Work in Progress


We tend to think of technology in terms of classical machines or computers. Neither are straightforward concepts.


It is not always clear to me what people mean by ‘a machine’. One notion is that an idealised machine is a ‘simple system’ in the sense that we can form a straightforward view of how it ‘should’ work:

  1. Has a definite boundary with definite ‘inputs’ and ‘outputs’.
  2. That it responds to its inputs according to some fixed classifying schema.
  3. That it only takes account of a limited range of factors (or ‘detail’) in its inputs, treating equally whatever seem the same.
  4. That its outputs are constrained according to definite rules linking outputs to inputs.
  5. That usually the either the outputs are either determined by the inputs or are determined stochastically by the inputs and perhaps ‘hidden variables’. (or at least they ‘make sense’ to us in some such way.

Engineered machines are usually intended to be idealised as a whole, and kept that way by engineers. As such, uncertainty plays no role unless and until they go wrong.


Computers can emulate any machine, and can change their characteristics, albeit in some ‘programmed’ way.

The following quotes from a pioneer, [Ada Lovelace ], illustrate some of our confusion:

[Computers] might act upon other things besides number, were objects found whose mutual fundamental relations could be expressed by those of the abstract science of operations, and which should be also susceptible of adaptations to the action of the operating notation and mechanism ..

The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform..

There is too much tendency to making separate and independent bundles of both the physical and the moral facts of the universe. Whereas, all and everything is naturally related and interconnected. A volume could I write you on this subject.

I may remark that the curious transformations many formulae can undergo, the unsuspected and to a beginner apparently impossible identity of forms exceedingly dissimilar at first sight, is I think one of the chief difficulties in the early part of mathematical studies. I am often reminded of certain sprites and fairies one reads of, who are at one’s elbows in one shape now, and the next minute in a form most dissimilar.

A sublety is that we can program a computer to adapt its categorisations, factors, rules, sometimes with unexpected results. That is, there is an element of uncertainty.

Alan Turing discusses this further in his ‘test‘.

See Also



Dave Marsay

%d bloggers like this: