[ExI] AI through human homogeny?

[Anders Sandberg]

On 05/01/2013 03:41, Will Steinberg wrote:
>
> The AI already exists by necessity, it is called Gaia and its 
> thoughtforms are inaccessible to us, for we are their fibers.
>

That is not certain. By the above claim, employees (or CEOs) of 
corporations would be unable to know what their organisation is 
"thinking". While this no doubt happens, it is not a necessary state. 
Consider the man in the Chinese Room thought experiment: he could 
actually happen to understand the mechanisms around him (including 
knowing Chinese and translation theory!) He might not be able to take in 
the full bandwidth of what the room is "thinking", but he could 
understand parts of its activity. This is more obvious when considering 
a simpler case, like doing numeric calculations by hand.

There is also the question whether Gaia is thinking in any deep sense. 
The universe is thinking since it contains thinking subsets and 
information flows between different parts, but I think it is a stretch 
to say that it is doing any cognition. Gaia is the same: there doesn't 
seem to be any feedbacks so far that are more easily explained as 
top-level activities than the result of the parts (compare that to a 
human, whose behavior is much easier to explain in terms of a mind with 
goals than a bundle of neuron interactions). Those who think there are 
regulatory top level processes should check out Ward's "The Medea 
Hypothesis" for some counterexamples.

-- 
Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University