar!, one of the earliest video games, made by the computer scientist Steve Russell. It ran on the PDP-1, a computer that also had a round screen. This simple combat game was passing from campus to campus and lab to lab at the time.

Spacewar!'s star-map backdrop, the gravity built in around the stars, the way the fight was controlled — these gave Sutherland the idea that he could design a real-time drawing system along the same lines.

He gathered his research into a paper titled Sketchpad: A Man-Machine Graphical Communication System, along with a video he recorded himself showing Sketchpad in use.

In the talk on his thesis, Sutherland showed not only a new way of drawing computer graphics but also proposed a new way of operating a computer: graphical human-computer interaction.

Sketchpad pushed the computer-aided design project forward quickly. A new discipline, computer graphics, was born out of it.

But its spread in so short a time owed a great deal to a piece of marketing by MIT: the institute made more than 200 copies of the video of Sutherland using Sketchpad, along with the project team's findings, and mailed them to research institutions, news outlets and companies in the industry.

On the tape, people saw a young man at the small screen of an enormous computer, drawing straight lines, dashed lines, curves and circles with a light pen. As he turned the controls, the shapes changed to match his instructions. Zoom in and out, delete, undo, copy, paste — every drawing function was there.

This material changed how Americans imagined design and production.

Even so, because advanced mainframes were expensive at the time, Sketchpad was used at first mainly by large industrial and military manufacturers — carmakers, aircraft makers — as a tool for designing parts and product shapes. Ford Motor Company and the aerospace manufacturer Lockheed, for instance, both invited members of MIT's computer-aided design team to help develop CAD programs of their own.

As graphical human-computer interaction won wider acceptance, Europe and the United States stopped seeing the computer as a machine of war and began to see it as an obedient "slave" or a "working partner." But at the time that change of mind amounted to little more than treating the computer as a substitute for pen and paper.

A prodigy, Sutherland rose fast. Sketchpad made him a member of the U.S. Advanced Research Projects Agency (ARPA). He then went to teach at Harvard, where he published a paper called "The Ultimate Display," setting out the basic idea of a virtual reality system that combined interactive graphic displays, force-feedback devices and sound cues. It was the beginning of virtual reality technology.

His years teaching at the University of Utah laid a still more important foundation for modern computer graphics. He trained a generation of masters of the field who shaped the United States and shape our lives today: Alan Kay, the pioneer of the graphical user interface; Edwin Catmull, a founder of Pixar; and John Warnock, a founder of Adobe.

Sutherland loved computers and he loved geometry, and he was keen to put the two together. Alan Kay once asked him how he had solved such a complicated problem in computer graphics in so little time. Sutherland's answer: "Because I didn't know it was hard."

In the Sketchpad thesis he did something original: he compared Sketchpad, a computer tool, to the act of design itself, and in doing so gave computer-aided design a precise definition. "We might regard the process of drawing on Sketchpad as the design process itself — laying out many points and lines is like the rich creativity that arises in designing; but the finite geometric shapes those points and lines form are also like the finite variety that design constraints can bring to a product's value."

He went on: "How well Sketchpad's geometric shapes can satisfy a designer's requirements is equivalent to how effectively a good designer can use limited materials and limited cost to make a design that meets all kinds of needs. In fact, designers in many fields today do nothing but draw. When such design conditions are added to Sketchpad's list of constraints, the computer can not only help the user produce beautiful drawings but also help people arrive at safer, more reasonable designs."

Sutherland drew a line between "building something on a computer" and "drawing something on paper," and at the same time pointed to the endless conflict between the variety design can express and the limits under which a design has to be carried out.

Daniel Cardoso Llach writes in Builders of the Vision: Software and the Imagination of Design: "That group of MIT computer-aided-design researchers turned the process of Design into Building, which gave rise to a notion — that the design process becomes a series of structured information-management processes, letting a designer infer, for a work of design, a form that does not yet exist."

Once personal computers took off, that shift brought design both blessing and harm.

