Benefits and Risks of Artificial Intelligence

According to a recent article in MedicalFuturist.com, artificial intelligence (AI) will redesign health care with unimaginable potential. The author sees great benefits, and so do I, but he dispels the risks – risks that visionaries like Bill Gates, Elon Musk, and Stephen Hawking warn against. They warn that full development of AI could spell the end of the human race, and I share that concern. That’s why I’m writing today’s article with a cautionary tone.

The accelerating pace of change

At issue is whether or not man will find ways to guard against the dangers of tech innovation accelerating exponentially and indefinitely. The questions start with, what will AI, automation and robotics eventually do to employment? Which jobs will be replaced first, and which are safe for now? What might AI do for (or to) government? I don’t share the author’s confidence and instead side with the visionaries. Here’s why.

Moore’s Law

Among the 100+ Future-oriented articles posted on this site, one of my favorites is Moore’s Law and the Future of Healthcare. As illustrated above, Moore’s Law describes the decades old trend of computing capacity doubling every 2 years or less, enabling medical and sensor devices to keep getting cheaper, smaller, more capable, and easier to use. As that trend continues, many medical functions associated with doctors in hospitals will be able to move down-market to consumers at home. This is a good thing that can greatly reduce the cost of delivering care and even improve health.

The Moore’s Law article also envisions a gradual blending of technology and science (INFO + BIO + NANO + NEURO). But what might be the future impact of computer systems eventually shrinking to the size of a cell and that cost just $0.01? And what if each of these tiny processors had the processing capacity of the entire human race? That’s the potential of exponential compounding, where each new technology provides the tools that enable the next generation to come even faster.

Metcalfe’s Law

And what about the added effect of connecting these tiny computer systems to individual neurons in the brain, or networking them with billions of other tiny processors that can then work in synergy? That introduces Metcalfe’s Law, where Ethernet inventor Robert Metcalf noticed that the value of a network increases exponentially with the number of nodes attached to it.

Put together, Metcalfe’s Law accelerates the acceleration of Moore’s Law, so my biggest concern is with humans and our natural resistance to change, plus out inability to change fast enough. I think we could easily be overwhelmed, and more quickly than we realize, because of that compounding effect. That would cause all sorts of social problems, including further widening the wealth gap as capitalists at the top are more able to benefit from tech investments.

Wealth Gap

Over time, through the agricultural, industrial and information ages, tech innovation has increased worker productivity and generated wealth, but more and more of that has flowed to the top. But few people, Democrats and Republicans alike, have any real idea what today’s inequality actually looks like. That’s why I refer you to Wealth Inequality, Healthcare and the Economy. A video infographic in that article is the best illustration of the problem I’ve seen so far, but it still doesn’t adequately show how the gap widens exponentially with tech innovation.

The chart above shows the 2008 population distribution and how the value of 1 hour of labor is falling. With control of capital markets, the wealthy 1% has been better able to invest in automation, robots and AI – to produce more with less effort (fewer workers). And because demand for labor fits the supply & demand economic model, when workers outnumber the need for them, wages fall or stagnate, or jobs disappear entirely. That further widens the wealth gap, makes more people dependent on government assistance, and further burdens our healthcare system.

This next chart shows that the top 1% make much more money ($27 million) than the bottom 90% ($31 thousand). But that’s not the half of it. Income distribution is widening, with those at the top earning 46% more in just two years, while the bottom 90% earned 15% less.

Clearly, AI will exacerbate this unsustainable trend, which already demands corrective action. But what actions might that be? How will AI effect them? And will society even be able to evolve as fast as technology? As a retired IBM technologist who understands the benefits but also the challenges of adoption, I doubt it.

Business and Public Policy Impact

Already, politicians have been unable to keep up with the need for policy changes brought on by tech innovation, and even as Congress needs to speed up, it seems to instead be slowing down. The Republican controlled Congress during the Obama administration, for example, passed so few new laws, and opposed nearly all Obama proposals, that it became known as “a Do Nothing Congress.”

The same concerns about slow decision-making also apply to corporate leaders, who measure success in business terms like profit margin, return on investment, quarterly stock price, and payback period. That focus causes them to make relatively short-term decisions without enough long-term visioning. Even though lawmakers could change the structure, oversight, and charter of corporations to influence behavior, there’s no political will to do that. Instead, corporations keep increasing their power and ability to control the political process.

The Corporation, a documentary

The Corporation describes a history of this legal entity and how it has morphed into what now must serve the investment interests of shareholders before everything else, even the public good. Because the US Supreme Court gave corporations the legal rights of a “person,” maybe we should examine corporate behavior in psychological terms. If so, we’d find typical behavior is like that of a psychopath. Could artificial intelligence evolve in a similar way?

Even with efforts to create ethical standards for AI, it’s not at all clear that corporations would adhere to them when profits have higher priority. Neither would rogue nations that want a competitive advantage. That’s why I fear that human nature, natural selection, and the concept of “survival of the fittest” would likely cause bad actors to circumvent any AI ethical standards put in place, for their own greed or ambition.

What We Can Do – You and I

Don’t despair. We citizens still have a lot of power, and it starts with our votes. If we’re unhappy with slow progress in Congress, we can vote out those who oppose it and vote in those with a vision and a plan to better exploit technology, make needed investments in education and infrastructure, make necessary changes in campaign finance and the lawmaking process, and even decide how to evolve the Constitution to allow this if necessary.