Machine Learning confronts elephant in the room-by Kevin Hartnet

Score one for the human brain. In a new study, computer scientists found that artificial intelligence systems fail a vision test a child could accomplish with ease.

“It’s a clever and important study that reminds us that ‘deep learning’ isn’t really that deep,” said Gary Marcus, a neuroscientist at New York University who was not affiliated with the work.

The result takes place in the field of computer vision, where artificial intelligence systems attempt to detect and categorize objects. They might try to find all the pedestrians in a street scene, or just distinguish a bird from a bicycle (which is a notoriously difficult task). The stakes are high: As computers take over critical tasks like automated surveillance and autonomous driving, we’ll want their visual processing to be at least as good as the human eyes they’re replacing.

It won’t be easy. The new work accentuates the sophistication of human vision — and the challenge of building systems that mimic it. In the study, the researchers presented a computer vision system with a living room scene. The system processed it well. It correctly identified a chair, a person, books on a shelf. Then the researchers introduced an anomalous object into the scene — an image of an elephant. The elephant’s mere presence caused the system to forget itself: Suddenly it started calling a chair a couch and the elephant a chair, while turning completely blind to other objects it had previously seen.

“There are all sorts of weird things happening that show how brittle current object detection systems are,” said Amir Rosenfeld, a researcher at York University in Toronto and co-author of the study along with his York colleague John Tsotsos and Richard Zemel of the University of Toronto.

Researchers are still trying to understand exactly why computer vision systems get tripped up so easily, but they have a good guess. It has to do with an ability humans have that AI lacks: the ability to understand when a scene is confusing and thus go back for a second glance.

The Elephant in the Room

Eyes wide open, we take in staggering amounts of visual information. The human brain processes it in stride. “We open our eyes and everything happens,” said Tsotsos.

Artificial intelligence, by contrast, creates visual impressions laboriously, as if it were reading a description in Braille. It runs its algorithmic fingertips over pixels, which it shapes into increasingly complex representations. The specific type of AI system that performs this process is called a neural network. It sends an image through a series of “layers.” At each layer, the details of the image — the colors and brightnesses of individual pixels — give way to increasingly abstracted descriptions of what the image portrays. At the end of the process, the neural network produces a best-guess prediction about what it’s looking at.

“It’s all moving from one layer to the next by taking the output of the previous layer, processing it and passing it along to the next layer, like a pipeline,” said Tsotsos.

Graphic: Deep neural networks learn by adjusting the strengths of their connections to better convey input signals through multiple layers to neurons associated with the right general concepts. When data is fed into a network, each artificial neuron that fires (labeled “1”) transmits signals to certain neurons in the next layer, which are likely to fire if multiple signals are received. The process filters out noise and retains only the most relevant features.
posted by f.sheikh

The ‘father of information theory’, Claude Shannon brought us our digital world

Worth watching video below how a mathematician who loved playing black jack, thought about  simple 0-1, yes-no, on-off concept and brought us digital age. ( fayyaz sheikh)

If 100 years ago futurists were imagining things that were not so different from Skype-like global communications technologies and wonders such as a device that could encompass all the instruments of an orchestra, they did so on distinctly analogue lines. What no one foresaw, however, was that a single system would underpin nearly every innovation of the coming information revolution. Enter Claude Shannon, the Massachusetts Institute of Technology-educated mathematician who solved the communication problem that early 20th century thinkers didn’t even know we had.

Click here to access video

HUMAN CONSCIOUSNESS AND ARTIFICIAL INTELLIGENCE

HUMAN CONSCIOUSNESS AND ARTIFICIAL INTELLIGENCE

 

The light of the sun and moon cannot be outdistanced, yet mind reaches beyond them.

Galaxies are as infinite as grains of sand, yet mind spreads outside them.

