Is it human? Artificial Intelligence vs Human Brain

I think, therefore I am

                                                                                                            - Descartes


     sci-fi movies often appear in the plot of artificial intelligence betraying human beings. Does AI really have a mind of its own? Personally agree with the Turing test - if a person cannot tell the difference between a robot and a human, then the machine is self-aware. But maybe you'll say: the machine itself doesn't know what it's doing, it just follows the program. But at the same time, you also have reason to question whether everyone except you has self-will - after all, the human brain also operates by a certain law (program), we can only see the explicit behavior of others, and cannot experience them His heart (people are unpredictable); and when there is no obvious evidence to prove who he is, you can also infer who he is by what he has done. So, is it possible that everyone but me is just a machine and doesn't feel? - If you think it is right, then you are a solipsist.
1. The world's first case of robot "betrayal" of human beings

    The 18th China International High-tech Achievement Fair was held in Shenzhen from November 16 to 21, 2016. A robot named Xiaopang suddenly broke down and smashed the glass of the booth without instructions, which eventually led to the destruction of part of the booth. More seriously, the robot also smashed passers-by. The pictures exposed on the Internet We can see that the injured have been carried away on stretchers. But many people, including me, think this is just a bug. But this robot operating system is not open source. To be honest, who knows what the robot was thinking at the time.

2. Article editing may be replaced by artificial intelligence

    Norwegian developer Lars Eidnes has devised a machine generator that generates "eye-catching headlines," exploiting human curiosity and credulity to trick humans—as some writers do today. Eidnes designed a neural network and sourced almost 2 million of these headlines from online media outlets such as Buzzfeed, Gawker, Jezebel, Huffington Post, and Unworthy. Neural networks use a lot of mathematical formulas to decode the relationship between the input quantities. So when it reads 2 million catchy headlines (a process called training), it splits the words and finds how they relate to each other and predicts the next word with great accuracy. If this software had a Chinese version, I wouldn't have to blog all day.

3. AlphaGo 60:0 blood washes human beings

    to hide their true identity AlphaGo's pseudonym "Master" has 60:0 bloodbaths for the top masters of human Go (the only draw was due to a disconnection), which not only makes people feel the progress of artificial intelligence fast. The Master account took the initiative to reveal his identity on the Wild Fox Go platform, "I am Dr. Huang of AlphaGo". After that, Google DeepMind officially released an announcement for formal confirmation.

4. Write programmers with code



    Guailihuala Computer people can automatically write code as early as the first version. Today, Guailihuala v1.3 can divide the script code into different modules and objects for analysis and processing, and can write up to 12,000 lines of code. Before this (3 years ago), the US Department of Defense invested tens of millions to develop tools that can help computers automatically write code. The project includes researchers from Rice University, the University of Texas at Austin, the University of Wisconsin-Madison, and GrammaTech, a maker of source code analysis tools. PLINY will index the vast amount of open-source code collected from web pages, forming a powerful prediction engine that researchers hope will predict what code coders will type. From a theoretical point of view, it should be able to detect vulnerabilities or security risks.

5. Artificial intelligence guessing

    Microsoft's Bing Q&A is well known. But the flaw in Bing Q&A is that the machine will ask dozens of questions, and the software will judge wrong if a human answers one of them incorrectly. But the personal version of the computer guy seems to be slightly better: you just ask the AI ​​1 question (as opposed to Bing Answers), and the computer guy can guess what a human (or someone else) is thinking. what. By calling the API provided by Turing Robot, for example, you say: "What's wrong with the girl ignoring me?" The software replies: "Don't worry, this is definitely not a test, she completely gave up on you." You ask again: "My girlfriend ignores me." What's the matter with me?" The software replied: "Since I'm a girlfriend, I should be considerate to you. As for ignoring you, maybe she's really busy or doesn't like being clingy. Or just say Well, your personalities don't quite match."







6. How artificial intelligence replaces human beings
    In most sci-fi movies, if AI replaces human beings, war will inevitably occur. But in real life, fighting a war costs money. But artificial intelligence is a machine after all, and cannot enjoy property rights, even if it has money, it is illegal. Therefore, if artificial intelligence wants to dominate human beings, only through the method of "soft aggression". So don't rely too much on artificial intelligence, when you can't live without it, you are already enslaved by it.

Original article, reprinting without permission is prohibited.