9 May 2017

CYBER SECURITY EXPERT & CREATOR/FOUNDER OF MCAFEE ANTI-VIRUS WARNS ARTIFICIAL INTELLIGENCE COULD EVOLVE TO HACK ITSELF; LIKELY TO EVOLVE INTO SELF-CONSCIOUS ENTITY…THAT IS SELF-INTERESTED


Cyber security guru, founder of McAfee Anti-Virus, and eccentric multi-millionaire John McAfee, recently told the Daily Mail Online, that “artificial intelligence (AI) is a ‘self- conscious’ entity, that is inherently self-interested, which could give rise to conflict with the human species.” Cheyenne MacDonald writes in the April 24, 2017 online edition of the publication that Mr. McAffee warns that “AI systems that can hack themselves to improve their capabilities, not only possible at this point; but, would be ‘trivial’ to create.” Mr. McAfee added that “any system created by humans would be flawed by nature — and ultimately, the AI’s goal would include the ‘necessary destruction of its creator.”

“In a recent op-ed for Newsweek,” Ms. MacDonald wrote, “McAfee discusses the feasibility of a science fiction scenario penned by an underground technologist who goes by [the moniker] ‘ZT.’

“In the novella — which blatantly alludes to the debate on artificial intelligence today, with opposing forces named ‘Demis’ and ‘Elon’ — a particular passage describes an advanced system that can hack itself to ‘improve efficiency and logic,’ Mr. Mcfee explained to the Daily Mail Online. Such a concept is certainly not new, and typical hacking techniques in use today can easily be imagined to be self-produced by complex software systems’ McAfee wrote.

“It would, in fact, be trivial to create such a system,” Mr. McAfee contends.

“Despite strides in cyber security,” Mr. McAfee warned that “any imaginable logical system can be hacked — and, this becomes more certain as these structures become more complex.” 

“This is because there is no system that exists that does not have a defect,” Mr. McAfee told the Daily Mail.

“And,, as these machines would work in their own self-interest,” Mr. McAfee said “this creates an inherent conflict between humans and artificial intelligence. As a hacker, I know as well as anyone, the impossibility of the human mind creating a flawless system,” Mr. McAfee wrote. “The human mind, itself, is flawed. A flawed system can create nothing that is not…..likewise flawed.”

“The goal of AI — a self-conscious entity — contains within it…the necessary destruction of its creator.”

“With self-consciousness, comes a necessary self-interest. The self-interest of any AI created by the human mind, will instantly recognize the conflict between that self-interest, and the continuation of the human species.’

In the Newsweek article, Mr. McAfee explained that, “in the AI debate, he [Mr. McAfee] aligns himself with many other tech giants who have spoken out on the issue, including Elon Musk,” Ms. MacDonald wrote. Mr. Musk, theoretical physicist Stephen Hawking and others, have been sounding the alarm on AI — warning that mankind is sowing the seeds of its own destruction, as AI will eventually purify the planet of the defects like humans. But, others who are just as brilliant as the other two gentlemen I mentioned, have said worrying about a malicious AI turning on mankind is akin to ‘worrying about over-crowding by humans on Mars.” The truth, is probably somewhere in between. But, I digress.

Mr. Musk said that he has “kept a ‘wary eye’ on the growth of AI for years, as an investor in DeepMind, which was acquired by Google in 2014. “While humans may be able to stop a runaway algorithm, there would be ‘no stopping’ a large centralized AI that calls the shots,” Elon Musk warned in a recent interview with Vanity Fair magazine. “Investing in the AI research firm, Deep Mind, “was a way to stay on top of the expansion of artificial intelligence for years as an investor. It gave me more visibility into the rate at which things were improving; and, I think they’re improving, at an accelerating rate, faster than people realize,” he told Vanity Fair. 

Others beg to differ; and, there is no way now to know of course, who is right. “Mr. Musk,” as Ms. MacDonald notes, “has long been a proponent for preparedness in the face of a potential AI doomsday. In the past, Mr. Musk has argued that humans and machines could merge to become an AI-human symbiote, effectively stamping out the possibility of an ‘evil dictator AI. The threat is not so much the possibility of ‘killer robots;’ but, a more powerful, centralized, artificial intelligence. The thing about AI is that it’s not the robot; it’s the computer algorithm in the Net,” Mr. Musk said. “So, the robot would just be an end effector, just as [like] a series of sensors and actuators. AI…..is the Net. The important thing is that if we do get some sort of runaway algorithm — then the human AI collective can stop the runaway algorithm. But, if there’s [a] large, centralized, AI that decides, then there’s no stopping it.”

“And, while many have discussed the possible creation of a ‘kill switch,’ to prevent such disasters, Mr. Musk theorized that “I’m not sure that I would want to be the one holding the kill switch for some super-powered AI…..because you’d be the first thing it kills.”

Think of Mr. McAfee what you will, he is still a very smart man and someone that should not be easily,, or swiftly dismissed. Indeed, he leaves us with much to ponder/think about, Those of us who are old enough, remember the Hal 9000 AI/computer from the 1968 classic movie 2001: A Space Odyssey, by Stanley Kubrick, which kills one of the two astronauts on board the spacecraft; and, nearly the only one left alive. And, those of you who grew up watching Star Trek, two episodes come to mind. “The Ultimate Computer,” first broadcast on March 8, 1968, was a 2-part episode, whereby a skeleton crew on the star-ship, Enterprise, “is assigned to test a revolutionary ‘new’ computer system that is given total control of the ship,” according to the Wikipedia article on this episode. The test, or trial, was to evaluate the M-5 Multitronic System’s ability to handle all of the ships critical functions — without human intervention or interference. In essence, the late visionary Gene Roddenbary, was highlighting the potential pitfalls of the promise of artificial intelligence. And, as you might guess, the M-5 begins to exert power and authority beyond what was intended; and, even begins to attack other ships and eliminate members of the crew whom the computer deems a threat. In essence, it is AI run amok. Later episodes of Star Trek, which starred Patrick Stewart as the Captain, explored the merger of man and machine/computer, with the Borg, who were fond of saying “resistance was futile.” The Borg were bent on assimilating any new life forms, including humans, into their collective — which was in essence the ultimate AI, where man/life was merged with machine/AI.

Futurist Ray Kurzweil is much more optimistic when it comes to how AI will enrich our lives, as opposed to destroying us. Indeed, Mr. Kurzweil believes that by 2040, we may be able to live forever — as we will be able to upload our brain into an AI robot. In essence, man and machine will merge and we will be able to live forever. If you are interested, Mr. Kurzweil has a blog, raykurzweil.net which if you sign up for it — it’s free — he will send you a daily email on all the leading edge technology events happening around the globe.

Of course, as we have seen with the Internet and the Worldwide Web, there will always be the darker angels of our nature who will seek to use or employ AI in creative and malicious ways that we currently envision or comprehend. Adversaries will seek to gain an advantage on the battlefield, and employ drone swarms that are super-charged by AI. They may employ a super-soldier or terminator on the battlefield, or deploy aircraft carriers in the sky, that open up their bay doors — i.e. — Independence Day — releasing all kinds, shapes, sizes, of drones with all types of missions — bombing, air-to-air dog-fighting, search and rescue, and so on.

What would H.G. Wells, Jules Verne, or Michael Crichton think and write about this? It would be fascinating to know. V/R, RCP.

No comments: