Artificial Intelligence –AI Introduction


liu, tempo Date: 2021-07-08 10:25:49 From:ozmca.com
Views:54 Reply:0

Artificial intelligence (AI) has entered the vision of the general public, and we can see many AI-related products in our lives. Such as Siri, AI beauty, AI face change…


Although everyone listens a lot, most people don’t understand AI, and there are even some misunderstandings. This article will not involve any technical details to help everyone understand the nature of artificial intelligence.


What is artificial intelligence?

Many people have some misunderstandings about artificial intelligence:
1.The robot in the movie is a typical representative of artificial intelligence
2.Artificial intelligence seems to be omnipotent
3.Artificial intelligence will threaten the survival of mankind in the future
4….


The reason why you have many misunderstandings about artificial intelligence is mainly because you only see some people’s speech, but do not understand the basic principles of AI. This article will help you understand the basic principles of AI. The essence of things is often not what everyone said so complicated.


We use traditional software to compare with artificial intelligence. It is easier to understand with a frame of reference.

Traditional software vs artificial intelligence


Traditional software

Traditional software is the basic logic of “if-then”. Humans use their own experience to sum up some effective rules, and then let the computer run these rules automatically. Traditional software can never exceed the boundaries of human knowledge, because all rules are made by humans.


To put it simply: traditional software is “rule-based” and requires artificially setting conditions and telling the computer what to do after meeting this condition.

This kind of logic is very useful when dealing with some simple problems, because the rules are clear, the results are predictable, and the programmer is the god of software.


However, real life is full of various complicated problems, which are almost impossible to solve by formulating rules. For example, the effect of face recognition through rules will be very poor.

Artificial intelligence

Artificial intelligence has now developed many different branches, and the technical principles are also diverse. Only the hottest deep learning at the moment is introduced here.
The technical principles of deep learning are completely different from the logic of traditional software:
The machine sums up rules from a large amount of “specific” data, sums up some “specific knowledge”, and then applies this “knowledge” to real-world scenarios to solve practical problems.


This is the essential logic of the development of artificial intelligence to this stage. The knowledge summarized by artificial intelligence is not like traditional software, which can be expressed intuitively and accurately. It is more like the knowledge learned by human beings, which is more abstract and difficult to express.

The above statement is still quite abstract, the following will help you understand thoroughly through several aspects:
 Artificial intelligence is a tool


AI is the same as the hammers, cars, and computers we use. It is essentially a tool.


Tools must be used by someone to be valuable. If they exist independently, they are worthless, just like a hammer in a toolbox. No one wields it without any value.


The reason why the whole society is talking about artificial intelligence is that it greatly expands the capabilities of traditional software. There were a lot of things that computers could not do before, but now artificial intelligence can do them.

Thanks to Moore’s Law, the power of computers has risen exponentially. As long as the computer can solve and participate in the links, productivity has been greatly improved, and artificial intelligence has allowed more links to catch the express train of Moore’s Law, so this change It is of extraordinary significance.


But no matter how it changes, traditional software and artificial intelligence are tools that exist to solve practical problems. This point has not changed.


 Artificial intelligence only solves specific problems
“Terminator” and “The Matrix”…Many movies have appeared against sky-defying robots. This kind of movie makes everyone feel that artificial intelligence seems to be omnipotent.
The actual situation is: current artificial intelligence is still at the stage of a single task.


Single task mode


Landline for phone calls, game consoles for games, MP3 for listening to music, navigation for driving…


Multitasking mode

This stage is similar to a smart phone. You can install many apps on a single phone and do many things.


But these abilities are still independent of each other. After booking a ticket on the travel app, you need to set the alarm clock with the alarm clock app, and finally you need to use the taxi app to call a taxi. The multitasking mode is just a superposition of a single task mode, which is far from human intelligence.

Integrate


You are playing Go with a friend, and you find that your friend is in a very bad mood. You could have won easily, but you deliberately lost to the other party, and you kept praising the other party because you don’t want to make this friend more depressed or more depressed. Irritable.


In this trivial matter, you have used a variety of different skills: emotion recognition, Go skills, communication, psychology…


But the famous AlphaGo will never do this. No matter what the opponent’s situation is, even if it loses the game, AlphaGo will win the game ruthlessly, because it can do nothing but play Go!


Only when all the knowledge is formed into a network structure can it be integrated. For example, military knowledge can be used in business, and biological knowledge can also be used in economics.


