Lab notes Rovelli 11/11

I'm in my lab to get my head around #quantumcomputing and using Carlo Rovelli's The Order of Time to melt my brain to an appropriate state

"There is no need in any of this to choose a privileged variable and call it 'time'. What we need, if we want to do science, is a theory that tells us how the variables change with respect to each"

Hence my droning insistence on articulating a theory of change

"The world without a time variable is not a complicated one. It's a net of interconnected events where the variables in play adhere to probabilistic rules, which we know for a good part how to write"

We're starting to edge into the important stuff. Probability is the language of uncertainty and the value of information is the reduction of uncertainty for important decisions. We can get into the quantum properties of data via the rules of probability

"The dynamic of these interactions [of the elementary grains of the universe] is probabilistic. The probability that something will happen - given the occurrence of something else - can in principle be calculated with the equations of the theory [of quantum gravity]"

Aka Bayesian inference, which needs to be somewhere in your data playbook

"We cannot draw a complete map, a complete geometry, of everything that happens in the world, because such happenings - including the passage of time - are always triggered only by an interaction with, and with respect to, a physical system involved in the interaction"

I've left a bit too much out here. But basically it means that the big data approach of swallowing up as much data as you can isn't going to get you past a certain point

"The world is like a collection of interrelated points of view. To speak of the world 'seen from outside' makes no sense, because there is no 'outside' to the world"

This is taking us to a place where asking 'what does the data say' is meaningless: the 'data' is less a recording of the world as it is a recording of the point of view of the recorder. Stance dependence becomes critical.

This is at the heart of rights issues around data: much of the data (well, at least here in New Zealand) that we 'have' consists of the viewpoints of white people. This leads us to  point where the search for objectivity ends up telling us less about the world

"If we give a description of the world that ignores point of view, that is solely 'from the outside' - of space, of time, of a subject - we may be able to say many things but we lose certain crucial aspects of the world. Because the world that we have been given is the world seen from within it, not from without"

This refers to indexicality, one of the crucial properties of data

"... the characteristic of certain words which have a different meaning every time they are uses, a meaning determined by where, how, when and by whom they are spoken... these indexical phrases make explicit reference to the fact that point of view exists, that a point of view is an ingredient in every description of the observable world we make"

Hence Melvin Kranzberg's first truism about the role of technology in society: 'technology is neither good nor bad; nor is it neutral'. What does this have to do with data? Data doesn't exist independently of our desire to obtain it, and we obtain it with technology.