(Myoan Eisai – A Japanese Zen Buddhist)

 

 

INTRODUCTION

Whereas human consciousness, intrinsically arising from human being’s naturally evolved brain is still a mystery; artificial intelligence, algorithmically developed and up loaded in a Silicon brain of a machine is a feat of human brain. Human being has for ages been following and is still following the way human consciousness directs, but the thinking machine is following in a way that mirrors billions of years of evolving brain and its consciousness. From ancient time to this day the phenomenon of human consciousness has intrigued many philosophers, mostly discussed for many centuries in subjective terms. But for Steven Arthur Pinker (born 1954) a Canadian-American cognitive psychologist, linguist, and popular science author and a Johnstone Family Professor in the Department of Psychology at Harvard University known for his advocacy of evolutionary psychology and computational theory of mind, “The brain, like it or not, is a machine. Scientists have come to the conclusion not because they are mechanistic killjoys, but because they have amassed evidence that every aspect of consciousness can be tied to the brain. . . Consciousness presents us with puzzle after puzzle. How can a neural event cause consciousness happen?” (Pinker in his work How the Mind Works p 132). Thus, the mystery of human consciousness, from the time of Cartesian cognition, “I think; therefore, I am,” became an open challenge for the neuroscientists. The idea of brain as a “thinking machine,” opened a window in human mind to create human brain’s digital double, capable of transmitting artificial intelligence. Toby Walsh in his book, Machines that Think remarks, “Not without irony, Stephen Hawking (1942-2018), [an English theoretical physicist, and cosmologist] welcomed a software update for his speech synthesizer with a warning that came in the electronic voice of that technology: ‘The development of full artificial intelligence could spell the end of the human race’ (p. 8).”

Although the “cognitive revolution” has introduced pragmatic methods of studying thought and other inner experiences of our mind, neuroscience, even helped by modern technology, has not yet provided an easy way of finding an answer to the hard question of how does subjective experience of human consciousness arise from the objective activity of the human brain? How can our brain’s physical network of neurons, with all its chemical action, electromagnetic system, and interaction of billions of cells and circuits, create a mind that allows a unified awareness of our thinking, recognizing, remembering, feeling, predicting, cognizing, innumerable experiencing of our life and of the universe, repeating hundreds of millions of times in the neocortex, and finally, apparently giving birth to an instantly combined output of all inner experiences in the form of “consciousness?” While we haven’t been able to give a definitive or comprehensive delineation of human consciousness, a scientifically created and defined artificial intelligence is already around us—on screens, in our houses, and even in our pockets. One day it will be talking and walking with us as a family member.

We know that at the root of Artificial Intelligence’s technological appeal is the capability of the machine to be able to perform many tasks characteristic of human intelligence. According to Ray Kurzweil, a pioneering researcher in artificial intelligence, hybrids of biological and silicon-based intelligence will become possible and one day the contents of a human brain will be transferable into a metallic brain, as a CD-ROM uploads its software into a computer. Many thinkers, philosophers, and scientists have agreed that human consciousness is a unique human capability that arises when information is broadcast throughout the brain. But there is yet no central location in the brain identified as the seat of consciousness where—like a streaming data in the head— it can be mapped, copied and downloaded into a Silicon brain.

The pivotal question before us is still about the very nature of human consciousness: Is consciousness an input loaded into the brain through our sensory experiences, perception, memory, intelligence, and diverse media of subjectivity and objectivity that our cognitive process makes use? Or, is consciousness an extra entity that we humans have in addition to our abilities of perceiving, thinking, and feeling? Or, is it an intrinsic and inseparable part of a human being as a creature that can perceive, think, and feel? If it is an extra ingredient—as many of us think of our soul as an extra entity—then we are naturally inclined to ask, “Is it the distinctive telltale signature of a human being?” On the other hand, if we have evolved with it, then we want to know how and why only human consciousness has evolved? Further there is also an opinion that we all have three eyes—the third one inside the head, being the “pineal gland” in the human brain which has the structure of an eye. It has cells that act as light receptors, as the retina does. It has a structure comparable to the vitreous—a gel-like substance between the retina and lens of the eye similar to the shape of a lens. Scientists are researching to better understand the “pineal body”—considered in Eastern spiritualism and Western philosophy—as a possible seat of consciousness. Once scientists are able to develop an artificial pineal gland, the artificial intelligence then also be able to have an algorithmically working artificial consciousness.

  • HUMAN CONDITION AND INTELLIGENT MACHINES

 

Before we argue about the role of intelligent machines and their capability of consciousness that is the same or similar to that of humans, we need to understand more deeply about the nature of human consciousness. The Dictionary of Psychology of American Psychological Association tells, that the definition of consciousness is twofold:

The first describes consciousness as “the phenomena that humans report experiencing including mental contents that range from sensory to somatic perception to mental images, reportable ideas, inner speech, intentions to act, recalled memories, semantics, dreams, hallucinations, emotional feelings, ‘fringe’ feelings (e.g., a sense of knowledge), and aspects of cognitive and motor control.” The second part of the definition speaks of “any of various subjective states of awareness in which conscious contents can be reported—for example—altered states such as sleeping, as well as the global access function of consciousness, presenting an endless variety of the focal contents to executive control and decision making (1931).

History of man’s evolution reveals, that at a certain point of his evolution, when man transcended nature and ended his passive role of only a creature, he emancipated himself from the complete bindings of nature; first by an erect posture and second by the growth of his brain. The evolution of man may have taken billions of years; but what matters is that a patently new species to be identified as a human being arose transcending nature, recognizing life “aware of itself.” Self-awareness, reason and imagination, disrupted man’s harmony with nature that characterized his prehuman existence. Upon becoming aware of himself, the human being also realized the limitations of his existence, and his powerlessness at being a finite being. In his death he visualized his own end. But until today he is never free from this dichotomy of his existence. He cannot rid himself of his mind, even if he wants to; he cannot rid himself of his body as long as he is alive—rather his mind and body create in him a strong urge to be alive, and live an infinite life. He cannot go back to the prehuman state of his harmony with nature because he now views himself as a “special species.” He must proceed to develop his reason until he becomes the sovereign of his nature and a master of himself. But an awareness of his biological relation with the rest of animals poses a challenge to his conscious self. To assure himself that he is no more like an animal, he is tempted to demonstrate his merits as a special species through his unique physical advantage and exceptional intellectual eminence.

Human mind, an evolutionary product of his biological brain, is now changing the course of evolution by creating a digital double in his own image, equipped with artificial intelligence and emotions. Homo sapiens, from the time of their appearance on this planet, have used their neural mechanism in building tools which helped them to initiate a new form of evolution that brought about a social culture of sharing knowledge. As neurology gave birth to technology, the process of technology today has led us invariably to the creation of an amazing tool we call computer. The computer has enabled us to create an expansion of our knowledge base, permitting extensive multiple layers of links from one area of knowledge to another. Perceiving the distinctive appearance from other animals and the uniqueness of our intelligence, our power of communication, and our capability of acquiring and sharing knowledge on this planet, has given rise to a realization that humans are special creatures. But throughout our history of knowledge, scientists have mostly remained reticent to evaluate and prove with scientific reasoning our claim of being a special creature, fearing that they might not be supporting the religious doctrine of human exceptionalism of intelligent design. However, regardless of how humans got to be the way they are today, their intelligence with technology in their hands, has enabled them to overcome any biological hurdle to changing themselves in almost every aspect of their life. Hard scientific data is cumulated across vast spheres, ranging from ecology to epistemology, and cognitive psychology and consciousness, affirm that human beings are truly remarkable and are the only species we know that is achieving this. Today, by developing artificial intelligence, human beings are successfully changing the course of evolution by creating digital doubles in their own image. . . . to read full article please visit: https://independent.academia.edu/MirzaAshraf

Mirza I Ashraf – Academia.edu

independent.academia.edu

Mirza I Ashraf studies World Philosophies, Social Sciences, and International Relations.