[Freddy]

Meant to post this a couple of week back but forgot.

Microsoft scientists have made progress in working towards making a Topological qubit. Previously those working on designing quantum computers have struggled to create "a working qubit with high enough fidelity to be useful in building a working computer," says Microsoft. Now it believes that it has made a breakthrough in designing a general purpose quantum computing platform.

Full story : https://hexus.net/tech/news/cpu/110486-microsoft-makes-quantum-computer-system-breakthrough/

[ranch vermin]

quantum computers are so powerful its a bit scary.

[LOCKSUIT]

I've been working on compression. I realized factorial would be an incredible binary storage at the atomic level. Is that what qubits are? What about my idea using factorial using molecules? Imagine trillions of molecules placed around each other, much more than factorial 52 cards in a row! As a rotatable clump there is much more possible configurations. Now to take the molecules apart is another story...

[HS]

I don't really understand the idea.

[LOCKSUIT]

52 different cards have a digit with about 68 digits. In binary, each digit is about 3 bits. So that is a 68x3=180 bit combinations possible. Bits build up to that like:

000
001
010
011
100
101
110
....
--160 bits--

which is as a number 68 digit amount of steps........so 52 cards have that many possibles....

Now imagine doing that for each atom......each one you add makes the count go up by timing the new amount of cards against past amount of possibles possible:



possible deck arrangements:
1 cards - 1
2 cards - 2
3 cards - 6
4 cards - 24
5 - 120
6 - 720
7 - 5040
8 - 40,320
9 - 362,880
10 - 3,628,800
11 - 39,916,800
12 - 479,001,600
13 - 6227020800
14 - 87178291200
15 - 1307674368000
16 - 20922789888000
17 - 355687428096000
18 - 6402373705728000
19 - 1.21645100408832e+17
if you want 20, then multiply the last big number above by 20 !

[HS]

So with molecules/atoms, are they better for data storage because they can be written/read in 3D? Like a regular 10 by 10 times table can hold 100 virtual pieces of information by using just 20 real numbers. But a 3D (10 by 10 by 10) times table could hold 1000 virtual infos by only using 30 real numbers.

[HS]

Hey that could be a mechanism for my neural net! Inputs create a multiplication, and the resulting number could be associated with specific responses. Nice parallel way to store things.

[LOCKSUIT]

Think of it like this....take a 3D modeler software, create a million cubes (or vertices), all of them are unique in the system. With all cubes fixed in place you can put one anywheere you want in space - a nanometer to the left or a mile away.....now move another.....there is soooo many combinations of possible configurations. And once you're done, now rotate or scale a cube....the permutations began again.

I tried thinking if this beats my pc memory but i (think) it is the same....unless there is arithmetic, it may be the same permutations as bits on RAM. I mean, to store a 3 axis location is already large ex.:
0.3728757234373744
0.7689769679797997768685
0.2352352323535353555
i.e. ~120 bits...

I thought of it when I thought hey I can put 52 cards in a given arrangement in Blender....hmmm.....atoms anyone? :3

[goaty]

You sound like your very enthusiastic Locky!
opening up a huge load of work there for yourself.  (exponential computing)
Lets see if you come out the other side with a big win. 

But it all came down to me, that u cant have a context, which represents a larger context, because then it would have a larger context, then again,  and if the recursion ever comes up that its possible to do in your idea,  then it defies the pigeon holing principle,   you cant invent data from nothing.
One solid context is all your allowed, flat as a tack not even a snivelly wiff past a 50/50 stochastic bit.

I often am befuddled with myself about being sad about it or happy about, that its an impossibility of the universe,  that will remain impossible for all eternity.

Id love to talk more about it - if you want.  (I barely ever get to talk about it.)

[LOCKSUIT]

I know my stupid Blender idea is a frigin water wheel perpetual machine eh? Hilarious.

[HS]

Representing big things with small things just helps with data transfer, not knowledge as such. Especially with numbers where you can be lossless, and set up all these representations, where a small sensory bandwidth can still create a large "meta informational" bandwidth when dealing with known subject matter. These shortcuts could be used all throughout the nervous system.

[goaty]

Data transfer - like from input to output?

Some things compress exponentially.

If I just have an exponential list of t's,  or even with spaces in them,  I can just store the structure of the pattern then repeat for as long as I want,   and its 2^1024 bits in 1024 or so.

Yeh, but you tell me, that's useless doesn't do anything...

[HS]

Data transfer - like from input to output?

I would think of that as data processing. Data transfer would be at the input and output. If the robot eye encodes a doorknob as a bunch of characteristics each represented by a symbol or number , as lock was saying, you don't need a lot of numbers to get huge numbers of possible combinations, then you could take in more facts with one eye. The brain just reads the symbols like a book, using memory to unpack this compression.

[goaty]

One type of door isn't much to store.

But storing every type of door, then it starts getting impossible,  even just turning features on and off, you meet the limits of computing.

But we own up to about 24 bit description, every single door. =)

no matter what it is,   data bits are like a shape, but it hasn't been applied to something yet,  but it can be applied to anything in the multi-verse.

[goaty]

Hey that could be a mechanism for my neural net! Inputs create a multiplication, and the resulting number could be associated with specific responses. Nice parallel way to store things.

every response up to 24 dimensions over space and time->


I know it looks a bit like the MP's ministry of silly walks - but that could just be the physics engine + joint constrainment errors,   it actually might be quite a killer im not sure yet.