Our extended consciousness

A problem-solving paradigm is rapidly transitioning from the classical determinism based paradigm to the non-determinism based paradigm. So the non-determinism based problem-solving methodology opens Pandora's box which enables us to solve a problem impossible to solve by the classical determinism based one. Concretely, problems impossible to solve via analytical methods such as Calculus, etc. can be solved via non-determinism based optimization methods such as the gradient descent, contrastive divergence, simulated annealing, quantum annealing, and so on.

A human-level reasoning machine for prediction/judgment with a relationship model can be created based on the gradient descent, contrastive divergence, and so on, with experiences. Also, a human-level reasoning machine to search for solutions from prior knowledge can be created based on energy-based optimizing methods such as the simulated annealing, quantum annealing, and so on. Such human-level reasoning machines become our extended consciousness parts, and individuals including small business owners and entrepreneurs can utilize them as their extended consciousness-level automatic reasoning tools to create and expand values by themselves.

Reasoning intelligent entity based on deep relationship models

To make an optimal decision, valid prior knowledge must be obtained. Through exploiting deep learning based relationship models with your experience data which are automatic, human-level reasoning intelligent entities, valid prior knowledge for an optimal decision can be induced automatically and fast, so you can make an optimal decision more efficiently and quickly than using analytical models. Also, by exploiting a deep learning based decision model (including a reinforcement learning model) with your experience data, you can get decision recommendations or make optimal decisions automatically. We can develop such reasoning intelligent entities with your experience data, using deep learning frameworks such as Tensorflow, and PyTorch. Especially, we developed and published the Keras Unsupervised framework in Github and PyPI , so we can provide more accurate deep learning based relationship or decision models via unsupervised learning although your small experience data.

Reasoning intelligent entity based on optimizing models

Without your experience data, you can infer solutions from only prior knowledge. However, Although that problem complexity is a little complicated, you can’t quickly infer solutions or can’t solve problems at all, by yourselves. By the way, based on energy-based optimizing methods such as the simulated annealing and quantum annealing, that inference problem can be quickly solved and completely automated. We can make energy functions by analyzing your problems, and then develop and provide automatic inference models based on the simulated annealing or quantum annealing.