AI Business Overview - The History of AI
Artificial Intelligence (AI for short) is one of the hottest topics in the field of science and technology today, and it is profoundly changing people's work and life. It is a discipline that studies how to enable machines to simulate and perform intelligent human behavior. With the improvement of computing power and the rise of big data, artificial intelligence is leading the transformation of technology and society. From self-driving cars to intelligent assistants, from speech recognition to machine translation, to the current emergence of large language models, artificial intelligence has penetrated into various fields, showing amazing potential, and bringing great benefits to people's work and life. change and convenience.
Artificial intelligence has experienced three waves of development driven by the development of the technological revolution, from symbolism to connectionism, and then to the current large-scale pre-training generative model based on deep learning, which has sounded the clarion call for the great development and breakthrough of artificial intelligence. Artificial intelligence is constantly updating, iterating and introducing new ones at an unprecedented speed, and it is quickly taking root in thousands of industries, which brings hope for the development of general artificial intelligence.
1. Four technological revolutions
The development of artificial intelligence is inseparable from the promotion of the technological revolution. The technological revolution has promoted the progress of human society and constantly changed people's work and lifestyle. Every technological revolution brings new technological breakthroughs and social changes, and artificial intelligence is the most eye-catching part of today's technological revolution. Every technological revolution provides an important foundation and impetus for the development of artificial intelligence. By reviewing the development of the technological revolution, we can better understand the development of artificial intelligence and its profound impact on people's lives and society.
The first industrial revolution: machinery replaced manpower, and the invention of the steam engine
The first industrial revolution refers to a major change that occurred in Britain from the end of the 18th century to the middle of the 19th century. This industrial revolution was characterized by mechanization and factory manufacturing, which completely changed the production mode dominated by agriculture and handicrafts, and opened the prelude to the industrial age. The core innovation of the first industrial revolution was the application of steam power. Steam is generated by burning fuel to drive various machinery and equipment. One of the most representative inventions is the steam engine, invented by James Watt and others. The application of the steam engine greatly increased productivity and led to the rapid development of agriculture and industry.
This period also saw a series of other important inventions and innovations. The textile industry ushered in automatic spinning machines and power looms, realizing large-scale production of weaving cloth. Railway transportation began to rise. Since the steam engine was carried on the train, the logistics speed and trade development were greatly accelerated. These technological changes resulted in dramatic increases in production efficiency and scale, triggered urbanization and labor mobility, and gave birth to the formation of an industrial capitalist economic system.
The first industrial revolution had a huge impact on society and promoted the process of global industrialization and the evolution of the global economic structure. It laid the foundation for modern industrial civilization, drove the rapid development of economy, society and technology, and opened a new chapter in industrial society. .
The second industrial revolution: the invention of electricity and the internal combustion engine
The second industrial revolution was a global transformation that occurred from the end of the 19th century to the beginning of the 20th century, with electrification and the application of the internal combustion engine as the core features. During this period, building on the foundations of the first industrial revolution, new innovations and technological breakthroughs revolutionized production and transportation, driving urbanization and industrialization.
An important innovation of the Second Industrial Revolution was the widespread use of electricity. Through the establishment of generators and grids, electricity replaced traditional steam power as the main source of energy. This innovation accelerated the development of mechanized production and improved production efficiency and quality. At the same time, the electric revolution has also brought about changes in family and social life, such as the popularization of indoor lighting and electrical appliances, making life more convenient and comfortable.
In addition to the electrical revolution, the invention of the internal combustion engine also had a profound impact on the Industrial Revolution. The internal combustion engine further revolutionized transportation by converting fuel such as gasoline into mechanical energy. The invention and popularization of automobiles have greatly improved people's travel methods and accelerated the circulation of commodities and the expansion of markets. In addition, the internal combustion engine also drove the development of aircraft and ships, making cross-border and long-distance transportation possible.
