To put that number in context: About 23,000 American men and women will die of leukemia in 2012, according to the National Cancer Institute.
But while the technology Google is working on isn't the cure for cancer, it's pretty impressive.
What is it?
Self-driving cars.
According to the U.S. census, 33,808 people died in car crashes in the U.S. in 2009 — 37,423 in 2008. According to the World Health Organization, 1.2 million people die in car crashes around the world each year.
A source familiar with Google's self-driving car program tells us that one of its primary goals is to eliminate the ~99 percent of those deaths that are caused by "human error."
That's obviously a lofty goal.
It's certainly admirable … but is it achievable?
There are some data points to consider before answering that question.
- Google's cars are much better drivers than you. Self-driving cars can "see" in 360 degrees. They do not get tired, drunk, old, or enraged. They do not have teenage hormones. They know exactly how to stay out of other drivers' blind spots. They can anticipate and adjust due to variables impercitible to humans.
- Google's self-driving car program is much further along than you might guess in terms of consumer-friendliness. When you hear the phrase "self-driving car," you might imagine a car interior stuffed with wires. Don't. We understand that the interior gadgetry associated with Google's self-driving cars is minimal these days. Imagine a few extra buttons on the steering wheel no more complicated than "cruise control," and a user-interface that's as consumer-friendly as an Android tablet.
- It's much easier for Google's self-driving cars to operate on the highway than on the street. On highways, the traffic is all moving in the same direction at generally the same speed in neatly divided lanes. On surface streets, cars can come from anywhere and go in any direction. The data is messier. Because of this, our guess is that auto-drive will be a mode that is only available for a driver to enable on highways in particular conditions in certain lanes. Again, imagine "self-driving mode" as something similar to "cruise control." This puts a limit on the number of lives self-driving cars can save. In 2011, about 7,000 of the approximately 35,000 car accident deaths in the U.S. took place at intersections. Still: Google's cars would almost immediately cease the more than 3,000 deaths each year due to distracted highway driving.
- Google is still facing a couple technological challenges. Its cars still have trouble in the snow because their laser sensors can't "see" the road to tell if it's the same as the one inside their computer brains.
So … is Google's goal achievable?
Mostly.
And when you're talking about 35,000 lives in the U.S. each year, and 1.2 million worldwide, "mostly" is pretty incredible.