FANTASY：In which way do you think the world will end?

时间：2016-02-19 01:38:31 阅读：247 评论：0 收藏：0 [点我收藏+]

标签：

In which way do you think the world will end?

　　The moment you are reading my essay, you are somehow connected with it. And I fancy that you have never realized this kind of subtle but amazing connection. More specifically, when you are reading, you receive the words’ images my essay has sent you and, in turn to say, the images’ receiver is you. And, while you are clicking the mouse and turn to the next page, you become the sender of the command and my essay is now the receiver of it. What mentioned above is a simple example that interaction, which transmits information, should have one sender and one receiver, as though a directed edge must have an initial vertex and a terminal one in order to biuld a directed graph.

　　WLOG, let’s assume that everything, all the receivers and senders in this world, is represented by vertexes consisting a huge graph and we use directed edges to represent their interactions and to connect all these vertexes. And I prefer to refer this huge graph as our world. On the graph, there are countless vertexes and edges, so when the essay question asked me to assume a way this world ends, I felt it quite impossible to destroy such a hulk for I have to delete the whole graph, and before deleting it, I must clear all the vertexes which, however, may include the ones representing trillions of atoms and the whole Sombrero Galaxy, the ungodly technological advances human made and my essay. Finding a method to turn all these things into void is far beyond my imagination.

However, ending a world is actually not that hard and can be as easy as every day I shut down the program which is running on my laptop.

Think in another way; the world mentioned here is no longer a stable graph, but can be a program that runs on the huge graph and the huge graph is part of it; the world here is not exactly the one that often appears in literature and slogans, but works as an intermediary between the senders and receivers. This idea suggests that, e.g., when you tell a rumor to your friend, you actually share it first to the world and then the world repeats the same story to your friend. It become a program that working for the vertexes, including the one representing you through making highly-accurate responses to and collecting detailed feedback from vertexes .

From this new idea, now the huge graph may be a little redundant. For every vertex now connects with their intermediary, the world, directly, it is no use of the graph. Besides this, it’s also ordinary to get an edifying conclusion:the world has somehow separated everyone (all the senders, receivers, or vertexes) into countless cells. Say, I am in my cell and I was provided all the data or information (like today’s weather, where is my mom and …) that I ought to have from the world. After processing, I transmit my response data to the world. And world gives me new data, I …, that cycles. It is worth noting that the only place I receive the data from is the world. Interestingly, in turn, the whole procedures somehow prove a sentence that said by Arthur Schopenhauer: Everyone takes the limits of his own vision for the limits of the world.

Take Alice’s experience as an example. Every day, Alice enjoys sunsets and chatting with her net friend, Bob, via Wechat, an app for social media. Bob is a nice guy. He is friendly and always gives response to Alice’s messages. However, Alice one day starts to wonder whether Bob is a real person that exists, but she never has a chance to confirm her doubt without Bob’s permission and no evidence to prove Bob exists. If I were Alice, I might even wonder whether the spectacular sunsets every day I see is not real either. For the world is acting like Bob, the world uses the same tricks as Bob plays on Alice by using internet to separate Alice and the world that Bob stays.

This understanding reduces complexity of realizing the program of the world. Say, without this thought, the program may looks like a great online game that supports PVP. However, now it looks more like a PC games which have one player. The program has two main parts: simulating the behavior of a vertex (sender/receiver) and generating its response data.

The first part simulates the behaviors, which is how a vertex processes the information. All the behaviors can be classified into two categories: behaviors to the vertex itself (like sleeping) and behaviors to the world (like chatting). As we know, the most complicated and sophisticated creature is human. So the nub of the first part is how to simulate human’s brain and its countless neurons. (Why choose to simulate a human’s world? Why not a tree’s? )

The second part is to generate the response data that are provided to the first part and to collect the feedback information. The hard-core part is how to make sure that the response data is logical and in accordance with the history and the laws of development. My suggestion to solving this problem is to add the number of simulators and to learn from the collected feedback data (build a training set), although it may add burden to the program when running.

However, the implementation of the both parts is beyond our current abilities. It is on the initial stage of artificial intelligence that we still have a long way to go before figuring out an approach for creating a robot as smart as human. It is at the prime time of the Big Data Time that we first turn to look at the big-size data and we still don’t do so well in dealing with big data which the second part needs. So, are we, human, the only possible or fittest author of this great program?

The answer is “No.”. Future robots with high artificial intelligence will be a better answer. According to the Singularity Theory that the growth of technological advancement is exponential, scientists will keep coming up with better algorithms in the field of big data and more powerful and cleverer robots that surpass humans. When the moment comes (in 50 years or shorter), which our ability in designing AI progresses to a certain level, the program will no longer be an assumption in the essay.

I highly doubt whether human still exists when the day that the program is realized. A book once said that robot with high artificial intelligence will be the last invention born from human’s hands. I agree with this sentence fully. You can understand it in two ways: robot with high AI will be so automated and clever to invent all other things, or humans will die out once they manage to invent it. I tend to agree the second one. In 50 years, robots in the labs will have great logics, decision-making and prediction-making abilities. However, it will still be difficult for robots to understand moral reasons. Lack of these important moral reasons to constrain their behaviors, robots will be extremely dangerous to human. For robots, their life goal is what their creators first set for them and what is carved on their chips forever and they can sacrifice anything or take any action to realize the life goal including eliminating human. Of course, robots are not spiteful to human. When you cut your hair, you are not spiteful to them either, but you simply cut them. Unluckily, robot is much clever than human and doesn’t suffer from some physical and biological limits which human has.

If human has disappeared, why robot will want to simulate an antiquated and backward world (say, now you are reading my essay)? According to the Darwinian Theory of evolution, when robots replaced humans, robots become the winner in the natural selection and should be seen as a new species or the evolutional version of human, from biological version updated to mechanical version. With the continuous progression for robots in accordance with the Singularity Theory, it will not be surprised that robots formed its own society and culture. Once they have their own culture, history will be important to them. For there are no differences between human and robot in nature, they are all the vectors of artificial intelligence. Future robots need to know and learn experience from how human, the old version, disappears and how their ancestors eliminate humans. Historian robots won’t simply satisfy in text and videos that record it, but want to touch this piece of crazy history through building a virtual world. This year 2016, to 2035 the predicted year that robot which is as smart as human appears, is not a long period, even to the year human may disappear (like 2055?). Now would be a great moment for future robots to start simulate the world, for now we are facing the big challenge of artificial intelligence and it is the best ever moment for us to reach the future.

When all the information that the future robots need has been collected, the program will be shut down and the world will end.