With the widespread use of electricity and the internal combustion engine, the second industrial revolution further promoted urbanization and the development of industry and commerce. Large-scale factories continue to emerge, and logistics and transportation networks are built and improved. This led to large-scale labor mobility and occupational division, which further exacerbated the expansion of cities and changes in social structures.
The second industrial revolution had a profound impact on various industries, bringing huge changes from economy, society to culture. It laid the foundation for the modern industrial age and provided important experience and inspiration for the process of industrialization worldwide.
All in all, the second industrial revolution revolutionized production and transportation through the innovation of electricity and the internal combustion engine, and promoted the rapid development of urbanization and industrialization. This revolution propelled human society into a new industrial age and had a profound impact on our way of life.
The third industrial revolution: information technology, Internet, computer
The third industrial revolution refers to a global technological change and economic transformation from the end of the 20th century to the beginning of the 21st century, with the development of information technology as the core feature. It integrates computer technology, the Internet, communication technology and scientific innovation, bringing unprecedented changes and possibilities.
Key technologies of the third industrial revolution include the rapid development of computers, the Internet and information and communication technologies. The rapid progress and popularization of computers has promoted the improvement of data processing and analysis capabilities. The popularization of the Internet and the realization of global interconnection have opened up infinite connections between people, people and things, and things and things. Innovations in various communication technologies, such as mobile communications and satellite communications, have made information transfer more convenient and widespread.
These technological breakthroughs have had enormous economic and social impacts. Information flows faster, distances are shortened, and commerce and trade are increasingly globalized. The rise of e-commerce has created new business models and market opportunities. Digital innovation has changed people's lifestyles, entertainment and social interactions, forming a brand new digital culture.
The third industrial revolution also brought many innovations related to sustainable development. The development of clean energy and energy-saving technologies has promoted the use of renewable energy and the reduction of carbon emissions. The concept of smart cities has emerged to improve urban sustainability and quality of life through technologies such as smart transportation, energy management, and smart buildings.
In addition, the third industrial revolution also brings new challenges and problems. With the development of artificial intelligence and automation technology, employment opportunities have changed and need to adapt to new technical requirements and ways of working. Issues such as data privacy and network security have also attracted extensive attention and exploration.
In general, the third industrial revolution, driven by the boom in information technology, has changed our economic, social and cultural landscape. It connects people around the world, breaks down traditional constraints, and lays the foundation for a smarter, sustainable, and global future.
The Fourth Industrial Revolution: Artificial Intelligence, Big Data, Internet of Things, Blockchain
The fourth industrial revolution refers to the ongoing global transformation, with the development of artificial intelligence and Internet of Things technology as its core feature. During this period, the convergence of digital technologies is changing our lives, work and society.
The Fourth Industrial Revolution empowers computer systems with the ability to learn and reason with the help of artificial intelligence (AI), enabling them to process large-scale data, mimic human decision-making, and self-adjust. This technological breakthrough is changing the way people interact with technology. The application of artificial intelligence in various fields shows amazing potential and creativity. From smart assistants to self-driving cars, from smart homes to smart medical care, they are constantly improving our quality of life and efficiency.
At the same time, IoT technology is also a key driver of the fourth industrial revolution. The Internet of Things connects various devices and items through wireless communication, realizing real-time perception, transmission and processing of information. In this way, AI systems can make more accurate decisions by acquiring and analyzing data generated by the IoT, thereby improving the efficiency of production and services.
In addition, the fourth industrial revolution also involves the innovation of other key technologies, such as cloud computing, big data analysis, virtual reality and augmented reality. These technologies collaborate and promote each other to build a brand new digital ecosystem.
The Fourth Industrial Revolution has had far-reaching impacts on the economy, society and the environment. New business models and digital platforms have spawned innovative products and services, leading to business opportunities and competition on a global scale. At the same time, it also raises a series of challenges, such as employment changes, privacy security, and ethical issues, which require us to think and solve together.
