[iDC] more on MTurk
geoff cox
gcox at plymouth.ac.uk
Tue Jun 16 17:16:48 UTC 2009
Hi
Following the digression of Ulises Mejias, I thought I'd jump in with a
fuller context of the Mechanical Turk, that opens Benjamin's 'On the
Concept of History' thus:
'The story is told of an automaton constructed in such a way that it
could respond to each move in a game of chess with a countermove that
ensured him victory. A puppet in Turkish attire, and with a hookah in
his mouth, sat in front of a chessboard placed on a large table. A
system of mirrors created the illusion of a table transparent from all
sides. Actually a hunchback dwarf, who was an expert chess player, sat
inside and guided the puppet's hand by means of strings. One can imagine
a philosophical counterpart to this device. The puppet known as
'historical materialism' is always supposed to win. It can easily be a
match for anyone if it ropes in the services of theology, which today,
as the story goes, is small and ugly and must, as it is, keep out of
sight.' (Esther Leslie translation of Benjamin, from _Walter Benjamin:
Overpowering Conformism_, 2000)
History, and the history of technology, is full of the use of trickery
to make it seem beyond the scope of human intervention.
The figure of the dwarf is rather obscure, linked to technology and
theology - evoking the labour of the operator - and that the success of
the automaton is contingent on the recognition that the dwarf has to
gain control of the technology. It's a complex allegory but an
investigation of the history of the chess playing machine reveals more
detail and the relative roles of puppet, puppeteer and opponent.
Benjamin is drawing upon a well-known example of The Turk, a
chess-playing automata built by Wolfgang von Kempelen in 1769. It
received widespread attention, and there was much speculation as to
whether the machine was driven by magic or by some other illusory device
- a spectre or demon. Part of the presentation involved Kempelen
demonstrating the clockwork mechanism beneath the automaton, opening
doors to compartments of the desk one by one and revealing what lay
beneath the Turkish attire (engaging Orientalist fantasies of the time):
the 'automaton stripped naked'.
After Kempelen's death, Johan Nepomuk Maelsel added some improvements
including speech - the announcement of 'échec' (check) by means of
bellows. It is this version that Norbert Wiener refers to as a
'fraudulent machine' in his note on the accomplishment of artificial
intelligence - part of the trick was that a machine could demonstrate
intelligence sufficient to play chess and speech is used to authenticate
intelligence. In this connection, the writer Edgar Allan Poe, compared
the chess automata to Charles Babbage's calculating machine, asking what
to think of a machine that operates, 'without the slightest intervention
of the intellect of man? It will, perhaps, be said in reply, that a
machine as we have described is altogether above comparison with the
Chess Player of Maelzel. By no means - it is altogether beneath it -
that is to say, provided we assume (what should never for one moment be
assumed) that the Chess Player is a pure machine, and performs its
operations without any immediate human agency.'
For Poe, machine-like agency simply serves to conceal the underlying
operating system. In Benjamin's allegory, the puppeteer appears to be in
the service of the puppet - suggesting perhaps that it is not the
machine that is life-like but that the human figure is machine-like
unless action is taken to correct the illusion (the position of the
historical materialist). This seems to concur with Esther Leslie in that
the dwarf has to gain control of the technology as it is the autonomy of
the machine that is fake. The theatrics simply reveal how technology
masks the underlying processes.
Pretending to reveal the actual mechanism has become an orthodoxy and
indeed part of the illusion itself of the interface, masking the labour
of people and machines.
Amazon's Mechanical Turk is a willful example of a similar trick.
--
geoff cox
http://www.anti-thesis.net/work
More information about the iDC
mailing list