ChatGPT alternative apocalypse: the last piece of the puzzle for ordinary people to control artificial intelligence? ｜Titanium Space Time

Write in front:

On April 14, 1980, in Los Angeles, the United States, the movie "The Kramers" won five Oscars: Best Picture, Best Director, Best Adapted Screenplay, Best Actor, and Best Supporting Actress. The grand prize, in an era of numerous masters, defeated the equally great Vietnam War epic film "Apocalypse Now". This great work seems to use its own destiny to reveal the truth of history in an allegorical way:

"There is no winner or loser in any war. The only winner is the war itself"

On November 30, 2022, in San Francisco, the United States, the American artificial intelligence research laboratory OpenAI released the artificial intelligence chat robot ChatGPT. Within 2 months, the number of users exceeded 100 million. This new generation of conversational "artificial intelligence" has become crazy around the world. Reaching 100 million users, it has become the fastest growing consumer app in history. According to the definition of "Father of Artificial Intelligence" Alan Turing: If a machine can have a conversation with humans (through telex equipment) without being able to identify its identity as a machine, then the machine is said to be intelligent. In this long-distance race between humans and artificial intelligence that spans more than 70 years, a new human revelation has surfaced:

"The world is on the eve of a new round of AI technology explosion."

How did China “miss” ChatGPT on the other side of Silicon Valley?

How can humans avoid being eliminated by ChatGPT?

When will the madness and bubble of ChatGPT and artificial intelligence burst?

BAT Who is the “number one player” in artificial intelligence in China?

What are the moral imperatives that will guide the future of artificial intelligence?

At 3 pm on June 7, 2023, Titan Time’s “Reading Moment” will bring a blockbuster live broadcast:

ChatGPT Alternative Apocalypse: The last piece of the puzzle for ordinary people to control artificial intelligence?

To find this piece of the puzzle, maybe we can start by looking back at history...

Artificial intelligence has gone through three historical development waves

► The first wave (1956-1974):

AI trend of thought gives machines logical reasoning capabilities. With the rise of the emerging concept of "artificial intelligence", people are full of imagination about the future of AI, and artificial intelligence has ushered in its first wave of development.

At this stage, artificial intelligence is mainly used to solve algebra and geometry problems, as well as learn and use English programs. Research and development mainly focuses on the logical reasoning ability of the machine.

Among them, the breakthrough development of natural language processing and human-computer dialogue technology in the 1960s greatly increased people’s expectations for artificial intelligence and brought artificial intelligence into the first wave of climax.

However, due to the lack of computer computing power at the time, and due to pressure from Congress, the US and British governments stopped allocating funds to artificial intelligence research projects without clear goals in 1973. The realization cycle of artificial intelligence R&D was lengthened and the industry was cold.

Keywords: recognition pattern, human-computer dialogue, expert system

► Second Wave (1980-1987):

Expert systems make artificial intelligence practical. The earliest expert system was the DENDRAL system developed by Feigenbaum in 1968, which could help chemists determine the molecular structure of a specific substance; DENDRAL proposed the definition of a knowledge base for the first time and also paved the way for the second wave of AI development.

Since the 1980s, "expert system" AI programs in specific fields have been more widely adopted. This system can deduce answers to professional questions based on professional knowledge in the field, and AI has thus become more "practical." ”, the knowledge base system and knowledge engineering on which the expert system relied became the main research directions at that time. However, the practicality of expert systems is limited to specific fields. At the same time, upgrades are difficult and maintenance costs remain high. The development of the industry has once again encountered bottlenecks.

The failure of the artificial intelligence DARPA project in 1990 announced that the second wave of AI had entered its trough. However, the proposal of BP neural network at the same time laid the foundation for subsequent machine perception and interaction capabilities.

Keywords: frame theory, computer vision, computer neurology, BP algorithm, behavioral robot

► The third wave (1993 to present):

Deep learning helps perceptual intelligence mature. The ever-increasing computing power of computers has accelerated the iteration of artificial intelligence technology and pushed perceptual intelligence into a mature stage. AI is combined with multiple application scenarios and the industry is revitalized.

The proposal of the deep learning algorithm in 2006 and AlexNet's major breakthrough in image recognition accuracy on the ImageNet training set in 2012 directly promoted a new wave of artificial intelligence development.

In 2016, artificial intelligence once again gained unprecedented attention after AlphaGo defeated professional Go players. From the perspective of technological development, in the first two waves, the logical reasoning capabilities of artificial intelligence continued to increase, computational intelligence gradually matured, and intelligent capabilities expanded from computing to perception.