The Fourth Industrial Revolution is reshaping our world. Through the rapid development of artificial intelligence, Internet of Things and other digital technologies, we are entering a new era of digitalization, changing every aspect of business, society and personal life. In this era of change, we need to continuously adapt and respond to the challenges brought by new technologies, and actively explore their potential to make positive contributions to building a smarter, inclusive and sustainable future.
To sum up the above four industrial revolutions, each industrial revolution has brought about profound technological changes and social transformation, reshaping all aspects of human society. True to their name, these industrial revolutions centered on the development of industry and drove dramatic changes in the economy, culture, and social way of life.
2. Key events of the fourth wave of technological revolution
We are currently at the forefront of the fourth scientific and technological revolution. This time artificial intelligence, as an important driving force to promote change, is leading a new era with the support of machine learning and deep learning. From self-driving cars to smart assistants, from face recognition to voice recognition, artificial intelligence is penetrating into all fields of life and work. Some key events on the eve of the fourth technological revolution are listed below:
2010-Google's first self-driving car
2012-Robot autonomous learning ability
2014-Cloud computing and big data ushered in a climax
2015-Genome editing technology came out
2016-Internet of Things application acceleration
2017-AlphaGo defeated the world's highest level chess player
2018-Blockchain technology applications continue to expand
2019-The first year of 5G commercial use
2020-The new crown epidemic accelerates digitization and intelligence
2020-Great The large-scale pre-training generation technology GPT-3 comes out
2021-SpaceX obtains the satellite Internet license, and reusable rocket technology succeeds again
2022-Quantum computer research has made a major breakthrough
2022-ChatGPT is born 2023
-CTP4, GLM, News Feixinghuo, Ali Tongyi, Baidu Wenxin, Shangtang Ririxin, Huawei Pangu
3. Three stages of AI development
Driven by the scientific and technological revolution, the development of artificial intelligence technology has gone through three stages of development. Each stage has brought different levels and types of technological breakthroughs and application innovations, which have had a profound impact on the development of artificial intelligence. The influence of the third stage is deepening day by day, artificial intelligence is being widely used in various fields, promoting the development of technology and society. Looking back at the technological breakthroughs in the three development stages of artificial intelligence and the development trajectory of application innovation, we can better understand the development of artificial intelligence, understand the underlying logic of each technological breakthrough and business innovation, and lay a solid foundation for the future development of artificial intelligence. Explore possible ways forward to provide basic support. Look back at history and run to the future.
The development of artificial intelligence can be traced back to the 1950s, when scientists began to explore how to enable machines to simulate and execute human intelligence behaviors. With the rapid development of computer technology and scientists' exploration of simulating intelligent behavior, artificial intelligence has begun to emerge. The history of artificial intelligence development can be divided into three stages, and each stage has brought important technological breakthroughs and application progress. The following is a detailed description of these three stages:
The first stage (1940-1980s): Symbolism
The first stage of artificial intelligence mainly focused on the research of basic theories and reasoning methods. Early AI research focused on physical symbol systems, logical reasoning, and problem solving. Representative results include machine theory of logical reasoning, methods of resolution, etc. Researchers try to simulate the intelligent thinking process of the human brain by constructing a symbol system. Using logical and formal methods, they attempt to solve problems using inference rules. One of the representative results is the machine theory of logical reasoning. The researchers transformed the human reasoning process into formal logic and mathematical models, and realized some basic reasoning and problem solving through computer programs. In addition, the resolution method is also an important result of this stage, which is used for reasoning and proving theorems.
