Thursday, October 21, 2010

Beyond science

Over at the Shuttle on Hugh's blogs there's been discussion about science or to be precise at what point does science stop. I mentioned there the table/stool point - what's the difference between the two? If I sit on a table does it become a stool? In other words things that cannot be empirically dissected. I also mentioned the uncanny valley where the features of Homer Simpson seem more 'realistic' than the most sophisticated computer generated image.

In than vein here are the latest benchmarks from 3DMark that use cutting edge technology to render two scenes - a jungle temple and underwater exploration. They look great, but I doubt anyone watching them would mistake it for reality or even a filmed sequence. Why not?

Science provides a framework in which to function, yet some things don't quite slot in and just sticks out the side; left hanging there. It's possible we may find the connecting pieces, it's possible there are no connecting pieces, but if we just launch ourselves off without a tether it's meaningless.


Orphi said...

The 3DMark stuff is impressive, but fairly obviously computer-generated. The visuals are just “too perfect”. The jungle scene also suffers from the colours being “not quite right”, in some way that I can't pin down.

That said, the shots of the underwater rock wall look like they could plausibly be photographs. It's just the man-made stuff that doesn't quite look right. (For a start, the propellers have no motion blur, and the water is oddly crystal clear, despite yielding a dense haze from the lights…)

Reality is a complex thing which is very difficult to match. There are lots of CG images that somehow don't quite look right, even though it's difficult to say why. For that matter, there are real photographs that somehow look wrong (but obviously can't be).

As to “what is science”, I believe the official definition is “the study of theories that are falsifiable” (i.e., things that you could plausibly prove wrong if they weren't true).

For example, the statement “God exists” is not falsifiable; there is no logical argument nor physical experiment you can perform which will conclusively prove that God exists or does not exist. Therefore the question is beyond the relm of science. Note that this is not the same as saying that the statement is false or meaningless. It is just to say that it's not science.

(People like Michael Behe don't seem to understand this. I read with astonishment when his book proclaims that he “doesn't see why” science should limit itself to what is experimentally verifiable. I mean, you know, other than that being the definition of what science is…)

FlipC said...

"in some way that I can't pin down." Heh which is entirely my point :-)

As for science, why can't the definition be changed? Put it like this - is the definition of science itself subject to the question of science i.e. is it a falsifiable definition?