Will humans one day elect an artificially intelligent robot as president?
After all, humans tend to make decisions based on ego, anger and self-reinforcing needs, rather than what is good for the masses. An AI president can be trained to work for the benefit of the vast majority without infringing on civil liberties. It might even know that the president tweets as little as possible, or not at all. At first glance, though, the idea of a robot as president seems far-fetched and absurd. I wonder if there’s an algorithm smart enough to host a state dinner, for example.
Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include information robot, editing robot, writing robot and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.
Nonetheless, machine politicians may well be the future. Think about cars. Tesla owners are driving their Model S on their own, and other automakers are racing to build cars that don’t even have steering wheels. Within a decade, thousands of people will happily entrust their daily commutes and their security to algorithm-based machines.
Why is that? Because it will make their lives better. The driver is no longer a driver, but a passenger, who can sit in the car to watch a movie or do his own work. This has a huge impact on human productivity and happiness. At the same time, smart cars will also improve driving safety. In the United States alone, more than 30,000 people die each year in traffic accidents, almost all of them caused by human error. That number will fall sharply when self-driving cars hit the market.
Similarly, humans are bad at managing themselves. The US government is Mired in gridlock, invective and partisanship. People no longer vote because of policy positions, but because they like the way they look or the way they talk. People elect politicians who want to embody their ideals and values, and are disappointed when they trick interns into handing out unmarked bills. Politicians are expected to embody the highest ideals of their people and their country. But usually politicians don’t. But a purer form of government is possible with an AI president that focuses on the realization of the national ideals it represents.
Voters can choose between democratic or Republican AI, but both robots promise to pass the party’s platform. Alternatively, voters could simply vote on a bunch of issues, and the AI, which reflects the will of the people on every one of them, would win. Of course, even with a clear platform, the answers to the big questions are fraught with complexity: should the constitution be taken literally, or interpreted differently to suit modern circumstances? How do we tackle poverty, inequality and welfare? Any action to address these problems could lead to unintended consequences. The president needs to react to the new situation and think about the many steps ahead and make tough choices. “When I have something on my desk, it means it’s really hard,” President Obama said in 2009.
The idea that an algorithm might do better than humans did when they entered uncharted territory seems unrealistic. But recent events have undermined that assumption. In the past 12 months, Google’s AI has won 60 matches against the world’s best Go players. To do so, it must master a game far more complex than chess. (In the known universe, go has more possibilities than combinations of atoms.) Ai is faced with a multitude of choices and has to consider dozens of steps. It requires making difficult decisions, developing a strategy that involves risk, and operating with incomplete information. It did all this, and it innovated.
Professional player Myungwan Kim said: “The robot did something we haven’t seen before, it won. We thought it would take 50 years for the software to beat the best player in the world, but in just five months it became the best player in the world. In fact, the chess program on your mobile phone in your pocket has more strategic knowledge, past or present, than most residents of 1600 Pennsylvania Avenue.
Remember artificial intelligence during the 2003 presidential term? The software analyzed decades of reports on Saddam Hussein, summed up various intelligence on WMDs, and concluded that invading Iraq was clearly a stupid idea, unlikely to spread democracy. The same is true of Vietnam. Of course, the real world is different from chess. But ai robots are already grappling with complex problems, and they’re getting better at it.
In a paper recently published in the journal Nature, researchers detail how AI was able to diagnose melanoma better than board-certified dermatologists with decades of experience. (Other A.I. s are also being trained to detect diabetic blindness and lung cancer.) These programs, driven by neural networks that mimic the human brain, can ingest vast amounts of information and learn from mistakes, two qualities sorely lacking in today’s politicians.
Facebook’s Jan Lehn, one of the leaders in deep learning, predicts that it won’t be long before machines begin to understand language, motivation and thought. “Our relationship with the digital world is going to be completely transformed by the fact that you can interact with intelligent robots,” Lern told MIT’s Technology Review in 2015.
We’re at the beginning of a deep learning revolution. Soon, AI will be mature enough to implement a core set of beliefs in a way that reflects changes in the world. In other words, AI will soon be more judicious than most politicians. All of which suggests that having an AI as president will only become more persuasive. In the modern world, human presidents have become increasingly inundated with information. We’re not designed for the digital age. Most of us can’t process email inboxes, let alone data from government agencies, the economy and the military.
An AI driver can simultaneously receive 360-degree information, see further and react faster than any human. Can an AI president do the same? Indeed, AI imbeds biases, blind spots, and the cultural and philosophical assumptions of its creators. But the promise of AI is that our ultimate goal is standard, and to figure out how best to get humans there. So, like all presidents, ai leaders will do their best to keep the majority of voters happy within the law. Unlike many people, AI can overcome biases and assumptions that don’t help achieve its goals.
To some extent, the emergence of artificial intelligence seems like a natural evolution. More and more people are entrusting their lives to machine intelligence. The most frequent crashes on long-flying commercial airliners occur when pilots switch to manual control. In manufacturing, there are more than a million robots working on everything from trains to computers. Every aspect of people’s lives is digital, from bank accounts to entertainment. Whether people know it or not, any given experience can be optimized by ARTIFICIAL intelligence to make life easier than it was a decade ago. Humans are also getting used to the idea that digital services know more about them than they do about themselves. When Amazon shows us a book or movie we might like, we’re likely to like it. People also appreciate email filters reading our messages and opening their inboxes. An AI president might be able to come up with policies that would be satisfying, just as Netflix has the same uncanny foresight for shows and movies that people love.
On the other hand, an AI president could be hacked or could turn into a demon because it can learn independently and carry out its mission of optimizing happiness for the majority of citizens. To demonstrate dystopian possibilities, Tesla CEO Elon Musk used the example of artificial intelligence to grow and harvest strawberries. To accomplish the primary task of picking as many strawberries as possible, ai might conclude that it makes sense to wipe out humans and turn earth into a giant strawberry farm. Imagine what would happen if we gave ai the nuclear launch code. How can we surrender our lives to a machine? Even if AI is set to maximise human happiness, it may decide that the threat of global warming is enough to justify killing millions of people to reduce carbon emissions.
The truth is that anything ai learns from its consumption of vast amounts of data is opaque to us. An AI president, like a person, would be a black box. We spend a lot of time imagining an anti-future associated with machine intelligence. In terminator, Skynet turns evil and tries to wipe out humanity. In the Matrix, machines turn humans into batteries, and in 2001, Space Odyssey Hal is a complex threat.
Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.
But that doesn’t have to be the future. Despite the risks, the potential benefits of AI cannot be ignored. Ai aims to keep humans away from things they are not good at, such as driving, doing medical diagnosis and management. It is entirely possible that ARTIFICIAL intelligence will lead mankind to a new era of prosperity and peace. None of this will happen quickly or immediately. Increasingly, leaders will turn to intelligent robots to help them analyze complex situations and game out the best options. Step by step, it is becoming clear that AI can handle more and more difficult decisions.