Skyney is Coming, No Really (Technological Singularity)

Last spring I was on vacation in Florida and took the hour drive from the beach into Orlando to visit Universal Studios. Many people elect for one of Disney's many, many family-oriented theme parks, but with no kids, and a severe affection for roller coasters and thrill rides, Universal Studios was my choice.

After spending, oh, I don't know, about $450 for a single-day pass for two people (seriously), I was entreated to a theme park full of movie-oriented attractions and rides representing today's most popular entertainment brands. Harry Potter. Shrek. Transformers. Despicable Me. Men in Black. Marvel Comics. The Simpsons. It was an all-out onslaught of merchandise and marketing. (If you're still with me, I promise, this is getting somewhere.)

Yet a few less-popular brands still exist in the park. Twister has its own ride. Beetlejuice has a show. I expected these fading brands to be replaced by more popular ones. I even attended a 3D live show of Terminator 2, complete with a Schwarzenegger imposter on a motorcycle, animatronic Terminators and mid-1990s 3D technology. It was obviously a let-down considering how developed 3D technology has become in the 20 years since the release of T2, and it was no surprise that the theater was less than a quarter-full during my showing.

So why does Terminator 2 3-D: Battle Across Time still stand as an attraction today? There are dozens of more profitable replacement options for an overage, unpopular attraction with a fading brand. Could it be that this live interpretation of humanity's battle versus Skynet is powerful reminder that the technological singularity is inevitable? Or is this blogger just grasping at straws while trying to segue into a post about such a topic?

It's definitely the latter, but at least hear me out.

You might remember the name Alan Horn--he is best known one of the lead developers of IBM's Watson, the supercomputer that appeared on Jeopardy! and is now used to assess treatments for lung cancer. Horn recently argued that whether humans want it to or not, technology will develop to a point where it begins to develop itself. It will be a cascading series of machines building machines, and it will make humans obsolete in every way, an event known as the technological singularity. Not only will humanity's thought processes be dwarfed by artificial superintelligence, but humanity's convictions on art and morality will be challenged in the name of optimization and efficiency. This will essentially make humans useless, but significant resources will still be spent on keeping us all alive, while humans contribute immense waste and pollution. This would be the point where the T-100s arrive and begin eradicating humanity. Many point out that an uncontrolled technological ascent has no reason to view humans as helpful or friendly. It wouldn't be actively malicious, but it would sometimes compete for the same resources, and would not be interested in furthering humanity's goals. Perhaps the best way of aligning our goals would be transhumanism, the concept of blending human physiology with bioengineering developments to create superhumans.

Futurologists have attempted to predict exactly when the singularity will occur. The median seems to be 2045, but the consensus is that it would definitely happen by 2100. This is based on the rate at which computational capacities compound, heavily influenced by Moore's law. Several things could slow down the rate of growth, such as software bottlenecks. Software has different development timescales than hardware, and even a superintelligent machine that is writing code would be playing catch-up to more advanced hardware. Also, according to other experts, eventually hardware reaches an upper threshold in computational speed--a threshold that would be reached in due time even with humans building machines.

Even if slowed, the singularity appears inevitable. For it to not occur, the Singularity University believes that intelligence augmentation needs to fail on six different impeding innovations: bioengineering, genetic engineering, nootropic drugs, AI assistants, brain-computer interfaces, and brain uploading. Even the success of one of these will lead to a singularity. It's so unavoidable, that Stephen Hawking believes superintelligence needs to be planned for today, because it could lead to lead to "technology outsmarting financial markets, out-inventing human researchers, out-manipulating human leaders, and developing weapons we cannot even understand."

Should we be worried about becoming obsolete ourselves, or is the concern of a technological singularity the stuff of science fiction?