 Know it, but don’t know why

The current artificial intelligence is to summarize knowledge from a large amount of data. This crude “inductive method” has a big problem:


Scams like Ponzi schemes take full advantage of this!


1.It uses super high returns to attract leeks, and then allows everyone who gets up early to participate in the transfer of money;
2.When onlookers discover that all participants have actually made money, they can simply sum it up as: historical experience shows that this is reliable.
3.So more and more people were jealous and joined until one day the liar ran away.


When we use logic to deduce this matter, we can draw the liar’s conclusion:

4.Such a high return does not conform to the laws of the market
5.Steady profit without losing? I don’t need to take the high risk of high return? Doesn’t seem reasonable
6.Why does such a good thing fall on me? Doesn’t seem right


It is precisely because the current artificial intelligence is built on “inductive logic”, it will also make very low-level mistakes.

Low-level mistakes made by AI

1.Left: The occlusion of the motorcycle makes the AI ​​mistake a monkey for a human.
2.Middle: The occlusion of the bicycle caused the AI ​​to mistake the monkey for a human, and the jungle background caused the AI ​​to mistake the handlebar of the bicycle for a bird.
3.Right: The guitar turns the monkey into a human, and the jungle turns the guitar into a bird


The image above shows the effect of a guitar on PS in a photo of a jungle monkey. This causes the deep network to mistake monkeys for humans and guitars for birds, presumably because it believes that humans are more likely to carry guitars than monkeys, and that birds are more likely to appear in the nearby jungle than guitars.


It is precisely because of inductive logic that a large amount of data needs to be relied upon. The more data, the more universal the experience summarized.


 The history of artificial intelligence


AI is not a brand new thing, it has been developed for decades! Below we introduce the three most representative development stages.

 The history of artificial intelligence

The picture above shows some milestone events in the field of artificial intelligence from 1950 to 2017. In summary, it will be divided into 3 major stages:


The first wave (non-intelligent dialogue robot)


1950s to 1960s


In October 1950, Turing proposed the concept of artificial intelligence (AI) and also proposed the Turing test to test AI.


Within a few years after the Turing test was put forward, people saw the “dawn” of the computer passing the Turing test.


In 1966, the psychotherapy robot ELIZA was born


People in that era spoke highly of him, and some patients even liked chatting with robots. But his realization logic is very simple, it is a limited dialogue library, when the patient utters a certain keyword, the robot will reply to the specific word.


The first wave did not use any brand-new technology, but used some techniques to make the computer look like a real person, and the computer itself is not intelligent.


 The second wave (voice recognition)

1980s to 1990s

In the second wave, speech recognition is one of the most representative breakthroughs. The core reason for the breakthrough is to abandon the idea of ​​the semiotic school and change to a statistical idea to solve practical problems.


In the book “Artificial Intelligence”, Kai-Fu Lee introduced this process in detail, and he was also one of the important figures involved.


The biggest breakthrough of the second wave is to change the thinking, abandon the thinking of the semiotic school, and instead use statistical thinking to solve the problem.

The third wave (deep learning + big data)
Early 21st century


2006 was a watershed in the history of deep learning. Jeffrey Hinton published “A Fast Learning Algorithm for Deep Belief Networks” this year. Other important deep learning academic articles were also published this year, and several major breakthroughs were made on the basic theoretical level.


The main reason why the third wave will come is that two conditions have matured:


After 2000, the Internet industry developed rapidly and formed massive amounts of data. At the same time, the cost of data storage has dropped rapidly making the storage and analysis of massive data possible.

The continuous maturity of GPUs provides necessary computing power support, improves the availability of algorithms, and reduces the cost of computing power.


After various conditions have matured, deep learning has exerted its powerful capabilities. Continuously set new records in the fields of speech recognition, image recognition, NLP and so on. Let AI products really reach the stage of usability (for example, the error rate of speech recognition is only 6%, the accuracy rate of face recognition exceeds humans, and BERT surpasses humans in 11 performances…).

The third wave came, mainly because the conditions for big data and computing power are available, so that deep learning can exert great power, and the performance of AI has surpassed human beings and can reach the stage of “usable”, not just scientific research.


The difference between the 3 waves of artificial intelligence

1.The first two crazes were dominated by academic research, and the third craze was dominated by real business needs.
2.Most of the first two crazes were at the level of market propaganda, while the third craze was at the level of business models.
3.The first two upsurges were mostly about persuading the government and investors to invest in the academic community. The third upsurge was mostly about investors actively investing money in academic projects and entrepreneurial projects in hot areas.

4.Questions were raised when the first two crazes were more frequent, and problems were solved when the third craze was more frequent.

If you want to learn more about the history of AI, I recommend reading “Artificial Intelligence ” by Kai-Fu Lee . The contents of the three waves above are all excerpted from this book.


 What can’t artificial intelligence do?


3 levels of artificial intelligence

When exploring the boundaries of AI, we can first simply and crudely divide AI into 3 levels:

1.Weak artificial intelligence
2.Strong artificial intelligence
3.Super artificial intelligence


Weak artificial intelligence


Weak artificial intelligence is also called restricted field artificial intelligence (Narrow AI) or applied artificial intelligence (Applied AI), which refers to artificial intelligence that focuses on and can only solve problems in a specific field.

For example: AlphaGo, Siri, FaceID…


Strong artificial intelligence


Also known as Artificial General Intelligence (Artificial General Intelligence) or Full AI (Full AI), it refers to the artificial intelligence that can do all the tasks of human beings.

Strong artificial intelligence has the following capabilities:

1.Ability to reason, use strategies, solve problems, and make decisions when there are uncertainties
2.The ability to express knowledge, including the ability to express common sense knowledge
3.Planning ability
4.Learning ability
5.The ability to communicate using natural language
6.The ability to integrate the above capabilities to achieve the set goals


Super artificial intelligence

Assuming that through continuous development, computer programs can be smarter than the smartest and most gifted humans in the world, then the resulting artificial intelligence system can be called super artificial intelligence.


We are currently at the stage of weak artificial intelligence. Strong artificial intelligence has not yet been realized (even far away), and super artificial intelligence is even invisible. Therefore, the “specified field” is still an insurmountable boundary for AI.


 What are the capabilities of artificial intelligence?

If we go a little deeper and explain the boundaries of AI capabilities from a theoretical level, Master Turing must be moved out. Turing was thinking about three questions in the mid-1930s:

1.Do all math problems in the world have clear answers?
2.If there is a clear answer, can the answer be calculated in a limited number of steps?
3.For those mathematical problems that may be calculated in finite steps, can there be an illusory machine that allows him to keep moving, and finally when the machine stops, the mathematical problem will be solved?

Turing really designed a set of methods, and later generations called it a Turing machine. All computers today, including new computers being designed all over the world, have not exceeded the scope of Turing machines in terms of their ability to solve problems.


(Everyone is from the earth, why is the gap so big??)


Through the above three questions, Turing has drawn a boundary. This boundary is not only applicable to today’s AI, but also to future AI.


Below we further describe the boundary clearly:

  1. There are many problems in the world, only a small part of them are mathematical problems
  2. In math problems, only a small part is solvable
  3. Among the solvable problems, only part of it can be solved by an ideal Turing machine.
  4. In the latter part (the part that the Turing machine can solve), there is only part that can be solved by today’s computers
  5. The problems that AI can solve are only part of the problems that computers can solve.

Worried that artificial intelligence is too powerful? You think too much!

In some specific scenarios, AI can perform well, but in most scenarios, AI is useless.

Will artificial intelligence make you unemployed?

This issue is the issue that everyone is most concerned about, and it is also the issue that has the greatest impact on every individual. So take it out separately.

First of all, it is an inevitable trend for artificial intelligence to replace “part of human behavior”

Every new technology or invention will replace part of the labor force:

Time reporting work-table

The work of pulling a rickshaw-the car

Well digging work-drilling machine

It should be noted that technology replaces only certain specific jobs. The digging machine can only help you dig a hole, but it cannot help you determine where you should dig a hole.

The same is true for artificial intelligence. It is not aimed at certain occupations or certain people, but instead replaces some specific labor behaviors.

Secondly, while unemployed, new and better jobs will appear

The history of several technological revolutions tells us that although the emergence of new technologies has caused some people to lose their jobs, they will also create many new occupations. The jobs that are replaced are often inefficient, and the jobs created are often more efficient. Think about pulling a rickshaw, and then think about driving a car.

When artificial intelligence liberates a part of the labor force, this part of the labor force can do more valuable and interesting things.