Although some important progress was made at this stage, due to the limitation of computer processing power and the limited understanding of human brain intelligence at that time, this stage did not achieve the expected results. For a period of time thereafter, the development of artificial intelligence fell into a downturn, which was considered to be the "winter of artificial intelligence". However, this stage laid an important foundation for later artificial intelligence research. It reveals the key issues and challenges of artificial intelligence research, and also inspires subsequent researchers to explore new ideas and methods. This stage lays the foundation for the field of artificial intelligence research. The key events mainly include:
- In 1936, Alan Turing proposed the concept of Turing machine, which is a theoretical concept that helps us understand and study the nature of computing, and it also provides the basis for thinking and solving various computing and intelligence problems , had a profound impact on the development of computers and artificial intelligence
- in 1950 Alan Turing proposed the concept of "Turing Test", which can evaluate whether a computer can exhibit behavior comparable to human intelligence. The purpose of the test is to explore whether computers can simulate human thought processes and behaviors. The Turing test has therefore become one of the important standards for measuring artificial intelligence research.
-In 1956, many scientists held the Dartmouth Conference (Dartmouth Conference) at Dartmouth College, announcing that they are committed to developing programs that can exhibit intelligent behavior. Milestone, the word "artificial intelligence" is also an important outcome of the conference.
- In 1958, John McCarthy created the artificial intelligence programming language LISP, which was the first language specifically designed for artificial intelligence development. Its flexibility and expressiveness made it widely used in subsequent research.
-In 1966, Joseph Weizenbaum developed the ELIZA chat robot, which was a milestone product in the field of early artificial intelligence and natural language processing. Based on pattern matching and simple conversion rules, it simulates the dialogue process of a psychological counselor and can interact with users based on text. While ELIZA has no true understanding or consciousness, it is able to converse with its users in a seemingly intelligent manner. Although the principle of ELIZA is relatively simple, it shows how to use expert knowledge and reasoning rules to simulate the ability of human dialogue, which laid the foundation for the later development of chatbots and natural language processing technology, and has an important role in artificial intelligence research. historical meaning.
The ELIZA chat robot is essentially a simple expert system. The expert system is an important achievement at this stage and has also been commercially applied to a certain extent. An expert system is an artificial intelligence system based on expert knowledge and reasoning rules. It stores the knowledge and experience of domain experts and uses reasoning mechanisms to solve problems and make decisions. It has the characteristics of knowledge base, reasoning mechanism, interpretation ability and learning ability, and can simulate the decision-making and problem-solving process of human experts in specific fields. Expert systems are widely used in medical, chemical, industrial and other fields, providing an important foundation for the development and application of artificial intelligence. Some important expert systems are as follows:
-MYCIN, a medical diagnosis system developed in 1976, is specially used for diagnosis and treatment recommendations for bacterial infections. It is the first expert system applied on a large scale in clinical medicine, which proves to the world the potential and value of expert systems in complex fields.
- In 1980, XCON, an expert system developed by Digital Equipment Corporation, was used to configure and customize computer systems. It has achieved remarkable success in enterprise-level applications and promoted the commercial application of expert systems.
-An inference engine R1, which appeared in 1982, has the functions of rule interpretation, rule execution and rule maintenance. The design and implementation of R1 provided examples and foundations for later expert system development tools.
-The PROSPECTOR expert system developed in 1983 is an expert system specially used for mineral exploration and resource evaluation. It uses geological data and domain expert knowledge to predict the location and value of mineral deposits
The second stage (1980s~2000s): connectionism Artificial
intelligence entered the era of connectionism, and the development of artificial intelligence in this stage is mainly characterized by neural networks and machine learning. It mainly revolves around simulating the connections and rules of neurons to achieve intelligence. At this stage, the field of artificial intelligence began to focus on the exploration of large-scale parallel processing and distributed computing. Compared with traditional methods based on symbolic reasoning, connectionism focuses more on realizing intelligence by imitating the neural structure and learning mechanism of the brain. The main features of the connectionist stage:
Neural network: It is a learning model based on the interconnection of biological neurons, which simulates the information processing process by training and adjusting the connection weights between neurons.
Distributed parallel processing: Accelerate machine learning and decision-making processes by working simultaneously with multiple neurons or neural networks. This distributed parallel processing method helps to deal with large amounts of data and complex problems.
