Will we at some point in the future elect an artificial intelligence (AI) to higher office?
Would you ever vote for an algorithm to represent you in Congress, to manage your city, to be your President?
Could we ever vote for an Abraham Lincoln 2.0 - an artificial intelligence built to replicate the decision making style of one of our most accomplished Presidents?
It sounds ridiculous doesn't it?
Electing a software reincarnation of the man consistently rated as the greatest American President certainly sounds ridiculous. And that's exactly why the idea is worth exploring.
Artificial intelligence is racing ahead. IBM's Deep Blue defeated chess phenom Garry Kasparov in May, 1997. IBM's Watson defeated Jeopardy! trivia stars Ken Jennings and Brad Rutter in February, 2011. In 2014 "Eugene Goostman," a chatbot created by PrincetonAI, fooled one third of the judges into thinking it was human and controversially passed the annual turing test held by the University of Reading. In 2014 Deep Knowledge Ventures, a Hong Kong venture capital firm, named an algorithm called VITAL to its board in order to augment the group's decision making.
But would an AI ever become the boss, giving orders to humans? That's already happened. All of the maintenance and repair for the Hong Kong subway system is run by an AI invented by Andy Chun of Hong Kong's City University. MTR Corporation, the owners of this technology, plan to scale it across other transit systems.
What will AI be capable of in 20 years? In 30 years? Here we have two illuminating clues. First, famous inventor and technologist Ray Kurzweil is working to build a powerful new artificial intelligence at Google. Kurzweil, the author of How to Create a Mind, is now hard at work attempting to replicate a human mind in digital form. Our second clue comes from the US Army Research Laboratory (ARL) and its June, 2015 report titled "Visualizing the Tactical Ground Battlefield in the Year 2050." Section 2.2.6 is titled "Cognitive Modeling of the Opponent." The report assumes that "a vastly improved capability to understand an opponent and predict their actions will enable a new and potentially disruptive capability in this time frame." In other words, the US military will replicate an adversary's personality and thinking with AI and test strategies against that AI opponent. This is not a new idea. It should be familiar to fans of the strategy game Civilization, in which human players compete against a predetermined number of AI opponents built to replicate the leadership styles of well-known national leaders (of course there is the well-known Gandhi glitch). It's clear that someone, somewhere will create AI that comes close to replicating a human's decision-making process. While it seems obvious that the likeliest creators will come from Silicon Valley tech or the military, my money is on marketers and the consumer packaged goods sector, as they have the most to gain.
As the saying goes - "Software is eating the world." Technology is disrupting almost every industry. And politics is certainly an industry. Almost every part of politics itself is being disrupted by technology, from advertising to communications to voter targeting. The candidate appears to be the lone exception. Replacing the candidate with artificial intelligence may sound insane today, but our species has been brilliantly ceding power to technology for a very long time.
As a starting point, let's consider three technologies that you use today, the elevator (or "lift" in the UK), the passenger jet and the automobile.
If you live in an urban area, you use an autonomous, robotic elevator every day. You push two buttons and trust that an unmanned, robot elevator will hoist you hundreds of feet into the air and deliver you to the correct floor without killing you. There was a time when elevators were operated by humans, but that job was automated decades ago. There are no protests on the streets against drone elevators. You have relinquished human control in this area because you have faith in the technology. People may be stuck in drone elevators now and then, but they rarely die in a drone elevator.
Now let's take flying. We don't need a pilot in the aircraft to fly the plane, but you still want a pilot on your plane. The pilot is in the cockpit, but you know that "autopilot" flies the plane most of the time you are in the air. We have relinquished most of the human control over flying and could relinquish all of it, if we weren't so afraid of our new found flying ability.
E.B. White famously noted that "everything in life is somewhere else, and you get there in a car." When he coined this phrase, people used maps, asked for directions and steered the car. We've already ceded the first two to technology and are working on the third. When was the last time you double checked the GPS by looking at a map? When was the last time you asked a human for directions?
Auto companies and technologists are now banking on the fact that you will outsource your car driving to AI. Doing this will almost certainly reduce your insurance rates, reduce traffic fatalities and reduce the flow of organs for transplant. It will increase road safety. You won't relinquish control of the wheel overnight. Instead, you'll likely do this in phases, first by moving to autonomous parking, then adding autonomous highway cruising and finally relinquishing everyday driving control. Manned driving could become an expensive pleasure activity that hearkens back to yesteryear.
In each case, with the elevator, the passenger jet and the automobile, you have ceded or will slowly cede personal control over a mode of transportation to artificial intelligence. Moreover, you have done this across a spectrum of activities, including your investments, which are largely a collection of stocks balanced and rebalanced each day by artificial intelligence.
The reader may agree with all of the above and still remain skeptical for three basic reasons:
1. Doubts about the ability of AI to ever replicate the workings of a mind
2. Skepticism that humans would ever elect an AI to lead them
3. Fear that AI may break free of its master and attempt to harm its creators
The first objection is based on a skeptical assessment of our technical ability to create extremely powerful AI built upon immense processing power. This is curious. We are living in a world of exponential change. The phone in your pocket would have been a national strategic asset in the 1950s. This first objection may really be based on a quasi-religious view about the exceptionalism and uniqueness of the human mind, that we are "a little lower than the angels." Many of us may feel that creating an AI version of a human mind is "playing G*d." But, we've been doing that for some time now with birth control, flight, atomic power and genetics.
The second objection is "the creepiness factor" of voting to put a piece of software in a position of authority. But this assumes that we will continue to interact with AI at arm's length. In reality, we will co-evolve with AI much as we co-evolved with dogs, using them for early warning, security and friendship. Children today interact with technology seamlessly. They ask Siri silly questions. Computing technology has inexorably moved from a distant thing to an intimate accessory. Just as computing has moved from a building to a room to a desktop to a laptop to a tablet to a phone and to wearables, AI will do the same. Will an AI candidate seem creepy when we all own one or more personal digital agents that make recommendations to us, warn us against danger, monitor our health and entertain us?
The final objection is based on the fear hard wired into us by millennia of evolution. Will AI become a threat? Many futurists and technologists feel that this objection is almost entirely due to Mary Shelley and her book Frankenstein; or, The Modern Prometheus (published in 1818). The dark and feverish dream is that we create something that destroys us. That this dystopian view of our creative potential is widely considered the first science fiction novel is even more problematic. The reasoning is simple. We are flawed. We create a flawed thing and the flawed thing kills us. This plotline is a recurring technology meme. Think HAL from 2001 Space Odyssey. Like so many things about the modern era, the singularity we know as Goethe may have gotten there first with his 1797 poem Der Zauberlehrling - what we know as "The Sorcerer's Apprentice" or the animated film Fantasia with Mickey Mouse. Here the apprentice is out of his depth. His magic runs amuck and only the skill of the sorcerer can save the day.
Of course, an alternative and more hopeful meme is that we create a thing that saves us, like a vaccine or a drone that delivers medicine. But, that may not be as entertaining.
There are six (6) reasons why we may someday vote for an AI candidate built to replicate one of our great leaders:
1. Exponential Technological Capacity: We will grow our computing and software capacity at an exponential rate and replicate the thinking of a human mind. Consider that in my 44 years of life computers have gone from something you house in a building to something you have in your watch.
2. AI as Person: There are a number of efforts to digitally replicate the thinking and personality of an individual. Just one example is eterni.me. Eterni.me asks consumers to "become virtually immortal." Eterni.me "collects your thoughts, stories and memories, curates them and creates an intelligent avatar that looks like you. This avatar will live forever and allow other people in the future to access your memories." Eterni.me wants to do for individual consumers what the US military wants to do to adversaries - create software that thinks and communicates like a specific human. These efforts will advance rapidly. Former Presidents provide us with an enormous amount of source material. Scholars have studied and mapped their decision making styles. They have left us with decades of speeches, correspondence, first-hand accounts and memoirs. All of this material could be used to build an artificial intelligence that replicates their thinking. Presidential libraries may be the drivers of this.
3. Comfort with AI: We will become more comfortable with AI over time and as we bring it ever closer to our person. Computing power moved from a building to wearables. The same will happen with AI. GenZ, the homeland generation, are digital natives and will be completely comfortable with AI that their Boomer grandparents may find creepy.
4. AI as Management: Distributed Autonomous Organizations (DAOs) will challenge more traditional organization structures and may disrupt senior management. If this happens, the idea of AI leadership will not seem so strange.
5. Golden Ageism: An AI candidate modeled on the greatest former President is a high tech appeal to nostalgia and golden ageism. It all may sound crazy today, but go and look at all the Presidential Twitter handles. Some impersonate a famous President, like @dick_nixon. Others helpfully transmit the famous phrases of their namesakes. It's worth noting that some of the largest cheering at Nationals Stadium is for the racing Presidents.
6.Novelty: Since it opened on October 1, 1971, Americans have flocked to the animatronic Hall of Presidents. In a more timely example, consider the "humanoid persona" and Japanese music star Hatsune Miku (translated as "the first sound from the future."). Her first "live" concert was in August, 2009. Adoring fans came to Saitama Super Arena to hear her "sing." A holographic avatar like the Tupac image dancing at Coachella, Hatsune Miku is a music sensation. An AI reincarnation of a past leader could develop a similar superstar following. Just imagine a Lincoln 2.0 debating contemporary candidates.
The futurist Jim Dator once wrote that "much that will be characteristic of the futures is initially novel and challenging. It typically seems at first obscene, impossible, stupid, 'science fiction', ridiculous. And then it becomes familiar and eventually 'normal.'"
The idea of voting for an AI candidate certainly seems ridiculous at first, but it is clearly not out of the question.
Perhaps one day we will see "Elect Lincoln 2.0" lawn signs and make a micro donation as we walk past.
Would it have surprised our 16th President, an avid user of the greatest technology of his day, the telegraph - the Victorian Internet?