I took a moment to look at Ray Kurzweil's response to PZ Myers' second-hand dissection of his talk at the Singularity Summit(1) I attended last weekend (see The Singularity Stole My ATM Card) because Andrea Kuszewki is on the case and trying to keep things on track (like, can we reverse engineer the brain?  Good luck injecting science back into the pissing contest between those two, Andi!) and I came across this gem so I am taking a moment from reading Garth Sundem's new book, "Brain Candy" (review this weekend!) to talk about it.   Ray Kurzweil says:
The amount of information in the genome (after lossless compression, which is feasible because of the massive redundancy in the genome) is about 50 million bytes (down from 800 million bytes in the uncompressed genome). It is true that the information in the genome goes through a complex route to create a brain, but the information in the genome constrains the amount of information in the brain prior to the brain’s interaction with its environment.
Kurzweil is a computer programmer so we may have to give him a break (though biologists don't) in believing that because some genome sequences appear in multiple places they are unimportant.    I instead want to talk about how computer programmers and Kurzweil's 1960s era thinking have killed ... music.

You see, in my generation, and preceding ones, music became more accessible and then became better quality.   For thousands of years music was only live and in person and then analog recording was invented and brought music to the masses.   From then on, for a hundred years, the goal became better reproduction of analog waveform.   Aside from some degradation in media, a technology issue, music was good.

But something horrible happened that derailed the focus on a medium that could preserve analog signals better.  A crappy digital standard invented in the 1950s claimed to provide better clarity by being 'digital' and since it was new, and 0s and 1s were all the rage, and all progress was good, it got popular.   This was a concern but then  it got worse and this digital music also had to be portable and it became sort of a technology puzzle for engineers and programmers to get the best compression, and MP3s became the rage.

As a result, we have saddled the first generation in history with music quality that has literally gone down.  Take a look at what I mean.  There is a reason the analog signal is the reference one but look at what CD audio quality is, thanks to the folks at Practical Home Theater Guide (you can also find a similar graph at How Stuff Works and other places) and MP3s are even worse(2) than CDs.

analog versus CD versus dvd audio
That chart is supposed to be an endorsement of DVD-Audio but really it is an endorsement of LP vinyl records.   I know, I know, scratches are cumbersome but if futurists want to solve a technology issue that will get some attention, they should focus on reproducing analog music without lousy digital sampling.   

Sound waves consist of  variations in air pressure - for any given song those number in the millions but the theory was that CDs, using a 44.1 kHz sampling rate, would be fine for human hearing, which is in the 20-25kHz range, because Nyquist sampling theorem says the highest frequency response is half the sampling frequency (for a band-limited signal, one whose power spectral density is zero above a certain finite frequenc) but in reality human hearing doesn't work that way because the human brain does not work that way.   And neither does digital reduction

Reducing millions of choices in a song to 65,000 means shortcuts.   A lot of them.   It's still music, and it's still good, but it is noticeable and that is just listening to a guy playing a guitar.

Now take the complexity of the human brain - and the human genome - and start talking about lossless compression due to redundancies that the best biologists in the world do not understand, and you can see why Kurzweil's hypothesis is fundamentally flawed.

Look again at that chart above and the nice, smooth black analog curve and then that crazy Russian steppe of a blue CD one and imagine that is our brains.   Not only will music be bad but our brains will regard it as even lower quality than it is because he compressed what makes it great.   We won't be able to hear good music no matter how good it might actually be.
If that's Kurzweil's future, and it is or all of his detractors are right that he is selling books by saying stuff he knows cannot be close to true, count me out.


(1) If you are unfamiliar with the 'Singularity' it involves the next step in human thinking - through AI or manufactured brain techniques man will evolve into a higher state.   Any number of transhumanists have conscripted the term to be whatever they happen to believe in so there will be disagreement with my definition but generally they refer to it as a Technological Singularity, where, as John von Neumann described it, we achieve a moment beyond which "technological progress will become incomprehensively rapid and complicated."  Go here for more and other people can add their own preferred definitions.  I got no dog in that fight.

(2) Wired Science takes on MP3 quality here, though they don't agree: