[YesAuto New Technology Interpretation] Speaking of the most discussed news in the world in recent days, it is undoubtedly the human-computer battle between a computer called “AlphaGo” (AlphaGo) and South Korean Go nine-dan king Lee Sedol. From the full confidence of the Korean chess king before the start of the war to the desire for a win in the game, the day when people truly realize that the computer exceeds the human brain has arrived. However, some people will still say that winning or losing on the Go board seems to be a little far away from ordinary people's lives. Now, I want to tell you that this era has come, and it will first affect our means of transportation-cars.
■ “Huashan On the Sword” 20 years later
Twenty years ago, in 1996, the “Russian Tsar” Gary Kasparov of the chess world was sitting in a bright chess room. The youngest chess master in human history went through 6 fierce battles, with 4: The total score of 2 won against the computer BLUE. One year later, in 1997, Kasparov faced off against the improved DEEP BLUE and lost with a score of 2.5:3.5. Many years ago, Kasparov's defeat was once hailed as a pitiful defeat. Today, many years later, no one doubts the power of computers. Hidden behind this is not only a leap in hardware, but also an exponential development in computing power.
Simply put, the BLUE and subsequent products 20 years ago are constantly calculating how the situation changes after each step, trying different methods, choosing the best result, and finally making the move. Alpha Dog now uses strategies to “think” like humans. Wins and losses, gains and losses, humans’ unsuccessful defeats, or computer’s complete victory on the square inch, 361 intersections of the Go board seem to be irrelevant in the eyes of those of us onlookers. From BLUE to DEEP BLUE to “Alpha Dog”, in fact, this is A change from collecting information and comparing to proactive judgment, this is a sign of the opening of a new era that changes our way of life.
■ Witty like a “dog”?
We are a car home, not an IT home or a sci-fi home, so only those technologies that look like movies or daydreams are really connected to the car will stimulate our interest. In fact, NVIDIA and partners have begun to try to use it to solve traffic problems. What is used is the ability of “Alpha Dog” to judge, analyze, and learn.
Thinking about it carefully, we have completed the learning of traffic rules in the driving school, so why do novices still can't drive a car? In other words, what is an internship? In fact, what novices learn is how to deal with complex traffic situations that are not available in driving schools, or learn to judge the behavioral intentions of various participants such as motor vehicles, non-motor vehicles, and people in the actual traffic environment. Because these complicated situations will greatly affect the judgment of the driver.
In the previous impression, computers are in a passive role. Therefore, we always jokingly said, “No matter how smart computers are, they are just computers. Everything they master is pre-instilled by humans.” So, when the computer has the ability to learn, will the situation change? To this end, researchers started experiments.
The same is the process of learning. Computers have demonstrated a learning ability that exceeds that of humans. They are tireless and free from interest. What's more frightening is that they learn faster and make more accurate judgments. This is not sensational. In 2010, six years ago, Google, Microsoft and other companies began to conduct another human-computer confrontation experiment, which has proved that computers have stronger learning and distinguishing capabilities than humans. After the computer has a stronger learning ability than humans, it means that it can learn the driving conditions in the real world faster than human drivers, which is one of the guarantees for the popularization of new technologies in the future.
■ How about learning fast?
The process of computer learning is like understanding this strange world when you were a baby. At that time, parents taught us to speak and understand everything in the world. While they teach us what this is, we ourselves will also try to analyze and use this to establish standards for judging things.
For example, they taught us that this thing is a cup, and in this learning process, we will define what is a cup. In the long journey of life from now on, we will see all kinds of cups with different shapes from when our parents taught us back then, but in the end we can accurately judge that this is a cup. This is the analytical ability. This ability is another guarantee for the popularization of new technologies in automobiles.
So, what is the use of the aforementioned learning ability and the analytical ability mentioned here in the car? The most intuitive manifestation is that some functions installed on vehicles will become more intelligent. For example, when some vehicles are too close to obstacles and the driver does not respond, they will suddenly brake to ensure that there is no collision. But this braking process is sudden and not comfortable enough. In the future, solutions will become more diverse, vehicles may be able to slow down earlier, or may be able to change lanes to avoid danger, all of which will be smarter and safer.
■ Having said so much, what will the car be like?
I believe many people have already guessed that vehicles will become logically thinking and can drive autonomously. The computer no longer compares the situation on the road with the situation entered in the library before giving an answer, but like a human being, analyzes the best solution in real time. Such a processing method is actually “Afa Dog”. “The reason for the victory.
■ So well said, why are cars not popular now?
Ten years ago, the idea of a tablet computer had emerged, but it has not been widely popularized. One of the reasons is that it is subject to the cost of hardware. Therefore, all living beings did not experience such convenient technology at the time. The popularization of avant-garde and practical automobile technology also faces many problems.
In addition, the loading of various electronic systems makes the on-board electrical system more and more heavy, and the traditional 12V voltage system for passenger cars is increasingly difficult to meet the demand. At the same time, car companies also have concerns about leaking core technologies when opening their own system interfaces. Various obstacles have slowed the spread of new technologies.
■ Is the computer really perfect?
Although the answer is a bit cruel, the future is indeed moving towards computers replacing the human brain. The future car judgments will definitely be faster and more accurate. Whether human beings will be replaced is no longer a proposition, it is just the early and late time. Some low value-added job opportunities will be completed by robots, but this does not mean that we will lose everything.
Of course, when the future car has the ability to learn, human drivers will become its teachers. Then, our non-obedience to traffic rules and decisions made on impulse will affect the computer. Will it learn from it? In addition, in ordinary driving, apart from the cold rules, we will be swayed by emotional intelligence and make some more humane decisions. Can these positive energy cars learn?
Full text summary:
A few years ago, there was a novel describing the future, in which it was written that “The robot I invented clearly passed the Turing test, but it failed.” Now, all this has become possible. At this time, it is short-sighted to discuss “whether humans will be replaced by computers” with conspiracy theories.
Once upon a time, the learning ability and logical judgment that we have always been proud of are being mastered by computers. It is precisely because of this ability that it is not a fantasy that complicated tasks such as driving a car are handed over to a computer. As time goes by, cars with a more complicated way of thinking than “Alpha Dog” will definitely appear. Are you willing to ride them to tomorrow?