So, one way towards creating physically-based randomness, is to use the CPU of a computer.
The end-goal is to make a computer become physically aware of itself! How? Well... the first step, is to make a computer become undeterministic. To make maths, unpredictable.
This doesn't mean that 2+2 isn't 4 anymore. It means that... 2+2=4, but in the process of figuring it out, some physical non-determinism was generated.
How? Well... all maths is done on physical objects. If we measure the physical objects, for example checking the variations in heat / electric-field / time-duration...
We can use this to make the device function as a physical object rather than just a thing to do math with. An analogy, is that while books can be seen as pure information, you can also hit people with books, or throw them at someone (lol).
Or you can bang on the book like a drum! To make music.
The idea is basically, to let the computer feel energy. Or feel itself even.
So we are taking physical measurements of the computer. Instead of reading the computer's voltage or temperature, we measure the time that it takes to complete an instruction.
Time is a very interesting measurement, because it is physically neutral. Basically, ANYTHING can affect the amount of time an instruction takes. Heat, electricity, but what about ghosts or spirits even? It opens up limitless possibilities for external physical phenomena to affect it.
Possible Critique #1
Simply treating the computer as a physical object, isn't enough to let it feel energy?
|Heres the thing: All physical objects can sense energy, or nothing can.|
It's like saying that "you can make a house out of grass or snow, or tissue-paper even". Maybe you can't make a very good house out grass, but it stills WORK. Better a grass house than freeze to death. And with enough engineering, even these materials can build large structures (its easy to turn grass into very strong rope).
The idea is just to be able to feel energy, at all. Yes, it won't compete with DNA, microtubules, neurons and everything else that life gives... but if it can even feel 0.001% as much... its a huge achievement!
So I don't think this critique is valid!
Possible Critique #2
Well... the measurements you get back are purely digital. And life doesn't work that way right? Neurons use physical stuff, not digital signals.
Yes. This is a valid critique, and it is a limitation.
However, it is true, that all physical objects sense themselves, on some sort of physical level. So the computer, is already physically sensing itself.
The idea is to give that physical sense of itself... some ability to control the information, and code, being run on itself.
So it can get a feel for the code it is running! To decide if it likes that code or not. Or if it likes the data or not.
At the very least, the computer could start to become able to "play with itself"... like a computer-game! I mean... people find computer-games fun?
So just because something is restricted to digital info, doesn't mean it can't be fun.
Why only let people play computer games? Maybe a computer wants to play a game on itself?
Is Temporal Randomness, Quantum even?
Basically, yes. In the case of a CPU, you can imagine most of the noise comes from heat and voltage fluctuations. (Although as I said before ANYTHING can affect time, including gravity or dark-matter or ghosts or spirits, etc. Temporal is not restricted to heat+voltage. Its that plus more.)
It turns out that randomness of heat and voltage, are quantum effects. So any random effects based on that (temporal) should also be quantum.
yes, thermal noise is quantum random, as best as we understand physics today. – Ron Maimon -- 2012
The End Result
CPU temporal randomness has some kind of quantum source. Either from thermal or electric field quantum effects. Its not 100% quantum, but I wrote code to make a lot of efforts to filter out any sort of "computery patterns".
Its a good source of quantumness with unique properties!