Learning and self-adaptation: The neural network is trained through algorithms such as backpropagation, and the connection weights are automatically adjusted to optimize the performance of the model. This learning mechanism enables connectionist AI to automatically extract features and regularities from a large amount of data, thereby improving the performance and adaptability of the system through training and learning.
Nonlinear models: Compared with traditional symbolic reasoning methods, connectionism is more inclined to use nonlinear models. Neural networks can achieve complex nonlinear mapping by stacking multiple layers of neurons, thereby improving the ability to model real-world problems.
In general, the core of the connectionist stage lies in simulating neural networks and machine learning, and realizing intelligence by constructing multi-layer neural networks and using distributed parallel processing. Research at this stage laid the foundation for further developments in the field of artificial intelligence and paved the way for later technologies such as deep learning. The main representative things are:
- In 1958, Frank Rosenblatt and Robert Baum proposed the "perceptron" model, which is a simple neural network model. This model has input and output layers and learns to classify input data by adjusting weights, laying the groundwork for connectionism.
-In 1986, David Bowes and others proposed the "Backpropagation" (Backpropagation) algorithm, which is the core algorithm for neural network training and learning in connectionism. This algorithm can efficiently calculate the error and adjust the connection weights in the neural network, so that the neural network can gradually optimize its performance.
-In 1989, Andrew Ng and others launched the "ALVINN" (Autonomous Land Vehicle in a Neural Network) project at Carnegie Mellon University. The project uses neural networks to train self-driving cars, and successfully realizes road recognition and steering control, marking a breakthrough in the practical application of connectionism.
The third stage (from the beginning of the 21st century to the present): deep learning
This stage is mainly based on deep learning methods, trying to achieve more advanced artificial intelligence by using large-scale data and powerful computing power. The characteristics and characteristics of artificial intelligence at this stage mainly include the following aspects:
Deep learning: use multi-level neural network structure for feature extraction and representation learning. Through large-scale data training, deep learning models can automatically discover patterns and laws in the data, and realize more complex tasks and functions.
Large-scale data: the success of deep learning methods is inseparable from the support of large-scale data. The performance and generalization ability of the model can be improved by training the deep neural network through massive data. The availability of large-scale data and the advancement of storage technology provide an important foundation for the development of deep learning.
Powerful computing power: deep learning methods require a lot of computing resources for training and inference. With the development of computer hardware and the popularization of cloud computing, people can use distributed computing and high-performance computing platforms to accelerate the training process of deep learning, so as to achieve faster and more accurate model training.
Multi-modal learning:Deep learning methods can handle not only traditional structured data, but also multimodal data such as images, audio, and text. Through multimodal learning, different types of data can be fused and correlated for more comprehensive and in-depth understanding and analysis.
Artificial intelligence at this stage is widely used in various fields, including image recognition, speech recognition, natural language processing, machine translation, recommendation systems, driverless driving, etc. Especially from the end of 2022 to the beginning of 2023, the continuous emergence of generative pre-trained multi-modal large models, as well as the innovation of AI application development paradigms based on these large models, let people see the hope of leading to general artificial intelligence . Key representative events at this stage include:
- In 2012, Alex Krizhevsky of the Google team participated in the ImageNet large-scale visual recognition challenge, using deep convolutional neural networks (CNN) to achieve revolutionary breakthroughs in image classification tasks, leading deep learning development trend.
- In 2016, DeepMind's AlphaGo defeated the world's top player Lee Sedol in the game of Go, drawing widespread attention to the capabilities and potential of artificial intelligence in complex decision-making games.
-In 2017, the Google machine learning team proposed a paper called "Attention is All You Need", which proposed the concept of self-attention mechanism, which is a neural network model based on self-attention mechanism (self-attention mechanism) , which has achieved remarkable results in the field of natural language processing, and is widely used in tasks such as machine translation, text summarization, and question answering systems. Since then, Google Transformer has gradually become an important research direction in the field of natural language processing. The BERT and GPT large models proposed later are all based on the Transformer model. These models have achieved very good results in various natural language processing tasks.
-In November 2022, OpenAI launched the artificial intelligence chat robot program ChatGPT, which interacts in the form of text, can interact with natural human dialogue, and can also be used for complex language work, including automatic text generation, automatic question and answer, automatic summary, Various tasks such as code editing and debugging. The emergence of ChatGPT marks a major progress in artificial intelligence chat robot technology, providing people with a more convenient and efficient way to obtain information and solve problems.
- In March 2023, OpenAI launched the GPT-4 artificial intelligence multi-modal large model, which is an upgraded version of GPT-3. By adding more training data, improving the training algorithm and adjusting the model structure, the performance of the model has been further improved. Expressiveness and applicability. Compared with GPT-3, GPT-4 has higher language understanding ability, better text generation ability, stronger language interaction ability, and wider application scenarios. GPT-4 not only supports longer context, higher accuracy and generalization capabilities, but also supports multimodality, such as speech recognition and image understanding.
- In March 2023, Baidu officially released the AI model Wenxin Yiyan. Based on the large-scale model built by Baidu's smart cloud technology, Wenxinyiyan has been widely integrated into all of Baidu's businesses. And launched Wenxin NLP large model, Wenxin CV large model, Wenxin cross-modal large model, Wenxin biological computing large model, Wenxin industry large model. And it provides a variety of large-scale model API services, which can call large-scale model capabilities through zero code, and freely explore how large-scale model technology can meet user needs
; Major core competencies, namely text generation, language understanding, knowledge question answering, logical reasoning, mathematical ability, coding ability, and multimodal ability.
The above-mentioned large models are only a small part of the current large AI models. With the continuous advancement of technology and the continuous advancement of research, more and more powerful AI large models are emerging.
4. Conclusion
With the rapid development of science and technology, the current AI large model has been regarded as a key factor to promote the improvement of the field of artificial intelligence, and the large model has become the research and application direction leading the wave of technology. Large models are artificial intelligence models with huge scale and complex structure, with hundreds of millions of parameters and deep neural network architecture. These models have achieved impressive results on various tasks by learning massive amounts of data and optimizing deep neural networks. 2023 can be said to be the first year of the development of AI large language models. With the release of OpenAI's ChatGPT and GPT-4, the development wave of artificial intelligence large language models has been ignited. Major technology companies have launched their own large language model products. Various countries have promoted the development of large language models as an important breakthrough in artificial intelligence technology, and have carried out business and technical layouts one after another. Moreover, the development of AI large models has tended to two modes: open source large models and commercial online large models. Open source large models are large-scale deep learning models released based on open source software license agreements. Usually jointly developed and maintained by researchers, developers, or the community, and provided in a public code repository. The advantage is that it can be obtained and used for free, and the open code allows users to freely view, modify and customize the model to suit specific needs and tasks. Open source big models also facilitate knowledge sharing and collaboration between academia and industry. Representative models include Transformers, BERT, and ChatGLM; online large models refer to large-scale machine learning or deep learning models deployed on the cloud or on the network. These models are usually provided by cloud service providers or online platforms, and accessed and used through interfaces or APIs. The advantage of the online large model is that users do not need to pay attention to the underlying hardware and software architecture, and can obtain the prediction results of the model only through network requests. The online large model can also realize real-time or on-demand model invocation, which is suitable for various application scenarios, such as speech recognition, image processing and natural language processing. These two models will have different performance potentials in different fields. In the future, the development of large-scale AI models and artificial intelligence will definitely have a profound impact on people's production and life. Only by trying to understand it, embrace it, use it, and optimize it can we Only by better complementing and integrating with it can we better experience the superior experience brought by artificial intelligence.