Will Computers Revolt? Preparing for the Future of Artificial Intelligence

Issue #1,000 

This issue sponsored by BricsCAD V19 catapults forward in BIM, mechanical, and traditional CAD. You work across all disciplines with the common DWG environment.   Discover the ABCs of BricsCAD at the forthcoming one-day Australasia BricsCAD Conference, 19 March 2019 in Brisbane, Australia: Learn how BricsCAD became the essential design application Tips and guidance to switch easily to BricsCAD Save money and boost productivity with BricsCAD Book your ticket now at http://www.trybooking.com/YIUA.  Early bird pricing valid until December 31, 2018 Sponsored by Sofoco of Australia and CAD Concepts of New Zealand Will Computers Revolt? Preparing for the Future of Artificial Intelligence Book by Charles Simon,  reviewed by Ralph Grabowski   When you replace parts in your car -- like wiper blades, an oil filter, the engine, some doors -- at what point is the car no longer the one you bought?  This is the Ship of Theseus thought experiment first conducted by Plutarch two thousand years ago. He mused about a ship whose timbers were rotting. If he were to replace each one with a new piece of wood, at what point would the ship no longer be the original?  Other philosophers expanded on the Theseus Theorem (named after the mythical king). Replacing an ax handle, and then its head. Water molecules flowing in a river. Cells dying naturally and being replaced in our bodies. The currently trendy version of Ship of Theseus theorem is where we replace a single neuron in your brain with an electromechanical one that operates identically to the organic one. You retain all your thoughts, and so you are the same person, presumably. Now replace all organic neurons and cells in your brain with electromechanical equivalents; still the same brain? Move the electromechanical brain (because we can do that) and store it remotely with a communications link back to the body. Still you? Philosophers are split. Some say the object remains the same no matter how much is changed, because the form remains the same (a car is a car; a ship is a ship). Other philosophers figure that changed objects are different from the originals, although they disagree at what point the changeover occurs. Some say it happens when a single item is changed, some when 51% is replaced, and others not until 100% is different. A difficulty enters the thought experiment when a copy is made, such as saving the original rotting timbers, reversing the rot, and rebuilding the ship with the original wood so that there now are two "original" ships. Or by making a copy of your electromechanical brain, keeping one in the skull and a copy somewhere else as a backup or a clone. Which you is you?   The Chemical of the Spirit The modern version of Theseus problem, however, fails to take into account the difference between non-living and living objects. The difference between a rock and your child is... what? We graft branches onto trees, transplant hearts into humans, but when we replace the human brain with a robotic equivalent, we don't know the result, because we've never done it. We do know that when we mess with a brain, the personality changes. Are you still you? Reductionists like Ray Kurzweil of Google and Alison Lowndes of nVidia enthusiastically affirm that brain-replacement is possible -- even necessary. They call the digital improvement to the brain's 100 billion nerves and trillions of synapse connections "the Singularity." Mr Kurzweil has scheduled it to begin in 2029 (starting with brain uploads to computers) and to be completed by 2045, precisely. The 48-year-old Ms Lowndes is less precise: "In my lifetime, the Singularity will happen." A reductionist says all entities can be reduced to their basic parts, and so it is inferred that human functions can be reduced to robotic equivalents. (See figure 1.) The philosophy leads necessarily to materialism: the only things that exist are materials that we can sense, which necessarily eliminates the spirit. Reductionists solve the spirit problem by assuming our sensation of having a human spirit is an illusion, the result of chemicals working with neurons and synapses -- somehow. Figure 1: Jacques de Vaucanson in 1738 invented a mechanical duck with 400 moving parts The reductionist-materialism path weds reductionists to the notion that love is also no more than brain chemicals reacting to stimuli; that altruism is an beneficial adaptive behavior conditioned by millions of years of evolution; and the shame that separates us from animals is an artificial construct concocted by religion to be eradicated by lots and lots of sex. (No surprise, then, at the rise in sex robots.) "Weds," because reductionism is a philosophy, not a science.  Materialism bottoms out at determinism: we have no free will, because we do only what chemicals in our body motivate us to do. In a handy parallel, robot-humans will do only what they learn from our programming and their experiences, conveniently enough.  The certainty of the reductionist-determinist world view falters, however, when he experiences the love of his children; he is unable to explain it well enough to satisfy his nerved spirit. In the most severe cases, reductionists are forced to deny their own existence, as their philosophy requires them to believe that even their existence is a trick played by brain chemicals, as in The Matrix. They have fallen prey to Descartes' malicious demon.   "Will Computers Revolt?" Charles Simon's Will Computers Revolt? describes the ramp that technology must climb to arrive at the Singularity. Early on, he gives the game away: he is a reductionist. After doing the neuron-replacement version of the Theseus Theorem, he states, "What makes you you incorporates the sum total of the experiences you've had." We are, of course, more than a collection of experiences. His thesis suffers when it disregards the intermediate steps needed to reach the desired conclusion. Here's one example: We don't know the brain's compute power, so we don't know the power computers need to simulate it Some estimate the brain's power to be 30 terraflops [trillion floating point operations per second] at the low end, and 1,000 terraflops at the upper end ??? Singularity by 2045! (Step #2 is a misdirect, because brains don't engage in floating point operations; they can't. As well, the range of 30 to 1,000 is a wild guess, not science.) In another example, this time of poor extrapolation, Mr Simon provides a chart showing the exponential growth in computing power through to the year 2100. Exponential growth is necessary to reach the Singularity by... well, by now you know the drill. The data in the chart, however, ends at 2000, perhaps because back in 2010 CPU speeds flat lined.  (He doesn't, oddly enough, bring up GPUs. They have taken over from CPUs to provide exponential increases in compute power, thanks to their brute force method of adding millions more processing units by decreasing their nanometer size, currently at 7nm. GPUs are expected to flat line once their circuit widths fall to 5nm, sometime after 2020, due to unsolvable leakage, resistance, and heat problems.)   How Will Computers Revolt? To make the decision to revolt, the computer-brain of the future must be conscious enough to experience revulsion, then be able to imagine a better outcome than exists now, and finally execute a revolt against others through free will. (See figure 2.)   Figure 2: Medical robot kills its first patient  In Chapter 12's "Free Will and Consciousness," the author sidesteps the problem. He notes that some readers may feel that free will, etc are outside the purview of science. But not to him. He asserts that these human attributes, along with the imagination, are located somewhere in the brain, and so can be simulated by computer code -- eventually. Never mind that he admits no one knows where these most human of traits are located or how they work. He relies on the hope (another uniquely human emotion) that quantum mechanics will rescue both he and his fellow reductionists.  In a podcast a few week back, Dilbert's Scott Adams reverses the question. He asked what has to be removed from a person to make him non-human: the brain. (See figure 3.) He has his own theory on what makes a robot human-like: to do whatever it takes to maximize human reproduction, because the decades-old dictum to robots, "Don't harm humans," is too vague for My Life vs. Your Life conflicts.   Figure 3: Scott Adams erasing everything unnecessary from the human body Mr Simon has no choice. To be a proponent of the Singularity, he must ignore the spirit, the soul, the tao, the pneuma, the psyche, the noumenon that is beyond the knowledge of science, outside the realm of technology, and unattainable by the Singularity. It will be delicious irony to see upcoming failures of science and technology make the longed-for Singularity impossible. Longed for, because the eschaton of the singularity is Eternal life, the dream by geeks for defeating their final foe, death. In the end, Mr Simon's thesis suffers from the Swiss cheese fallacy: his passion for the bright, inevitable, necessary future of the Singularity stumbles over the gaps that yawn between the technology of today and of tomorrow. The subject looks like a tasty whole, but his "we could imagine...", "not yet...," and "we don't know..." fail to plug the holes that define the cheese. His best argument on whether robots will have real consciousness and real emotions is as lame as, "I accept that you are a conscious being, even though I can't tell for sure." Oh, the agony of the determinists! So, will computers revolt? The answer, after reading this book, is, 'Definitely. It depends. Maybe. We'll see." The best this book can prepare you for is to see the gaps in the path of logic that is supposed to be leading us to the Singularity. Will Computers Revolt? Preparing for the Future of AI by Charles Simon 249 pages, paperback $19.95 Published by Future AI, 2018 ISBN-13: 978-1732687219 http://willcomputersrevolt.com/default.aspx Or as Nihilist Autodesk University (@AuNihilist) put it, "If you're feeling overwhelmed by the complexity of our BIM 360 product offerings, just remember that choice is an illusion constructed by random electrical impulses coursing through your primate brain."   ARES KUDO: MODERN DWG EDITING IN THE CLOUD   ARES Kudo is the only online CAD solution offering a full set of 2D features to read, create, modify, and share DWG and DXF drawings. Nothing to install. It runs in your Web browser and your files follow you on any device. ARES Kudo is integrated with cloud storage services like Google Drive, Dropbox, Box, OneDrive, OneDrive for Business -- and industry-specific solutions such as Onshape and Trimble Connect.  Visit www.graebert.com/areskudo to learn more and activate your own free trial at kudo.graebert.com Notable Quotable "In my lifetime, the singularity will happen. But why does everyone think they’d be hostile? That’s our brain assuming it’s evil. Why on earth would it need to be? People are just terrified of change."     - Alison Lowndes, head of AI developer relations, nVidia       metro.co.uk/2017/07/27/the-end-of-humanity-as-we-know-it-is-coming-in-2045   *5823