Don’t be afraid!  AI has many positive effects upon human life

There are 2 points mentioned above:

  1. The essence of artificial intelligence is a tool, and it needs people to use it
  2. Artificial intelligence is not replacing people, but certain parts of work

Therefore, don’t be afraid of artificial intelligence replacing yourself. You should take the initiative to learn AI, become the first person who can use AI, and become a person who can use AI wel .

Think about people who used computers and the Internet 20 years ago. They were very scarce in that era, so they earned the dividends of the Internet era. In the same way, the dividends of the intelligent age will belong to those who can use AI.

Which jobs will be replaced by artificial intelligence?

Kaifu Li proposed a basis for judgment:

If a job takes less than 5 seconds to make a decision, then there is a high probability that it will be replaced by artificial intelligence.

This kind of work has 4 characteristics:

  1. Not much information is needed to make a decision
  2. The decision-making process is not complicated and the logic is simple
  3. Can be done alone, no collaboration required
  4. Repetitive work

Scientists have summarized 3 skills that artificial intelligence is difficult to replace:

  1. Social intelligence (insight, negotiation skills, empathy…)
  2. Creativity (original power, artistic aesthetics…)
  3. Perception and operation ability (finger sensitivity, coordinated operation ability, ability to cope with complex environments…)

 How to meet the intelligent age?

Artificial intelligence will sweep the world like the industrial age. In this case, what we have to do is not to escape, but to embrace this change. Here are some specific suggestions:

  1. To understand the underlying logic and basic principles of the intelligent age, you don’t need to learn to write code, but you need to know what may happen and what is impossible.
  2. Artificial intelligence will infiltrate all walks of life like computers in the future. You have to understand artificial intelligence as much as possible, learn how to use it to solve existing problems, and become an early adopter of artificial intelligence.
  3. Make a good career plan. Don’t choose three no professions (no need for social interaction, no creativity, no strong perception and operation skills)

 To sum up

The basic principle of artificial intelligence: The machine summarizes the laws from a large amount of “specific” data to form some “specific knowledge”, and then applies this “knowledge” to real scenarios to solve practical problems.

On the basis of this basic principle, there are 3 characteristics:

  1. Artificial intelligence is essentially a tool
  2. Artificial intelligence skills can only solve specific problems, not everything
  3. Artificial intelligence belongs to inductive logic, which can tell you what it is, but it cannot tell you why

 So far, artificial intelligence has experienced 3 waves:

  1. 1950s to 1960s: non-intelligent dialogue robots
  2. 1980s to 1990s: speech recognition
  3. Early 21st Century: Deep Learning + Big Data

 Artificial intelligence is divided into 3 levels:

  1. Weak artificial intelligence
  2. Strong artificial intelligence
  3. Super artificial intelligence

 On the issue of unemployment, artificial intelligence will indeed replace some human jobs, but at the same time some new and more valuable jobs will appear. There are 3 skills that will not be easily replaced by artificial intelligence in the future:

  1. Social intelligence (insight, negotiation skills, empathy…)
  2. Creativity (original power, artistic aesthetics…)
  3. Perception and operation ability (finger sensitivity, coordinated operation ability, ability to cope with complex environments…)

 “Attached” 2020 AI Development Trend

Let’s review the important changes in artificial intelligence in 2019:

  1. Important progress has taken place in the NLP field, and pre-training models such as BERT, GPT-2, and XLNET have already played an important role in the product.
  2. The infrastructure is further improved: PyTorch is growing very fast, and TensorFlow is deeply integrated with Keras.
  3. GAN has developed rapidly, and popular products have appeared. DeepFake and ZAO let the public experience GAN technology.
  1. It is also because of DeepFake that the social impact of artificial intelligence is being paid attention to, and AI-related laws are being improved all over the world.
  2. Auto-ML lowers the threshold of AI and makes the deployment of artificial intelligence very easy.

What is the development trend in 2020?

  1. The implementation of 5G will digitize more of the physical world and will further promote the development and popularization of AI.
  2. The integration of data science teams and business teams will be closer.
  3. It is possible to see the development of the multi-task AI model and move further towards general artificial intelligence.
  4. Get rid of the dependence on data, use less data to obtain excellent models.
  5. The NLP field has achieved greater breakthroughs and development.
  6. Improve the interpretability of AI and solve the black box problem
  7. Social problems have intensified, and there have been more discussions on issues such as personal data security, privacy, and algorithmic bias.

Leave a comment

You must Register or Login to post a comment.