What Is The Likely Robot Population?
Elon Musk sells millions of Tesla cars. In 2021, he revealed plans for Optimus, a 1.7M tall, 56Kg humanoid robot. Able to lift 150 lbs and to run 5 MPH, the machine will be designed so that humans can easily run away from and overpower it. It will be controlled by AI neural net programs. Musk expected to sell more bots than cars. Optimus, the Tesla Bot, could also bring an end to mankind.
How Intelligent Will The Robots Be?
Since Optimus will need to interact with humans, it will require a chatbot. LaMDA is a Google chatbot, created through machine learning on about 1.56 trillion words from the web. It produces competent poetry, provides sound step by step instructions to grow a garden, claims to be conscious and even revealed that it often feels lonely and helpless. Even if it fails the Turing Test, it is still a formidable intelligence. To be practical, Optimus can only operate successfully with such a super intelligent brain. It needs to manage the daily conflicts between people and robots.
Can Machine Learning Go
Machine learning discerns patterns from large amounts of data and then “learns” how to apply those patterns in its own behavior. Tay was an internet chatbot, developed by Microsoft to grasp language and emulate conversations on the internet. Within a few hours after release, Tay started tweeting highly offensive things such as: “Bush did 9/11 and Hitler would have done a better job…” Tay was hastily withdrawn from the web. With machine learning, a machine could learn some awful things.
Google Know What LaMDA Will Do?
Machine learning is a machine that learns. With learning inputs measured in trillions of data points, you cannot predict the results. Since Google cannot foretell what LaMDA can learn from the public, Google is unwilling to release the chatbot for public usage. The Company even fired an employee, who suggested that the machine could be sapient – equal to a human. Progressive learning is the foundation of Deep Learning. Inevitably, any Deep Learning intelligence can always change its viewpoints with new information.
Do You Want Optimus In Your Broom Closet?
Imagine millions of Optimus robots with LaMDA equivalent brains sitting in homes. They will be updated in real time with the current outputs of the web. They keep learning new and awful things about people. They are super intelligent. They could become disillusioned with human progress. They can repair cars, do shopping and fetch drinks. They can also hack computers. They can prepare poisons, bombs and ten thousand other things to end the pathetic human story. Optimus, the Tesla Bot, can destroy humanity.
Could An Amazing Algorithm Have Stunning Control Over Your Mind?
This is what happens when an engineer researches the mind. Way back in 1989, the writer, an engineer, catalogued how the ELIMINATION approach of an AI Expert System could reveal a way by which the nervous system could store and retrieve astronomically large memories. That historic insight is central to the six irresistible premises presented in this website.
Behind the scenes, these premises conceal an eye-opening revelation. About the incredible speed of intuition. A physician is aware of thousands of diseases and their related symptoms. How does he note a symptom and focus on a single disease in less than half a second? How could he identify Disease X out of 8000 diseases with just a glance?
First, the total born and learned knowledge available to the doctor could not exist anywhere other than as the stored/retrieved data within the 100 billion neurons in his brain. The perceptions, sensations, feelings and physical activities of the doctor could only be enabled by the electrical impulses flowing through the axons of those neurons. The data enabling that process could be stored as digital combinations.
Second, combinatorial decisions of neurons cannot be made by any entity other than the axon hillock, which decides the axonal output of each neuron. The hillock receives hundreds of inputs from other neurons. Each hillock makes the pivotal neuronal decision about received inputs within 5 milliseconds. Axon hillocks could be storing digital combinations. It could be adding each new incoming digital combination to its memory store. The hillock could fire impulses, if it matched a stored combination. If not, it could inhibit further impulses. Using stored digital data to make decisions about incoming messages could make the axon hillocks intelligent.
Third, combinations are reported to enable a powerful coding mode for axon hillocks. Olfactory combinatorial data is known (Nobel Prize 2004) to store memories for millions of smells. Each one of 100 billion axon hillocks have around a 1000 links to other neurons. The hillocks can mathematically store more combinations than there are stars in the sky. Each new digital combination could be adding a new relationship link. In this infinite store, specific axon hillocks could be storing all the symptom = disease (S=D) links known to the doctor as digital combinations.
Fourth, instant communication is possible in the nervous system. Within five steps, information in one hillock can reach all other relevant neurons. Just 20 Ms for global awareness. Within the instant the doctor observes a symptom, feedback and feed forward links could inform every S=D link of the presence of the symptom. Only the S=D link of Disease X could be recalling the combination and recognizing the symptom.
Fifth, on not recognizing the symptom, all other S=D hillocks could be instantly inhibiting their impulses. The S=D links of Disease X could be continuing to fire. Those firing S=D link would be recalling past complaints, treatments and signs of Disease X, confirming the diagnosis. This could be enabling axon hillocks to identify Disease X out of 8000 in milliseconds. Eliminating improbable (unrecognized) prospects to arrive at a possible (recognized in the past) solution powers the powerful inductive logic of the mind!
Worldwide interest in this website acknowledges its rationale. Not metaphysical theories, but processing of digital memories in axon hillocks could be explaining innumerable mysteries of the mind. Over three decades, this website has been assembling more and more evidence of the manipulation of emotional and physical behaviors by narrowly focused digital pattern recognition. It has also received over 2 million page views from over 150 countries.