should make decisions based upon the probability of all outcomes, including the best- and worst-case scenarios, and be willing to take big risks when you are fully prepared for all eventualities.
It’s also important to remember that good decisions, based on an accurate analysis of the risks involved, can still lead to bad outcomes. That’s because risk is still involved. Here is a simple example: soon after I got out of school I was offered a job I wasn’t sure was a great fit for me. After several days carefully considering the opportunity, I decided to turn it down, assuming that soon thereafter I would be able to find another job that was a better match. Unfortunately, the economy turned south quite quickly and I spent months looking for another job. I kicked myself for not taking that position, which started to look more and more appealing. I had made a good decision, based upon all the information I had at the time, but in the short run it wasn’t a great outcome.
As in this situation, under most circumstances you have to make decisions with incomplete information. That is, you have to make a choice and take action in the face of considerable uncertainty surrounding each option. So, how do you fill in the gaps of your knowledge? I suggest looking to “Stanley” for inspiration. The inner workings of Stanley, the autonomous vehicle designed and built by Stanford’s Artificial Intelligence Lab and Volkswagen Electronics Research Laboratory for the DARPA Grand Challenge, offer clues to decision making with incomplete information. DARPA, the Defense Advanced Research Projects Agency, is a government agency charged with the development of cutting-edge technology for the military. In the DARPA competition, driverless cars must navigate a 212-kilometer off-road race. Each must pass through three narrow tunnels, make more than a hundred sharp turns, and navigate mountain passes with steep cliffs on each side. Despite very low odds, Stanford’s car won the race, due in large part to its ability to make quick decisions with incomplete information.
Stanley had a lot of powerful technology on board, including three-dimensional maps of the terrain, GPS, gyroscopes, accelerometers, video cameras, and sensors on the wheels. The on-board software analyzed and interpreted all incoming data and controlled the vehicle’s speed and direction. But the key to Stanley’s victory was its superior skill at making decisions with incomplete information. The designers accomplished this by building in the ability to learn the way humans do. They created a database of human decisions that the car drew upon when making judgments about what to do. This data was incorporated into a learning program tied to the car’s control systems, and greatly reduced errors in judgment.
This story highlights the fact that learning from others can significantly reduce your failure rate. You don’t have to figure everything out yourself. Like Stanley, you should gather all the data you can from your environment, and then tap into the wisdom of those who have gone before you, in order to make the best possible choice. All you need to do is look around to see hundreds, if not thousands, of role models for every choice you plan to make.
If you do take a risk and happen to fail, remember that you personally are not a failure. The failure is external. This perspective will allow you to get up and try again and again. Your idea might have been poor, the timing might have been off, or you might not have had the necessary resources to succeed. As Jeff Hawkins says, “You are not your company. You are not your product. It is real easy to think you are and it is real easy to get wrapped up in it…. But if you fail, or even if you are successful, it is not you. Your company may fail, your product may fail, but you aren’t the failure.” Keep in mind that failure is a natural part of the learning process. If you aren’t failing sometimes, then you probably
It’s also important to remember that good decisions, based on an accurate analysis of the risks involved, can still lead to bad outcomes. That’s because risk is still involved. Here is a simple example: soon after I got out of school I was offered a job I wasn’t sure was a great fit for me. After several days carefully considering the opportunity, I decided to turn it down, assuming that soon thereafter I would be able to find another job that was a better match. Unfortunately, the economy turned south quite quickly and I spent months looking for another job. I kicked myself for not taking that position, which started to look more and more appealing. I had made a good decision, based upon all the information I had at the time, but in the short run it wasn’t a great outcome.
As in this situation, under most circumstances you have to make decisions with incomplete information. That is, you have to make a choice and take action in the face of considerable uncertainty surrounding each option. So, how do you fill in the gaps of your knowledge? I suggest looking to “Stanley” for inspiration. The inner workings of Stanley, the autonomous vehicle designed and built by Stanford’s Artificial Intelligence Lab and Volkswagen Electronics Research Laboratory for the DARPA Grand Challenge, offer clues to decision making with incomplete information. DARPA, the Defense Advanced Research Projects Agency, is a government agency charged with the development of cutting-edge technology for the military. In the DARPA competition, driverless cars must navigate a 212-kilometer off-road race. Each must pass through three narrow tunnels, make more than a hundred sharp turns, and navigate mountain passes with steep cliffs on each side. Despite very low odds, Stanford’s car won the race, due in large part to its ability to make quick decisions with incomplete information.
Stanley had a lot of powerful technology on board, including three-dimensional maps of the terrain, GPS, gyroscopes, accelerometers, video cameras, and sensors on the wheels. The on-board software analyzed and interpreted all incoming data and controlled the vehicle’s speed and direction. But the key to Stanley’s victory was its superior skill at making decisions with incomplete information. The designers accomplished this by building in the ability to learn the way humans do. They created a database of human decisions that the car drew upon when making judgments about what to do. This data was incorporated into a learning program tied to the car’s control systems, and greatly reduced errors in judgment.
This story highlights the fact that learning from others can significantly reduce your failure rate. You don’t have to figure everything out yourself. Like Stanley, you should gather all the data you can from your environment, and then tap into the wisdom of those who have gone before you, in order to make the best possible choice. All you need to do is look around to see hundreds, if not thousands, of role models for every choice you plan to make.
If you do take a risk and happen to fail, remember that you personally are not a failure. The failure is external. This perspective will allow you to get up and try again and again. Your idea might have been poor, the timing might have been off, or you might not have had the necessary resources to succeed. As Jeff Hawkins says, “You are not your company. You are not your product. It is real easy to think you are and it is real easy to get wrapped up in it…. But if you fail, or even if you are successful, it is not you. Your company may fail, your product may fail, but you aren’t the failure.” Keep in mind that failure is a natural part of the learning process. If you aren’t failing sometimes, then you probably
Similar Books
