An art installation about the Entscheidungsproblem

Last night all my BTEC class did art installations. Mine involved 16 PCs all running a Flash movie and was named the Entscheidungsproblem.

The 16 computers all sit in a computer room, 8 along each wall. Its a long narrow room used for nothing else.
The same Flash movie was set to loop endlessly on each, with no attempt to synchronise the instances. (Quite the opposite in fact). The movie simply consists of the first 30 or so letters of Alan Turings essay, “ON COMPUTABLE NUMBERS, WITH AN APPLICATION TO THE
ENTSCHEIDUNGSPROBLEM”. These letters show one by one on the screen, and are each read aloud.

The Entscheidungsproblem was defined by Hilbert: broadly, it “asks for a computer program that will take as input a description of a formal language and a mathematical statement in the language and return as output either “True” or “False” according to whether the statement is true or false.” Turings paper, written in 1936, was one of the fundamental documents of the computer age. It is based around his concept of the Turing Machine, which was the first statement of the basic principles of the modern programmable computer.

The installation, although simple in concept, was surprisingly effective, producing a roomful of sounds and letters which were individually comprehensible but very difficult to relate or interpret. It was as if the computers were talking to each other, although using our language. It shows quite clearly that our brains need to chunk up data to understand it easily. The individual words would be meaningful, but the letters, one after another, require brain time to interpret, and the multiple channels override your ability to do this.

As part of my sketchbook I also submitted several entries from this blog, dealing with the large amounts of data that machines now hold, and my suspicions that our attempts to use machines to derive meaning from massive datasets doesnt always give us the results we think we are getting. The machines are saying one thing, but we think we are hearing another. Turings paper, of course, proved that a general solution to the problem is NOT possible – in other words, “it is impossible to decide algorithmically whether statements in arithmetic are true or false”. Theres a lesson there for heavy users of datasets.

I was very impressed by the high standards of installations prodiced by my course-mates. Almost all were good and three were simply brilliant. Its amazing how weve all grown in confidence, ideas and delivery skills since those first agonising attempts at drawing perspective or (my particular worst memory) covering an A3 sheet with charcoal and making a drawing by rubbing it out. If ever anything would convince you of the benefits of vector graphics, it is that classically grubby and futile art-school exercise.

I noticed one advantage of being a net artist. Whilst my colleagues were struggling upstairs with bags and cases, or earnestly stapling and glueing things together, I arrived with nothing at all, and simply logged on 16 machines to the same URL.

Whilst they produced large cumbersome objects that had to be destroyed at the end (unless someone has storage space for them!), mine survives, and can be recreated any time in a few minutes, by anyone with multiple PCs at their command.

Leave a Reply

Your email address will not be published. Required fields are marked *