30 December 2017

Checkmate For Humanity?: In 4 Hours, Robot Teaches Itself Chess, Then Beats Grandmaster With Moves Never Devised In The Game’s 1,500-Year History — A True Watershed Event

John Naish, posted an article on the Daily Mail’s website, December 21, 2017, opening with this provocative question: “Will robots one day destroy us? It’s a question that increasingly preoccupies many of our most brilliant scientists and tech entrepreneurs,” he wrote.

Artificial intelligence (AI) used to be an area of research that belonged to a relatively small sector of the scientific and academic/research communities. Not any more. If there was one area that experienced a surge in advancement, practical applications, as well as showed signs of breath-taking potential in 2017, AI would be near the top of the heap, along with Bitcoin.

AI, machine learning, big data mining/analysis, virtual reality, augmented reality, 3-D printing, driverless cars, miniature/micro robotics/drones, autonomous systems, and blockchain technology are likely to be areas ripe for disruption and potentially staggering advancements in 2018. These are all areas that will be hiring like gangbusters in 2018, as these skills will be in very high demand.

But, as with almost any disruptive technology, there can be a downside; and, those that will sow fear and anxiety about what these potential downsides might be. I am not one of those who believe that we’re sowing the seeds of our own destruction with AI; and, I believe, like the Internet and Worldwide Web — that AI is going to be a huge net positive and will enrich our lives in ways we can’t even comprehend or understand right now.

“This year,” Mr. Naish wrote, “a leading Oxford University academic, Professor Michael Wooldridge, warned Ministers of Parliament that ‘AI could go rouge,’ that machines might become so complex, that the engineers that created them, will no longer understand them, nor be able to predict how they will function.” A case in point…..

“AlphaZero, an AI computer program, this month, proved itself as the world’s greatest ever chess champion, thrashing a previous title-holder, [and] another AI system called Stockfish 8, in a 100-game marathon,” Mr. Naish wrote. “What’s so frighteningly clever about AlphaZero,” he added, “is that it taught itself chess in just four hours,” grand champion, master chess that is. “In those 240 minutes of practice, the program taught itself not only how to play; but, developed tactics [and strategy] that are unbelievably innovative — and, revealed its startling ability to trounce human intelligence. Some of its winning moves had never been recorded in 1,500 years that human brains have pitted wits across the checkered board.”

“Employing your King as an attacking piece? Unprecedented,” Mr. Naish wrote. “But, AlphaZero wielded it with merciless, self-taught logic.”

Former world World Grandmaster Chess Champion Gary Gasparov, who famously defeated IBM’s supercomputer, Deep Blue in 1997 — when it was programmed with what was then the best moves — said: “The ability of a machine to surpass centuries of human knowledge…is a world-changing tool,” and perhaps a watershed event.

Simon Williams, the English Grandmaster remarked, “this is one for the history books and joked: ‘On December 6, 2017, AlphaZero took over the chess world….eventually solving the game and finally enslaving the human race as pets,’ Mr. Naish wrote.

“The wider implications are chilling,” Mr. Naish asserts.

As Mr. Naish explains, “AlphaZero was born in London, the brainchild of a U.K. company called, DeepMind, which develops computer systems that learn for themselves. DeepMind was bought by Google for 400M pound sterling in 2014. “The complex piece of programming that created AlphaZero,” Mr. Naish notes, can be more simply described as an algorithm — a set of mathematical instructions or rules that can work out answers to [complex] problems. The other term for it,” he adds, “is ‘deep machine learning’ tool. The more data that an AI such as AlphaZero processes — the more it teaches itself — by reprogramming itself with the new knowledge,” it has acquired. We talk about compound interest when investing. This is the equivalent of compounding knowledge. “In this way, its problem-solving powers become stronger all the time, multiplying its intelligence at speeds and scales far beyond the [known] capabilities of the human brain. As a result, it is unconstrained by the limits of human thinking [comprehension] as its success in chess proved,” Mr. Naish observed.

Will these super-intelligent AI’s evolve into the Hal 9000 of 2001: A Space Odyssey? — with a mind of its own? “How we keep control of [super-intelligent]deep-learning machines that will soon be employed in every area of our lives is a challenge that may prove insurmountable,” Mr. Naish speculates. “Already,” he adds, “Top IT experts warn that deep-learning algorithms can run riotously out of control because we don’t know what they’re teaching themselves.”

There is a lot of fear out there; and, a lot of hype. I believe that AI is going to enrich our lives beyond anything we can envision today. Yes, there will always be the darker angels of our nature that will use this relatively new domain for nefarious, and perhaps evil purposes. We will find a way to deal with those darker angels like we do now.

The 1982 movie, Blade Runner, staring Harrison Ford, depicts a future in which synthetic humans known as replicants, avatars, super-intelligent AI’s who all have a terminal switch or time bomb inside them. Left unsaid, is that this kill switch was embedded into the AI’s at the five year mark, so that they could not grow so strong and powerful — that they could wipe out humanity. Will we…..embed a backdoor, ‘hidden’ kill switch when the Blade Runner future — becomes reality?

One thing that the AI’s and avatars won’t have — at least in my lifetime — is ‘good old human emotion.’ Compassion, forgiveness, purpose, creativity, human spirit, the will to survive, the incurable and insatiable desire to explore and achieve….a human soul — these are the things that machines will likely never have. But, there may be another outcome.

Futurist Ray Kurzweil believes that by the 2040s it may be possible for us to live forever — when machines and humans merge into the Signularity. Mr. Kurzweil believes that by 2029, AI will pass the Turning Test, and surpass the level of human intelligence. In essence, machines will be smarter than us at that point. And, by 2045, Mr., Kurzweil predicts humankind will achieve Singularity — the point at which we will multiply our effective intelligence one billion fold; and, merge with the AI that we have created. At that point Mr. Kurzweil predicts — we will be able to live forever. 

Singularity “leads to computers having human intelligence, our putting them inside our brains, connecting them to the cloud, expanding who we are today,” Mr. Kurzweil predicts. “Today, that is not just a future scenario [or science fiction]. It’s here in part [already]; and, it’s going to accelerate.”

AlphaZero’s watershed victory, is validation that Mr. Kurzweil’s futuristic outlook is not as fanciful as some might think. We do indeed live in interesting times. RCP, fortunascorner.com

No comments: