The Limits of Automatic Design
Today there are huge software packages available to help designers of all types: programmers, engineers, inventors, architects, artists, composers, and writers. The idea is to free the designer from mundane tasks and encourage unbridled creativity. Designers will make the soaring leaps of imagination, letting the design program ensure that their designs are scientifically feasible, technically sound, sufficiently strong, safe, reliable and economical.
Composers no longer have to copy their music laboriously onto music staffs, working out the details of harmony as they go. As soon as they have a new melody in their heads, all they have to do is play it on a piano keyboard connected to a computer. The score appears automatically on the screen, replete with standard harmony, ready for printout under the composer’s copyright.
Writers may give free flight to their fancy, and let the computer worry about spelling and grammar. Search engines automate the research, online encyclopedias verify the facts, and computerized anthologies supply the quotes.
Even mathematicians are finally learning that computers can help them to find mathematical proofs.
We still have not achieved a master system like the one science fiction writers described half a century ago. Once artificial intelligence reaches the level of creativity, some authors said that we should ask the computers how to build more powerful computers. The machines will then “lift themselves by their own bootstraps.”[i]
Engineers use machines to help them design and build new, more powerful machines. The present relationship between engineers and machines may merit the term “symbiotic.” Engineers and machines work together intimately in a mutually beneficial relationship. Together they are bootstrapping upward. However, at present the machines are not creative. They do not make great leaps of imagination or find new relationships between variables or synthesize new concepts. Hands must still lead the machines.
In the early 1960’s, engineers were just beginning to see the possibility of applying computers to their design tasks. The most adventurous of us learned to write programs in languages like FORTRAN (Formula Translation). Other programming languages came afterwards. Initial successes set some engineers to dreaming of the future. Hollywood immediately picked up the theme. In the future, according to the movies, we would only have to talk to a computer, and new designs would pop out, fully elaborated, automatically.
Success with programming languages led on to dreams about human languages. Isn’t it wonderful now that computers can do automatic translation? Soon we will achieve broad understanding between different cultures. Wars will cease, and everything will be peace and harmony.
This rosy picture is wonderful from a distance. Up close we notice a few flaws. In the last half-century we have discovered that there are limits to automatic design. It falls far short of human creative design.
[i] We “boot up” our computers when we start them. The start-up program is called the “bootstrap” because it loads itself. This, however, is nowhere close to the goal of getting computers to design better computers.
Composers no longer have to copy their music laboriously onto music staffs, working out the details of harmony as they go. As soon as they have a new melody in their heads, all they have to do is play it on a piano keyboard connected to a computer. The score appears automatically on the screen, replete with standard harmony, ready for printout under the composer’s copyright.
Writers may give free flight to their fancy, and let the computer worry about spelling and grammar. Search engines automate the research, online encyclopedias verify the facts, and computerized anthologies supply the quotes.
Even mathematicians are finally learning that computers can help them to find mathematical proofs.
We still have not achieved a master system like the one science fiction writers described half a century ago. Once artificial intelligence reaches the level of creativity, some authors said that we should ask the computers how to build more powerful computers. The machines will then “lift themselves by their own bootstraps.”[i]
Engineers use machines to help them design and build new, more powerful machines. The present relationship between engineers and machines may merit the term “symbiotic.” Engineers and machines work together intimately in a mutually beneficial relationship. Together they are bootstrapping upward. However, at present the machines are not creative. They do not make great leaps of imagination or find new relationships between variables or synthesize new concepts. Hands must still lead the machines.
In the early 1960’s, engineers were just beginning to see the possibility of applying computers to their design tasks. The most adventurous of us learned to write programs in languages like FORTRAN (Formula Translation). Other programming languages came afterwards. Initial successes set some engineers to dreaming of the future. Hollywood immediately picked up the theme. In the future, according to the movies, we would only have to talk to a computer, and new designs would pop out, fully elaborated, automatically.
Success with programming languages led on to dreams about human languages. Isn’t it wonderful now that computers can do automatic translation? Soon we will achieve broad understanding between different cultures. Wars will cease, and everything will be peace and harmony.
This rosy picture is wonderful from a distance. Up close we notice a few flaws. In the last half-century we have discovered that there are limits to automatic design. It falls far short of human creative design.
[i] We “boot up” our computers when we start them. The start-up program is called the “bootstrap” because it loads itself. This, however, is nowhere close to the goal of getting computers to design better computers.