The 1990s were the decade when CAD came into large-scale use in the design industry. Design masters including Marc Newson, Philippe Starck and Jasper Morrison all began trying out computer drawing to make product designs.

As computers grew more powerful, the engineering errors that slide rules and calculators used to produce shrank, and the boxy forms that strict building codes used to dictate were broken at once.

The master architect Frank Gehry, designing the Guggenheim Museum Bilbao in Spain, used CATIA, a 3D parametric modeling system first developed by Ford Motor Company and later improved and released by France's Dassault Group. He was an architect who loved computer-aided design; he even developed a digital modeling system of his own and founded a technology company, Digital Project. Jacques Herzog used that same system to design Beijing's "Bird's Nest" stadium.

In the past, to be safe, engineers had to design to a fixed set of rules. But the Japanese architect Junya Ishigami, building the KAIT Workshop at the Kanagawa Institute of Technology, used a CAD program he wrote himself to design a glass house that broke the traditional grid arrangement of columns. Its 305 columns stand scattered irregularly through the building, and no two have the same cross-section. It is a wholly rule-breaking, experimental piece of design; without CAD, such a building could not have been built.

Then there is the other side. Computer graphics and computer-aided design have changed the world, but CAD has gradually turned into a "way" of designing rather than a pure tool. (Architecture schools do their best to have students treat CAD as a tool, but it can do far too much.) So architects, designers and cultural scholars alike sound a little wary when they get to this point.

Richard Sennett, a professor of sociology and anthropology, writes in The Craftsman: "Modern computer programs can indeed learn from their own experience by updating their form, because algorithms can rewrite themselves from feedback data. But the trouble is that people may leave the learning to the machine and merely witness and consume this ever-expanding capability, rather than take part in it. The cases of CAD misuse show that when the hands and the brain are separated, it is the brain that suffers."

One example is the Airbus A380 project of more than a decade ago. As one of the most complicated vehicles ever designed, the A380 plainly could not have been designed or built without CAD tools. But the project was delivered two years late and ran several billion dollars over budget because of a problem in the design of its internal wiring.

It started when engineers building the first A380 prototype in Toulouse, France, found that wiring already in production to the design specification could not meet the demands of the plane's complicated internal wiring-integration system. The wires were all too short.

It took Airbus management several rounds of internal review to work out why: the design and manufacturing project involved 16 teams in four countries, different design teams used different CAD software, and the calculation of the bend radius for wires running through the fuselage came out differently in different programs.

In the end, 1,100 engineers had to camp out overnight in Toulouse trying to fix the mistake.

Polly Brown, a researcher at Stanford University, points out in a paper that before CAD tools came into general use, designing was complicated and full of repetition: getting from an idea to a physical object meant going back and forth among sketches, drawings, models and a final model before production. After CAD, drawing and revising a design became very easy, and designing shrank to a few basic steps: idea, CAD drawing, model, production.

That difference is part of why a whole line of CAD software succeeded commercially — people could use it to turn out beautiful drawings that hid the problems that would show up in construction or in use.

Apple's design director Jony Ive complained about this, mildly, at a discussion at the Design Museum in London: "We've now interviewed too many designers who know how to design but don't know how to make. Because obviously, the creative workshops at school are expensive, but a computer is cheap. And CAD can make a bad design look acceptable."

Henry Petroski raises a different worry in Design: Human Nature (《设计，人类的本性》): "Because computers are cheap and computer software is easy to get, young engineers lack judgment about the designs the computer calculates." That casts doubt on the safety margins of products; it also, because copying became easy and cheap, led to wholesale copying of product drawings and a flood of knock-offs.

"What do we use to measure whether a piece of work is excellent? One standard is how a thing ought to be done; the other is whether the work is useful. That is the difference between correctness and functionality. Ideally the two do not conflict; in the real world they often do," says Richard Sennett. The craftsmen who do the making, he argues, think more fully about how a product is made and used, yet they usually have no part in designing it. "This is exactly where the problem of skill lies: the separation of the brain from the hand is not for lack of knowledge; it is made by society."

Cover image from: Pinterest