Currently, the capabilities of perception technologies such as speech recognition, speech synthesis, and machine translation are approaching human intelligence.

Keywords: transfer learning, autonomous driving, generative adversarial network, unsupervised deep learning, large language model

Titanium Media & Titanium Time, together with the asynchronous community of People’s Posts and Telecommunications Press, launched an in-depth peak dialogue on the new round of AI technology development, paying in-depth attention to the development trends and business changes in the AI ​​field. In specific The key nodes in the development process of artificial intelligence, trying to reveal with many industry practitioners and participants: the last piece of the puzzle for ordinary people to master artificial intelligence.

Introducing important guests

Stuart Russell

In 1962, a British man, Ph.D. in Computer Science from Oxford University Physics and Stanford University, author of "Artificial Intelligence: A Modern Approach", was the first to publish a signed article with Hawking and others calling on everyone to be wary of people who may see threats to us through artificial intelligence. , Vice Chairman of the Artificial Intelligence and Robotics Committee of the World Economic Forum, University of California, Berkeley, where he serves as a professor in the Department of Computer Science and served as department chair and director of the Center for Human-Compatible Artificial Intelligence.

Nico Lee

Partner of TMTpost Media Group, president of ChainDD, a well-known blockchain media data platform under TMTpost Media Group, author of "100 Questions on Blockchain: Understanding Blockchain in Simple and Easy Ways", and the author of the representative article "Shenzhen The Mystery of the Sudden Death of Sailong Company", "36 Krypton Equity Crowdfunding Platform Fell into Fraud", "36 Krypton Fell into Fraud Again, Is China's Equity Crowdfunding Coming to an End?", "Investigation of Jia Yueting and Faraday Future's Assets in the United States", Major reports such as "OKEX's Suspected Illegal Futures Trading Insider" shocked the industry and even the country.

Introduction to important works

In 1995, "Artificial Intelligence: A Modern Approach" was created and is known as the "standard textbook in the field of artificial intelligence."

This book comprehensively and in-depth explores the theory and practice in the field of artificial intelligence (AI). In a unified style, it integrates today's popular artificial intelligence ideas and terminology into applications that have attracted widespread attention, truly integrating theory and practice. Combine.

The book is divided into 7 parts, with a total of 28 chapters. The theoretical part introduces the main theories and methods of artificial intelligence research and traces related ideas from more than 2,000 years ago. The content mainly includes logic, probability and continuous mathematics, perception, and reasoning. , learning and action, fairness, trust, social welfare and safety; the practical part perfectly implements the "modern" concept, and the practical application chooses the currently popular microelectronic equipment, robotic planetary detectors, and online platforms with billions of users Services, AlphaZero, humanoid robots, autonomous driving, artificial intelligence-assisted medical care, etc.

At present, the book has been updated and released on the original basis. A large amount of the fourth edition contains new content or has been rewritten to adapt to new changes in the field of artificial intelligence and present a more unified picture of the field of artificial intelligence. The whole book covers in-depth New achievements in learning and other aspects, as well as new developments in the field of artificial intelligence in the past ten years.

Ti Kong Time｜Reading Time

"Ti Kong Time｜Reading Time" live broadcast room—— is a special reading column of Titanium Time.

Books can be used as a measure of time, as an inscription of a place, as a footnote to a character. Every detailed reading, every exchange of thoughts that runs through the pages of the book, can produce a magical spirit in the titanium space. Power and resonance.

This is the spiritual place for emerging entrepreneurs and a new generation of innovators. Reading, knowledge, and culture are used as links to connect each unique individual and deconstruct the thinking trajectory of each unique soul.

Reading a new book together, you can get a glimpse of the future world from any angle

Reading an old book together, no matter how long it takes, you can find the beauty of human history

We try to explore the essence of the event through dialogue, while listening to the most cutting-edge ideas, while seeking ways to make progress together with them!

Welcome to scan the QR code below to follow Titan Time

More exciting things not to be missed

Lock the live broadcast room to participate in interaction and enjoy the benefits of the live broadcast room

Ti empty time

Titanium Time is a knowledge sharing platform under Titanium Media Group that focuses on pan-finance and new business. It is committed to building independent thinking that spans classics and cutting-edge, redrawing the spiritual map of a new generation of entrepreneurs and innovators, and creating An exclusive thinking space for "traveling through cycles with cognition".

Click to read Dr. Kuan’s recommended book list for the year!

The above is the detailed content of ChatGPT alternative apocalypse: the last piece of the puzzle for ordinary people to control artificial intelligence? ｜Titanium Space Time. For more information, please follow other related articles on the PHP Chinese website!