On Predicting the Future

In 1999, well-known futurist and popularizer of the "singularity" Ray Kurzweil published a book called The Age of Spiritual Machines. In it he made some claims about how the world would transform in the coming century. Here are some random samples of what he thought the world would look like in 2019:

"People routinely use three-dimensional displays built into their glasses... These "direct eye" displays create highly realistic, virtual visual environments overlaying the "real" environment. This display technology projects images directly onto the human retina, exceeds the resolution of human vision, and is widely used regardless of visual impairment."

"Keyboards are rare, although they still exist. Most interaction with computing is through gestures using hands, fingers, and facial expressions and through two-way natural-language spoken communication."

"A new computer-controlled optical-imaging technology has replaced most lenses with tiny devices that can detect light waves from any angle. These pinhead-sized cameras are everywhere."

"People read documents either on the hand-held displays or, more commonly, from text that is projected into the ever-present virtual environment using the ubiquitous direct-eye displays."

"Deaf persons routinely read what other people are saying through the deaf person's lens displays... Generally, disabilities such as blindness, deafness, and paraplegia are not noticeable and are not regarded as significant."

"The all-enveloping tactile environment is now widely available and fully convincing. Its resolution equals or exceeds that of human touch and can simulate (and stimulate) all of the facets of the tactile sense, including the sensing of pressure, temperature, textures, and moistness."

"Automated driving systems have been found to be highly reliable and have now been installed in nearly all roads."

Now, for brevity's sake, I'll stop here. In the book, Kurzweil goes on to give his predictions for 2029 and 2099. No doubt these predictions will seem equally outlandish when those dates arrive. My aim here isn't to pick on Kurzweil--he's a brilliant guy and has made his fair share of predictions that have turned out to be true. He's certainly not the first to get his predictions about the future wrong: history is full of examples of science futurism that has turned out to be science fiction. Instead I'm interested in why it's so hard for people--even smart people--to predict the future. Why do we know so much more about the past than we know about the future? That's a good question and I'm not sure I have a satisfying answer, but a few come to mind.

One obvious reason is compound error. As soon as you predict event A, all your predictions from what happen after A rely on your assumption that A has happened. If you miscalculated and were slightly off-base with your prediction for event A, your predictions for events B, C, and D will be even more off-base than A. Eventually this chain of predictions becomes wildly inaccurate. Compounding errors like this have crashed rockets--NASA's Mariner I, crashed due to a floating-point arithmetic error that got out of hand. Wars have been fought and couples have divorced over mistaken predictions about future behavior that spiralled out of control. And they have led reasonably smart people into making outlandish predictions. For all the things that humans are good at, predicting the future is not one of them.

Another has to due with the nature of computation. In computational theory there's a concept known as the undecidability. In laymen's terms, this is the idea that there are certain "undecidable" problems that can't be determined in advance by an algorithm. The only way to "decide" what will happen in such problem is to wait and observe what happens. The future is undecidable, in the sense that there is no way for humans to forecast an arbitrary number of years into the future. We live in a huge world made of tiny sub-atomic pieces, each of which lies in an unmeasurable state somewhere between a wave and a particle, and if we can't even be sure about the current state of every atom in the universe, have can we be sure about the state of every atom in the universe 5 seconds from now. This is a huge problem that humans have only just begun to explore seriously.

The third problem is psychological and it's related to the first two. Our ideas about the future are always inseparable from our ideas of the present. We spend so much time in our lives playing catch-up and trying to figure out what's going on in the present, that by the time we begin to have ideas about the future they are already entrenched in the past. This is why so many science-fiction books have aged poorly--they are reflections of their own time and place, rather than the future.

Maybe someday we'll be able to accurately forecast one year--or ten years--into the future, but that doesn't make it any easier to us to predict what will happen one hundred years in the future. The exponential nature of compounding errors make this problem computationally intractable. No matter how big our brains get, and no matter how much computational power humans can get their hands on--the vast size and timescale of the universe is guarantee that it will always outsmart us. For all our self-congratulation about scientific progress and the power of the human race, we will always have to contend with the fact that we live in a world that is mostly unknown and unknowable.

April 15, 2019