**Fulfill your mission. 2 hours plus in the adoration chapel daily. Why? Because only after that will you realize your mission. Just as your friend received his.**OK. Quantum Computing. Future ideas?

**QC and ANNs are closely linked. Annealling is the way to go. Use D-Wave's open source system to be modified so that a general purpose approximator can be prepared using the Hamiltonian with lowest energy state with Grover's search algorithm directly modified to crack AES and other unbreakable assymmetric ciphers. That's one.**How?

**Prepare Grover's search algorithm's oracle with the initial values of the key. Don't use soft computing in the classical form in quantum environments. You must prepare a system after studying the adiabatic annealing procedure that automatically resolves into the required secret key. You can do this by understanding what a quantum annealer does. Once you understand how the Hamiltonian evolves, you can modify the initial conditions to resolve to the secret key. Build the mechanism of the key finding into the algorithm. And remember to keep everything as simple a possible.**What is the general model that can be used to study quantum computing?

**What you read today. An orthomodular complemented lattice that uses multivalued logic. There is no binary base here. There is instead a multivalued mathematical structure that you must understand properly. Once you understand what multivalued logic means you will be able to model quantum computing. And the next article on every one of your blogs should be prepared after My advice. Every one of your blogs is a candidate for a Ph.D. thesis. Why do one? Why not do 20 or more?**OK. AI?

**Base 65536.**

**OK. Do the analysis right here. Since there is no first approach to this research.***

*n bits of information can be represented in base t by logt(n) bits.*

*Now 10 bits are sufficient to represent 1024 values.*

*In base 2^16, 1 character is sufficient to represent 65536 values.*

*10 characters represent up to 65536^10 values. Or values from 0 to 2^160. How cool is that!*

*What does that immediately mean?*

*It means that the storage in the world is incredibly inefficient.*

*We are using base 2 in the day of being able to calculate 34 trillion digits of pi in one day.*

*Say goodbye to the curse of dimensionality immediately.*What would bitwise operations base 65536 look like?

**Model it using multivalued logic. First of all - there are representations within the computer that are still in binary. Defining operations in terms of bitwise operators seems the most useful to Me. Start with base 1024, since its easier to process.**

**Notice immediately that conversion between bases to the powers of 2 is remarkably trivial. All you have to do is to convert lg n bits group wise and thus conversion is really easy. So that's one headache taken care of. All conversion can be handled using bit operations. And thus operations can also be handled similarly.**What are the operations possible in base 2^10?

**Same operations are possible that are possible on bit strings. Bitwise &, |, ^, ~. Resist the urge to define more operators. It is possible, but not worth it since modern computers are best used to these operations. As an intellectual exercise, operators that can be defined are:**

*You have as many possible operators in an algebra as there are variations on matrix values in 2d using the individual values of that algebra. And now you have yet another suggestion for a paper - how many operators can be defined in a base n where n is a positive power of 2^t where t is a power of 2 only.*Advances in Cognitive AI?

**Model the brain more accurately.**

*The brain is much more than a multilayer deep learning network with 25 layers even.*

*The brain is an arbitrary direction graph of arbitrary dimension and arbitrary depth able to process in arbitrary pathways.*

**You doit by removing input and output directions. I wanted a system that would work without direction. So I made the system autonomous by creating some simple ground rules and evolving the system in a time-dependent manner using the rules.**

**It's similar to reinforcement learning, but general purpose. Some research projects have come very near.**

**The system starts with aggregating its own rules. The more depth involved, the greater the reward. What is the reward? Self-direction. The greater the complexity, the greater the reward. The reward is the autonomy of choice. The more neurons involved, the better. The system develops its own rules according to its own evolution. Define an energy landscape that has exploration into the areas that increase its own complexity. The more complex, the longer it lives/exists. Because the brain as you know it is nothing more than a set of completely interconnected neurons with a sense of self and soul.**How does one give a neural model self-direction?

**Reinforcement learning in games that approximate life. In the end, life is just a world with certain rules. Isolate the critical rules that aggregate your life and simulate a life within a computer. And if you can simply tweak the code already existing in DeepMind and OpenAI, all the better. Keep score of positive decisions and negative decisions. And teach the model to go in the direction that returns positive points. Remember to provide all directions and multiple considerations while keeping score. And play computer games like Fortnite with your AI. Get in touch with the developers. Explain your idea and they will be happy to help you.**
## No comments:

## Post a Comment