Computer Program Re-Discovers Laws of Physics

Posted on April 6, 2009

A computer program developed by Cornell researchers has re-discovered the laws of physics in less than two days, an accomplishment that took humans centuries. This is obviously a large step in Artificial Intelligence (AI) development. To make this discovery even more unsettling, the developers insist that the program deduced the natural laws without a shred of knowledge about physics or geometry.

Before going forward anymore, let's take a look at a few of the past attempts in AI development (list courtesy of Wired.com):

• Half a century ago, IBM's Herbert Gelernter authored a program that purportedly rediscovered Euclid's geometry theorems, but critics said it relied too much on programmer-supplied rules.
• In the 1970s, Douglas Lenat's Automated Mathematician automatically generated mathematical theorems, but they proved largely useless.
• Stanford University's Dendral project, was started in 1965 and used for two decades to extrapolate possible structures for organic molecules from chemical measurements gathered by NASA spacecraft. But it was ultimately unable to assess the likelihood of the various answers that it generated.
• The $100,000 Leibniz Prize, established in the 1980s, was promised to the first program to discover a theorem that "profoundly affects" math. It was never claimed.

The software was initially designed with only a few basic mathematical equations and understandings of equation checks and balances. At first there were more errors than promising results, but after a number of trial and error processes the system began to "learn" how to develop the correct equations from the start. "Turns out, some of these equations were very familiar: the law of conservation of momentum, and Newton's second law of motion."

Don't worry, just like when you learned long division in school and had the excessively harsh teacher scrutinizing every piece of your work there are a team of scientist continuously testing each equation that comes from this inanimate box of intelligence. So the same phrase that haunted you in school applies to this wonder program, "Be sure to show all work or you get no credit." This group of glorified math teachers presiding over the computational output are also needed to decide which direction the system should take next. In other words, what is important enough to continue work on and what isn't?

We can be certain that we are a long way from seeing any HAL-9000s dictating how humans should live their lives, but it is inevitable.