Reading as a god

Chapter 222 Layout Artificial Intelligence
Women are emotional animals~
Zhang Shan fooled Gu Youyou into agreeing to develop this kind of website with some rhetoric.

But Zhang Shan doesn't really care about developing this website.

After all, if you only care about websites, it is not difficult to develop a hundred with Zhang Shan's resources now.

It's just that since Gu Youyou can be easily coaxed to do one chore, it must be easy to coax her to do other things.

Keke, this is not right.

Mainly as a good product manager.

What Zhang Shan is planning now is not just a simple website for pushing books!
At this time, his eyes have already been set elsewhere - artificial intelligence!
Artificial intelligence (English: artificial intelligence, often abbreviated to AI).

Also known as omniscience and machine intelligence, it refers to the intelligence exhibited by machines made by humans.

Generally artificial intelligence refers to the technology of presenting human intelligence through ordinary computer programs.

The definition of artificial intelligence in general textbooks is "the research and design of intelligent agents". Intelligent agents refer to a system that can observe the surrounding environment and take actions to achieve goals.

John McCarthy defined it in 1955 as "the science and engineering of making intelligent machines".

Andreas Kaplan and Michael Haenlein define artificial intelligence as "a system that correctly interprets external data, learns from this data, and uses this knowledge to achieve specific Ability to achieve goals and tasks".

Research on artificial intelligence has a long history.

It is different from the current emphasis on methods based on cybernetics and neural networks.

At that time, people have been working to symbolize artificial intelligence.

When digital computers were developed in the 20s, researchers began to explore whether human intelligence could be reduced to symbol processing.

Research has focused on Carnegie Mellon, Stanford, and MIT.

Of course, the research on AI in these schools has its own independent research style.

Researchers in the 60s and 70s were convinced that symbolic methods could eventually succeed in creating machines with strong artificial intelligence, and this was their goal.

These people try to break the situation from the following four angles:

Cognitive simulations: Economists Herbert Simon and Alan Newell studied human problem-solving abilities and attempted to formalize them, while they laid the groundwork for fundamental principles of artificial intelligence, such as cognitive science, operations research, and Business science.Their research team used the results of psychology experiments to develop programs that simulate human problem-solving methods.This method has been followed by Carnegie Mellon University and developed to its peak in Soar in the 80s.

Logic-based: Unlike Alan Newell and Herbert Simon, John McCarthy believed that machines need not simulate human thought, but should try to find the essence of abstract reasoning and problem-solving, whether or not people use the same algorithms.His lab at Stanford University works on using formal logic to solve a variety of problems, including knowledge representation, intelligent planning, and machine learning.Also working on the Logical Method was the University of Edinburgh, which led to the development of the programming language Prolog and the science of logic programming elsewhere in Europe.

"Against logic": Stanford researchers (such as Marvin Minsky and Seymour Papert) discover that solving hard problems in computer vision and natural language processing requires specialized solutions: they argue that there are no simple and general principles (like logic) capable of all intelligent behavior.Roger Schank described their "anti-logic" approach as "scruffy".Common sense knowledge bases (such as Douglas Leinart's Cyc) are examples of "scruffy" AIs because they have to manually code complex concepts one at a time.

Knowledge-based: Around 1970, large-capacity memory computers appeared, and researchers began to construct knowledge into application software in three ways.This "knowledge revolution" led to the development and planning of expert systems, the first successful form of artificial intelligence software. The "knowledge revolution" also made people realize that many simple artificial intelligence software may require a lot of knowledge.

……

Zhang Shan knew that early AI researchers directly imitated humans for step-by-step reasoning, just like how humans think when playing board games or logical reasoning.

By the 1980s and 1990s, AI research had also developed very successful methods for dealing with uncertain or incomplete information, using concepts from probability and economics.

For difficult problems, a large amount of computing resources may be required, that is, a "possible combinatorial explosion" occurs: when the problem exceeds a certain scale, the computer will require astronomical amounts of memory or computing time.Finding more efficient algorithms is a priority AI research project.

