The Practicality of an Impractical Education

Neil deGrasse Tyson, Astrophysicist and Director of the Hayden Planetarium, reminds us that “Art and Science have defined civilization since the beginning of time.” Biologist and Naturalist, Edward O. Wilson’s observation enfolds an explanation: “Although the two great branches of learning, Science and the Humanities, are radically different in the way they describe our species, they have risen from the same wellspring of creative thought.”

The current fascination with the STEM disciplines is to be expected in a technological and digital age, but supporting STEM at the expense of the Liberal Arts and Humanities ignores half of human thought.

While I embrace the idea that study in the Liberal Arts and Humanities contributes to intellectual growth, I am aware of complaints that asserting that the Liberal Arts and Humanities develop creativity, critical thinking, and mental flexibility is elitist. Therefore I seek a common ground by posing four questions.

1. 1. How does study in the Liberal Arts and Humanities develop intellectual growth?

2. 2. Do thinking skills gained through Liberal Arts and Humanities become practical?

3. 3. Why are the Liberal Arts and Humanities currently perceived to be irrelevant?

4. 4. How can we reestablish relevance of the Liberal Arts and Humanities in the minds of the public?

How does study in the Liberal arts and Humanities develop intellectual growth?

People enamored with STEM education accept a future based on robots and Artificial Intelligence, but designers of Artificial Intelligence examine human thinking processes to imitate them.

The initial and simplest form of ArtificiaI Intelligence is known as “answers in a box.” Like the digital help on your computer, such programs answer the questions its designers anticipated; it is inflexible; it does not learn. Like a human provided with a procedure or protocol, it repeats what it has been told.

Recently, Artificial Intelligence scientists have moved toward a tipping point. A team at the University of Maryland designed a procedure to teach a robot how to learn. Using You Tube videos of people making salads, the robot was shown humans picking up a knife, slicing a cucumber, etc. The intent was for the Intelligence in the robot to learn to perform those acts as well as what to watch for, such as gauging the thickness of the slice and removing a stuck slice from the knife. Watching is one way human apprentices learn. Yiannis Aliomonis, a scientist with the project, explained that the team gave the robot a challenge to prepare it to figure out new motions when facing new tasks.

Nick Bostrom, Director of the Martin Programme on the Impact of Future Technology at the University of Oxford is cautious about Artificial Intelligence that can learn. In a TED talk Bostrom argued that when Artificial Intelligence is able to learn beyond managing human assigned tasks, technology could make choices for itself. When Artificial Intelligence makes choices, it will employ the values of machines rather than humans. Machines with Artificial Intelligence could then control humans as humans now control chimpanzees, because human brains have a fixed size and the neural pathways, used or unused, exist. The mechanism which runs Artificial Intelligence can be endlessly expanded.

Both Artificial Intelligence and humans employ pattern recognition and what Edward de Bono calls “lateral thinking”: transferring information or observations acquired in one set of circumstances to another context. For instance, when the much celebrated Hubble telescope initially provided only blurred images from space, Jim Crocker, Chief engineer for NASA, was told the lens had been ground too fine. The telescope needed glasses, but installing such a large item in space was impossible. Later, in his shower, Crocker observed multiple streams of water from the shower head coalescing into one and realized that small corrective lenses could be installed on Hubble to refocus the blurred images into one clear picture.

Crocker applied an observation made in one environment or pattern into another vastly different one. Artificial Intelligence still obtains residual information and the task information from humans. Both the need to teach robots to...