IEET > GlobalDemocracySecurity > Vision > Advisory Board > Daniel Faggella > HealthLongevity > Futurism > Innovation > Artificial Intelligence > Biosecurity
The Future of Artificial Intelligence – Separating Facts from Fictions
Daniel Faggella   Oct 25, 2015   Ethical Technology  

The news is a tough nut to crack in today’s over-stimulated and often sensational, media-driven world. This is true more than ever in the coverage of artificial intelligence (AI). Many of us are not sure if AI is going to wake up any moment and wreak insidious havoc, taking over or destroying society as we know it.

Dr. Andras Kornai is all about separating AI fact from fiction.

In a recent interview, Andras also made clear that he is adamant about convincing the public that much of their assumptions about AI are real and not relegated to fiction. AI is with us now, and it is everywhere – but it’s nowhere close to becoming aware of itself and making intention-driven decisions.

While many are aware of the reach of AI into our interwoven, digital lives – our computers, phones, and other “smart” and mobile devices – many are unaware of the algorithms that drive much of this technology, which has led to the boom in the past decade in deep-learning. In fact, the biggest societal change in the past 20 years, says Andras, was with the introduction of cell phones. “Instant communication, it’s not overtly visible…it’s subtle,” he says.

AI, the Invisible Shaper

The same concept is true of much of the AI that is already shaping our worldly interactions. In high frequency trading (HFT), for example, computer algorithms already perform about 90% to 95% of the quotes – humans have already, more or less, been uprooted in this industry. Not many people outside of the financial industry see or know that this is the case. “The mere idea that we have given up on human decision-making capabilities in favor of better algorithms…this means we haven given up a huge area of human competence,” says Kornai.

Other industries have also felt the algorithmic pull. The medical industry has seen a surge in algorithms, which sorts through unstructured and numerous data to help doctors make better diagnoses. The fact that IBM has entered the market with Watson Health is evidence of the lucrative and far-reaching potential of this technology.

Algorithms are all the rage in self-driving cars, a hot topic in part due to Elon Musk’s and Tesla’s pioneering success, followed closely by Google, as well as a series of automakers entering into research partnerships to help pave the way for their place in the future industry. Andras points out that as of now, we (humans) still make the decision behind the wheel to swerve into a human or into another object, but these decisions will soon be turned over to algorithms.

Autonomous drone vehicles and weapons also need no introduction, which at this point are a prominent military concern. Humans still control these drones, but they certainly have the capability of being programmed and then put in control of their own actions, says Kornai. While the drones are not “out to get us”, the capability and the potential threat of use in the wrong hands is a real and present concern, not a future one.

A Complex Future

Where will we notice increased involvement in the next 10 years? Andras believes it will be in many of the same industries mentioned, but the ethical and legal issues and arguments will become more rounded and complex.

The current legal constraints with new technology used by Tesla, for example, will yield legal clashes with other car manufacturers. Self-driving cars could also become a point of conflict with companies like Uber, who want to take advantage of these vehicles, but who are already having run-ins with governments worldwide (France being notable).

Kornai also notes that there will be a shift in our national and global economies, with increased automation (and there seems to be no going back) leaving many people unemployed. We’ll have a much better idea of the industries that will be affected, but at this point there seems to be almost no limit (though many would exclude the arts) into which AI will, eventually, be able to take the place of humans, simply due to better cost, efficiency, and accuracy.

In the medical domain, Andras suggests that legislation is much more conservative; he believes we can expect the development of AI to be much slower, with no runaway tendencies. In contrast, the business world may have fewer barriers than in other arenas. Businesses are increasingly recognized as legal entities, and will be more independent and have more of an opportunity to freely invest in AI, which will play a bigger role right in the boardroom. Andras poses that “AI will be very visible in the Fortune 500 world, but less visible to the everyday person on the street.”

Last but not least in the public’s mind, what about all the hype surrounding the petition and comments made by Hawking, Gates and others from the Future of Life Institute surrounding autonomous and dangerous AI? “The arguments are reasonable…I don’t look at it as an investment (developing AI), but more like insurance,” remarks Kornai, who is working on developing bounded AI  and other beneficial forms of AI. “These are relevant concerns, even if the Hollywood fears are creeping in.” After all, farsighted is not science fiction, says Andras; it means you think ahead 25 years, well within our lifetimes, and counter the possible negatives with positive solutions.

Daniel Faggella is the founder of TechEmergence, and blogs at SentientPotential.com.



COMMENTS

AI doesn’t know it’s replacing jobs. It doesn’t have the intention of doing that. It doesn’t know that it’s work is more profitable and can’t judge the value of that. Yet.

The employment problems come from the exponentially accelerating pace of tech innovation, automation and AI, enabled by Moore’s Law, while humans evolve far more slowly. In short, the machines are getting smarter faster than we are, and people like Gates and Hawking worry about what happens when they become smarter and develop their own wisdom and morality.

One way humans can cope is through public policies that encourage lifelong learning to develop new skills for your next career without needing yet another degree, and another student loan, that may only last for a year or two before doing that all over again.

We can provide financial safety nets for displaced workers while they retool and find new purpose, but at some point we may lack the intellect or ability to outthink the machines when something goes wrong.

YOUR COMMENT Login or Register to post a comment.

Next entry: Hey Humans—Robots Are NOT Better Than You!

Previous entry: Monsters on the Brain: An Evolutionary Epistemology of Horror