Saturday, April 8, 2017

Artificial Intelligence and Singularity

I am sitting in a Beverly Hills surgeon's office and thinking about mortality. I'm waiting for a friend to get out of surgery and reading an article about A.I's threat to our very existence. We are so fragile. I feel like human civilization is an oxymoron. For most of our existence we have acted like the kid who gets his first bike and decides to go practice on the 405 freeway: a reckless daredevil who is astonishingly ignorant, supremely overconfident in our abilities, battling our own death impulse and the gravitational pull toward the grave with this addiction to adrenaline. We are shocked when we die or when our bike trip goes poorly. I'm shocked that it doesn't go wrong more often. It's strange to think that our innovation and reckless pursuit of an faster, bolder and less ethical future has put me in this office. It has created this world that is going faster than our ethical understanding of life and I think it might be too late. We have passed to many historical markers where we could have stopped ourselves and reflected on the impact of our progress.

AI is tied to our treatment of ourselves in history because this intelligence will learn from humans how to treat us when technology soon become more intelligent, powerful, and omniscient. Singularity is around the corner and we have not prepared for it. In recent human history there are two pivotal moments that we could have learned from in our rush toward this point. The first one is the end of the Cold War. For the latter half of the 20th century communism was the main threat to Western culture.  Our society was focused and organized around. When the Cold War ended a power vacuum opened up. At that point we should've gotten together, sat down, and drawn up some tech ethics for future research. When the nightmare of communism was removed as the focal point of our progressive struggle, there was only one place to look toward for our next enemy: the mirror. We were going to be our worst enemy in the 21st-century with no obvious imminent evil. That is not to say that terrorism is not a problem or that we don't have evils, but there was something about the Nazis and communists that captured all the attention so fully it was actually productive to organizing technological innovation, ethics, and laws.

The next period of time sounds weird at first but it immediately popped into my head: after the American Civil War. During the Reconstruction of America we could have created a different narrative for this country. Instead government just reinforced all our old hatred and racism and use that as a guide into the next chapter of American history. So much of what we do that is intentionally distractive in this culture comes from the original sin of a so-called free society trying to find ways to control slaves in antebellum and then freedmen after the Civil War while keeping poor White people in the dark. We have created an entire system of justice and laws and concepts and ethic that are destructive. But we keep these systems in place because they are extremely effective in preventing the average man from every uniting together for their betterment. Here is why that matters with AI: self-learning technology enhances human tendencies.

A few years ago Microsoft put AI bots on the Internet who could train themselves based on human interaction. The bots became extremely racist and misogynistic within a few hours. They embodied all the trash and awfulness of mankind...and it only took a few hours of ingesting our society. They were picking up on the way we train ourselves. We are not good models for artificial intelligence. In America particularly we're not a good model because our society has been designed around a post Civil War concept of degrading and controlling people by race. AI will use that. It will be like Trump's blitz of lies and hatred but much worse. AI will be like 1 million little bot Trumps who are playing on human stupidity, ignorance, and hatred to drive us over a cliff.

The red eye flight I just took back to Los Angeles was almost certainly flown mostly by autopilot while we, the human cargo, alternated between snacking and napping in front of blue screen. A cab GPS'ed me home from LAX and when the driver asked if I had cash I looked at him like he just requested to be paid in pirate's treasure or the beads that bought Manhattan. I am sitting in a sleek and automated doctor's office with one receptionist. Years ago this office would have been filled with multiple receptionists, rows of paper files, and the messy stench of...humans. Now it is quiet and streamlined. Even though the job of the surgeon is cutting open human flesh and is intimately messy, the process feels bloodless from the outside. I could be sitting in an architecture's office or the HQ of a new tech start-up. The automation and streamlining of our existence should be something we're thinking about a lot more, but AI hums quietly in the background of every single thing we do and it makes me vaguely suspicious, like a Luddite who is introduced to zippers and thinks the tailor is a witch. I write this on an iPhone in a surgeon's waiting room thinking about our fragile human body and our strange recklessness as we rush into the future.

No comments: