Why could we not have a hat or a wolly that has protecting features. Like a helmet but still looking like a wolly. So the idea is that we innovate a reinforced winter hat. That can protect the head if you fall. At least it should give some protection.
My idea is simple.
Information or a signal can be described as many sin waves of different frequencies. Like a Fourier series with low frequencies to the high.
I guess then an answer from a theory is a signal. So it will have different frequencies.
So my idea is that a theory of everything should handle and output everything from the very low frequencies to the very high.
So I had the idea that maybe biomimicry exist in mathematics also. Maybe it does but I thought what would I do to enhance optimation.
If you have a function F(x,y) and you want to find x and y that minimizes this function. You probably don’t get the best values at the beginning. Call them false positives.
What I wonder is if you can simply add a fill-function that spans over the domain and prevents the optimation from looking or going towards those false positive positions.
F(x,y) -> F(x,y) + fill1(x,y) + fill2(x,y) + fill3(x,y) + …
I don’t know what the different fill functions should look like though. The idea is that this simplifies F(x,y) in search for the best values for x and y.
I speculate that maybe it would be possible to build a series with similarity as the base function. So you have.
[x0, x1, x2, x3, … ]
where x1 is most similar to x0 and x2 is most similar to x1 and x3 is most similar to x2
but all x are unique.
Even if this is somewhat inaccurate. There should exist something like an evolutionary like math series.
So if you start with triangle and a square I imagine you could end up with a circle.
Another idea is the guessing series
1 2 3 … Filling in the rest is pretty easy.
365 52 … Filling the rest is easy but it shows a guessing series need preference data.
My idea is simple.
So I was wondering. What is new information?
Inspired by this I realized that. If you recognize noise in the mix of frequencies then part of it doesn’t form information. Or at least this the guessing sentence with delivery value.
You can transform information so it sounds like noise but then it is not directly understandable information anymore.
So my guess is that its possible in theory to generate new kind of information by just mixing sin wave signals and check for noise.
From this little thought experiment my guess is that you can generate new type of error functions. Here the error function is a noise detection function.