The Affective Neuron™ quantum machine learning module may revolutionize artificial intelligence. Affective Neuron™ learns from transducing human intuition and emotion by importing engrams implementing human knowledge. Conditioned, not trained, this quantum machine learning module enables unbiased learning in real-time, continuously.
Affective computing is the study and development of systems and devices that can recognize, interpret, process and simulate human affects (emotions).
Affective Neuron™ (AN) is a novel real-time quantum unsupervised reinforcement learning synthetic neuron that can be to reproduce human affects including but not limited to moods, feelings, attitudes and memories. The AN can be used singularly or collectively to achieve complex human emotions as described by Plutchik’s wheel of emotions infograph.
Affective Neuron™ implements complex human emotions and memories via engram transduction. Engram transduction is the process of taking human belief systems and memories and representing them as probability distributions (the engrams). The distributions are then transduced into the AN by mapping the distribution to the AN track the AN locator moves on. The engram transduction process includes forces that induce resistance and/or acceleration to the AN locator movement. The result is the AN is initialized with the human engram distribution, which itself can evolve over time from the initial engram transduction process.
Human engrams are captured by questionnaires, clinical observations and through direct brain transport using devices such as MRI and EEG attached to a human subject presented with specific visual and auditory stimuli to illicit responses captured by experimental apparatus. Once the engram traces are captured, they are AN mapped through transduction.
The Affective Neuron™ is a new patent pending unsupervised reinforcement learning technology that mimics human emotions, including intuition by which it learns and the feeling of epiphany. The AN is a new fine grained technology that learns learns continuously in real-time. It may revolutionize reinforcement learning.
The following is a validation experiment showing the AN demonstrating intuition and the feeling of epiphany and showing post traumatic stress disorder behavior (PTSD). Contrary to PTSD in humans, this AN functionality enables a new paradigm of machine learning systems where AN hyper vigilance is used to detect anomalies.
First as a control we show a single AN trained to recognize a consistent signal (Figure 1). Note that the AN becomes aware that the signal is always the same (see the surprise Physiological Response) and then settles down. AN trials are not the same as in all current ANN, since AN learns continuously and AN learning is instantaneous with micro-second response. This significance of the AN control must be understood. The single AN response here is no different here than a living system being exposed to a consistent stimulus repeatedly and suddenly realizing that the stimulus is consistent, i.e. an intuitive Epiphany.
An epiphany (feeling) is an experience of sudden and striking realization. Generally the term is used to describe scientific breakthrough, religious or philosophical discoveries, but it can apply in any situation in which an enlightening realization allows a problem or situation to be understood from a new and deeper perspective. Epiphanies are studied by psychologists and other scholars, particularly those attempting to study the process of innovation.
Next we expose a fresh AN participating in the PTSD experiment and continuously expose it to an erratic signal, which induces a PTSD response (Figure 2). Note that the AN quickly becomes aware of the erratic signal (see the Physiological Response). Unlike in the control AN, the Chronic Anxiety of this AN begins to rise. Both elevated metrics indicating the PTSD acute stress response. The AN PTSD training is suspended where the Physiological Response blue points start to erratically spike near the end of the AN PTSD exposure session.
Finally we expose the AN that has been damaged by PTSD and expose it to the same consistent signal as the experimental control AN (see Figure 3). This induces the AN PTSD response precisely as seen in living systems.
Note the PTSD AN Physiological Response quickly over reacts to the stable control stimulus. The AN Chronic Anxiety is also elevated. The PTSD AN eventually understands that the signal is consistent shown by the Physiological Response and its Chronic Anxiety leveling. The PTSD AN maintains a state of hyper vigilance, to the possibility of the original PTSD signal exposure returning.
The AN is applied in novel ways with capabilities not possible with current machine learning technologies. The AN enables a whole different paradigm of implementing machine learning and recognition. We are applying the AN to detect cryptocurrency mining targets of opportunity.