Sci Fi TV science or fiction?

Auguste Compte 1842: "We can never learn the stars internal constitution, nor, in regard to some of them, how heat is absorbed by their atmosphere. We can never know anything of the planets chemical or mineralogical structure; and, much less, that of organized beings living on their surface."

 

Lord Kelvin 1895: "Heavier-than-air flying machines are impossible."

 

Albert Einstein 1934: "There is not the slightest indication that nuclear energy will ever be obtainable. It would mean that the atom would have to be shattered at will."

 

Black holes were considered impossible when light was believed (correctly) to be massles, so it could not be affected by gravity (incorrectly).

 

Yes clever men make assumptions that later prove to be incorrect - that goes for you as well as it does me. That is not what I was talking about when I made the sucking-eggs jape.

 

Hence my added footnote:

 

Dean wrote: [for future reference the idiom "teaching granny to suck eggs" would mean that Einstein would be attempting to teach Maxwell something he already knew - that is, that light was a electromagentic wave for example. It precisely does not mean he would have been teaching him something new - that is "teaching an old dog a new trick" - a different idiom with a compeltely different meaning]

 

 

 

 

It was a small array of data compared to the amount of tablular data that Kepler was able to produce after he formulated his three laws of planetary motion, Kepler's data was also considerably more accurate because it did not contain observational error. As I said in my previous post - the data was still needed in tabular form because it was not practical to compute it afresh everytime it was needed. Also, redundant data is not compression, it is redundancy - a different thing altogether. Again - if you don't know where the planet is and what its orbit is relative to the Sun then Kepler's laws will not tell you where it will be or where it was (six pieces of data are required: three for position and three for orbital velocity) - that is not data compression since the same amount of data is needed to be put into the system for what you get out. If you have three packets of data x, y and z and apply a transformation the you get x', y' and z' - three packets of data in = three packets of data out and zero compression ensues (you can now discard x, y, and z because you can recompute them by applying a reverse transformation) - the transform is not the data - you cannot put only x and y in and hope to get  x', y' and z' out. When you then apply the transform for the second time you get x'', y''and z'' ... (you can now discard x', y' and z'). However if you wanted to get from  x, y, and z to  x'', y''and z'' then you'd either have to apply the transform twice or use a different transform. BUT each time you still need to know the previous state - the transform alone tells you nothing - Kepler's three laws do not tell you the orbital speed of Venus at perihelion or aphelion just by looking at the forumlae.

 

However, to this specific point:

 

Whether we  use my simple guesstimate of the number of atoms in the human body or the Leicester University calculation of the total data content the figure we arrive at is an enormous number; whether we use my simple guesstimate of the data-transmission time based upon known limitations of bandwidth and data-rates, or the calculated value from Leicester University, the time taken to transmit the data is also an enormous number. Eitherway the compression ratio required to a) get the total data content down to a managable number and b) then get the transmission time of that data down to a practical value requires an extraordinary level of compression. The problem with any data compression is that there is a finite limit beyond which information is lost with no chance of recovery - there is a finite limit of irreducibility in any set of data. Unfortunately your proposals do not constitute data compression, a formula is a transformation.

 

For example. We know we can analyse a section of waveform using fourier analysis which gives us a table of frequecies and amplitudes that can be used to reconstitute the waveform by applying a reverse transform. In principle this is an impressive level of data compression because a relatively complex time domain waveform is apparently described by a small array of frequecy domain data. But this has not compressed the data at all, it has convolved it from the time domain to the frequency domain and back again and in doing so some information of the original waveform has been lost (it's called windowing and results in sine x over x distortion) also the resulting reconstituted wave repeats continously whereas the analysed section could have changed in the next section.

 

We cannot apply a single fourier analysis to an entire piece of music and then reconstitute the same piece of music - from the analysis we would know that specific frequencies were present in the music and we would have a figure for their average amplitudes over the entire peice, but we would not know whether a specifc tone was present for the entire duration of the piece or just for a single note and we would not know the variations in amplitude of those tones through the piece. What we would get would be a continuous complex tone that lasts for the entire duration of the piece of music.

 

To encode the entire peice we would need to analyse the whole waveform continuously at twice the highest frequency we were interested in and the duration of the section we would have to analyse is one cycle of the lowest frequency we were interested in and then repeat the analysis forthe next window shifted by time interval of one sample - the resulting frequency domain data-pack is considerably larger than an equivalent time domain data-pack.

 

While this is a grossely over-simplified example of the problem it illustrates the limitations in reducing a complex non-repetitive array of data down to a simpler array of data and a formula (or set of formulae). Repetitive data is a doddle but as soon as you get non-repetitve impulse or stochastic data within that data set then you need to record each one individially - and that limits the reducibility of the data.

 

 

 

Now tell me it will someday be possible to compress all the books in the British Library to single 8-bit ASCII Character.

 

Yes, a set of simple Laws of Physics has created a complex Universe, but  the amount of matter has not increased - what has changed is how the matter is arranged, it has become more complex and the data that fundamentally describes it is unchanged. Each "piece" of matter can be described by four basic physical dimensions - length (distance), mass, time and charge - with those we can determine (ie derive) all the physical properties of the "piece" of matter such as its location, velocity, momentum, energy, temperature, etc... and that is a fixed amount of data relating to each "piece" of matter, even when the matter is at rest it still has the properties of velocity and momentum, the values of those properties just happens to be zero (and we need to know that they are zero to know that it is at rest). So when we move it (ie apply a transform) we create some new values for that data - it now has velocity and momentum - but we have not increased the amount of data associated with it, it's just that some of those parameters now have values. To move it we had to put something "in", we had to act upon it with a force that comes from another piece of matter that also has a fixed amount of data associated with it. The transform we are applying is transfering some of the data values from mass B to mass A, in this case the momentum of mass A is obtained from mass B - the amount of data associated with mass A and mass B is unchanged so the number of bits of data needed to describe each mass is unchanged. The laws of physics are not transforming the matter, they are transforming the properties of the matter, and those properties are data values - they do not create new data, (the amount of data is unchanged, its values change), hey transform old data values into new data values.

 

The information that is associated with matter only came into existence when matter came into existence, without the matter there is no information. If you wanted to be really pedantic about it you could even argue that more data was annialated in the quark-epoch than survived it. I could care less whether the Universe is deterministic or not, metaphysical philosophical bullcrap does not interest me, whether it is or isn't changes nothing - we experience one and only one ever-changing universe - whether it could have produced something different is irrelevant, it produced this one and this is the only one we know. As I have shown several times before, unpredictability does not require random input:

 

 

 

However, all this is a futile digression that avoids the issue.

 

It would be easier if you would simply stick to the point of explaining how the complex arrangement of atoms that form a human bean can be described by a set of forumlae and a set of reduced data such that they can be used to reconstitute a replica of the human that is exact in every way (including all memories and personality), because at present you have not provided any evidence to suggest that it is ever going to be possible, in fact your example of the evolution of the Universe and its laws suggests that it will not produce an exact replica. [It is unlikely that the same rules and initial conditions would even result in a human, let alone one specific human out of 108 billion possible humans (the estimated total number of humans who have ever lived)].