The human problem-solving mode usually uses the quickest and intuitive judgment, rather than conscious, step-by-step derivation. Early artificial intelligence research usually uses a step-by-step derivation method.

Artificial intelligence research has made progress in this "sub-representational" problem-solving approach: Embodying agent research emphasizes the importance of perceptual motion.Neural network research attempts to reproduce this skill by mimicking the structure of the human and animal brains.

Speaking of which, Zhang Shan has maintained a supportive attitude towards artificial intelligence for a long time, and has invested in Lin Ning's company before.

But speaking of it, it was nothing more than a small fight. At that time, it seemed that a large investment could only be said to be a drizzle for Zhang Shan, who is now rich and powerful.

Although the investment at that time was not much, Zhang Shan never stopped paying attention to the field of artificial intelligence.

Therefore, he also made up a lot of content related to artificial intelligence.

Artificial intelligence is actually a very broad topic!
After all, the research of artificial intelligence is highly technical and professional, and each branch field is in-depth and different, so it involves a wide range.

The study of artificial intelligence can be divided into several technical problems.

Its subfields are mainly focused on solving specific problems, one of which is how to use various tools to complete specific applications.

The core issues of AI include the ability to construct reasoning, knowledge, planning, learning, communication, perception, moving objects, using tools and manipulating machines that are similar to or even superior to humans.

Artificial intelligence is still a long-term goal in this field.

One of the core research problems in the current field of artificial intelligence is knowledge representation.

The goal of the so-called knowledge representation is to allow the machine to store the corresponding knowledge, and to be able to reason and deduce new knowledge according to certain rules.

One of the more popular definitions of artificial intelligence, and an earlier definition of the field, was proposed by John McCarthy of the Massachusetts Institute of Technology at the 1956 Dartmouth Conference:

Artificial intelligence is to make the behavior of the machine look like the intelligent behavior shown by humans.

Another definition refers to artificial intelligence as the intelligence exhibited by artificial machines.

Generally speaking, most of the current definitions of artificial intelligence can be divided into four categories:
That is, the machine "thinks like a human", "acts like a human", "thinks rationally" and "acts rationally".

Here "action" should be broadly understood as taking action, or making a decision to act, rather than physical action.

Regarding the current development of artificial intelligence, many of them are weak artificial intelligence!

The view of weak artificial intelligence believes that it is "impossible" to create intelligent machines that can "really" reason and solve problems. These machines only "look" like intelligent, but they do not really possess intelligence and will not have autonomous consciousness. .

Speaking of which, the research of artificial intelligence was in a state of stagnation for a time, and it was not until the neural network had a strong computing power to simulate it that it began to change and greatly advanced.However, artificial intelligence researchers do not necessarily agree with weak artificial intelligence, nor do they necessarily care about or understand the content and differences between strong artificial intelligence and weak artificial intelligence, and debate the definition endlessly.

As far as the current artificial intelligence research field is concerned, researchers have created a large number of machines that "look" like intelligence, and have achieved quite fruitful theoretical and substantive results:

For example, in 2009, the Eureqa computer program developed by Cornell University professor Hod Lipson and his doctoral student Michael Schmidt, as long as some data is given, the computer program can deduce the Newtonian mechanics that Newton spent years of research in just a few tens of hours. formula.

This is equivalent to rediscovering the Newtonian mechanics formula by yourself in only a few tens of hours.

This computer program can undoubtedly be used to study scientific problems in many other fields as well.

These so-called weak artificial intelligences have made great progress under the development of neural networks, but there is no clear conclusion on how to integrate them into strong artificial intelligence.

At present, weak artificial intelligence has achieved preliminary results, and even some unilateral abilities in image recognition, language analysis, board games, etc. have reached the level beyond human beings, and the versatility of artificial intelligence means that the ones that can solve the above problems are the same The AI ​​program can directly use the existing AI to complete the task without redeveloping the algorithm, which is the same as the processing ability of human beings, but it takes time to study to achieve integrated strong artificial intelligence with thinking ability. The more popular methods include statistical methods, computing Intelligence and AI in the traditional sense.At present, there are a large number of tools that apply artificial intelligence, including search and mathematical optimization, and logical deduction.Algorithms based on bionics, cognitive psychology, and probability theory and economics are also being gradually explored.

If there is weak artificial intelligence, there will naturally be strong artificial intelligence!

The term "Strong Artificial Intelligence" was originally coined by John Rogers Searle for computers and other information processing machines, and is defined as:
"The view of strong artificial intelligence holds that the computer is not only a tool for studying human thinking; on the contrary, as long as the appropriate program is run, the computer itself has thinking." (J Searle in Minds Brains and Programs. The Behavioral and Brain Sciences , vol. 3, 1980)

The debate about strong artificial intelligence is different from the broader debate about monism and dualism.

The crux of the argument is: If the only working principle of a machine is to convert coded data, does the machine have a mind?
Searle thinks this is impossible.

He gave an example of a Chinese room to illustrate that if the machine only converts data, and the data itself is a coded representation of certain things, then the premise of not understanding the correspondence between this code and the actual thing Without it, it is impossible for a machine to have any understanding of the data it processes.

Based on this argument, Searle believes that even if a machine passes the Turing test, it does not necessarily mean that the machine really has self-thinking and free consciousness like a human.

Zhang Shan knows that the so-called Turing test is a thought experiment proposed by British computer scientist Turing in 1950.

The purpose of the experiment is to test whether machines can exhibit human-equivalent or indistinguishable intelligence.

Conversation channels in the test were limited to the use of text, such as computer keyboards and screens, so the results did not depend on the computer's ability to convert words into audio.

……

It should be pointed out that strong artificial intelligence is not completely opposed to weak artificial intelligence.

That is, even if strong AI is possible, weak AI still makes sense.

At least, the things that computers can do today, such as arithmetic operations, were considered to require intelligence more than 100 years ago.

Moreover, even if strong artificial intelligence is proven to be possible, it does not mean that strong artificial intelligence will definitely be developed.

The development of artificial intelligence is not only technically difficult!
There are also many obstacles in the society:

Many media have predicted that some occupations will be replaced by robots.

Japan's Nomura General Research Institute also conducted a joint investigation with researchers from the University of Oxford in the United Kingdom and pointed out that 10% of occupations (20 occupations) in Japan may be replaced by machinery and artificial intelligence in 49 to 235 years, which will directly affect the development of the world. 2500 million people.

Examples: supermarket clerks, general clerks, taxi drivers, tollbooth operators and cashiers, marketers, customer service personnel, manufacturing workers, financial intermediaries and analysts, journalists, telephone company clerks, anesthesiologists, soldiers and security guards High-paying intellectual occupations such as lawyers, doctors, software developers and traders, stock traders, etc. will be the first to be hit.

Some people have even worried that artificial intelligence may lead to the third world war, because the previous two industrial revolutions caused two wars~
While the cause of war was not these innovations themselves, it was the mismanagement of the impact of the invention on the way many people in society lived.

Zhang Shan feels that there is nothing wrong with it. New technology creates new jobs in society and replaces old jobs, creating new losers and winners~
Just pay attention to the means and methods, and pay attention to the step-by-step implementation process.

Zhang Shan felt that all these possible troubles could be avoided!

Not only are there employment issues, there are also concerns about the safety of artificial intelligence.
Stephen Hawking, Bill Gates, Elon Musk, Jaan Tallinn, and Nick Bostrom have all publicly expressed concern about the future of artificial intelligence technology.

Zhang Shan knows that these people are worried that if artificial intelligence surpasses human intelligence in many aspects, continuously updates, self-improves, and then gains control and management, whether human beings have enough ability to stop the "arms race" in the field of artificial intelligence in time, whether Maintain supreme control?

After all, the existing facts are: machines often go out of control and cause casualties. Whether such a situation will appear on a larger scale, history obviously cannot give a reliable and optimistic answer.

Elon Musk called artificial intelligence an act of "summoning the devil" at a symposium marking the centennial of the MIT aerospace division.

(End of this chapter)