Depth | What is AIGC? Why McKinsey believes: 50% of occupations will be replaced by AI...

Original Source: AI Dark Horse

Image source: Generated by Unbounded AI‌

Preface:

Before the Spring Festival, Mr. Gua of ClassIn often mentioned ChatGPT and Notion.

In March, both the venture capital circle and the workplace circle began to fall into “AI swiping and AI anxiety”. Zhang Ying of Jingwei Ventures also said bluntly: “This time, the iteration of AI is too fast, which makes me feel more or less anxious about learning. What impressed me very much is that in the third week of March this year, there were consecutive GPT- 4. Baidu Wenxin Yiyan, Microsoft Copilot, Alpaca 7B of Stanford University, ChatGLM-6B of Tsinghua University…”

In May, I felt more deeply about AIGC. At the AIGC conference on 36kr, Baidu showed “Wen Xin Yi Ge”, and at the AIGC conference, which is a dark horse in entrepreneurship, 360 brought “Digital Man-Zhou Hongyi”.

In June, AIGC has amazed and even amazed me.

That’s the AIGC industry class in the dark horse of entrepreneurship:

1) Blue Cursor: Full moratorium on outsourcing, AII In Ai, becoming a marketing leader in generative AI. At present, the Blue Label team has used AIGC tools to serve large customers such as Starbucks, Yili, and Lenovo.

**2) **Rococo: AIGC is used for product design, and Arabic pots are designed for foreigners. One designer can produce 100 plans in two days. The Rococo team also started to use AIGC for LOGO, packaging, and AI drawing.

**3) **Dark Horse Apocalypse: In the past, 1 consultant could write application materials for 1 company in 2 weeks. After passing AIGC, now 1 consultant can write 2 companies in 1 day. “Cost divided by 15, efficiency multiplied by 20”.

**4) **Baidu: Demonstrated “Smart Composition Correction”, which directly circles good sentences and paragraphs, corrects punctuation, and corrects typos on students’ compositions. Generate ratings and comments at the same time. (Scoring details include content, expression, structure, behavioral norms, and scoring. Comments include some evaluation and revision suggestions. They are very professional and attentive!)

**5) **Tsinghua University: Demonstrated the AIGC industry map and scene application. For example, AIGC+ industry, AIGC+ medical care, AIGC+ education, AIGC+ media, AIGC+ e-commerce, AIGC+ games, AIGC+ graphics, AIGC+ video, AIGC+ digital human, AIGC+ scientific research…

I still remember that I typed two sentences in the memo:

*First sentence: Compared with AI large models, we should pay more attention to AIGC applications. The second sentence: This wave of AIGC will inevitably overturn a number of enterprises and employees. *

The text content includes:

1. What is AIGC?

2. How to embrace AIGC? 10 Typical Scenarios and Cases

3. The 50 most noteworthy AIGC companies and applications in China

**01 What is AIGC? **

Sequoia Capital The article stated that AIGC represents the beginning of a new round of paradigm shift.

“In the past, machines could not compete with humans in terms of creation. They could only do some analytical or mechanical cognitive computing. But now it is different. AI has developed to a new stage. Machines have begun to create meaningful and aesthetically pleasing objects. something.”

Tencent Research Institute report believes that some creative work of human beings may be completely replaced by generative AI.

“The biggest impact of AIGC technology is to reduce the marginal cost of creation and knowledge work to zero to generate huge labor productivity and economic value. The key impact of AIGC is to realize low-cost or even zero-cost automated content production.”

Baidu Robin Li publicly stated: "In the future, China will have its own ecosystem, with its own large language model and basic model. At least one, and possibly two or three basic large models will appear in China. Support people to develop various AI-native applications.”

“Use the way of thinking and concepts of AI native applications to reconstruct every product and every business of Baidu.”

You see, this AI wave, Robin Li has gone deep into the sea!

“Robin is both excited and anxious. Sleeping is a waste of time. He didn’t sleep 20 hours a week.”

It seems that Robin Li in 2023 remembered the days of “1417”.

Inside the company, Robin Li reminded his subordinates more than once:

“It’s not easy, it’s hard! Because of the 40,000 Baidu people among us, none of us is an aboriginal in the AI era. We have all experienced the mobile era, and many of us have also experienced the PC era. Our thinking The way is solidified in a sense!”

Like Baidu, there is also a blue cursor that attaches great importance to this AI revolution.

“Completely suspend outsourcing!”

In April of this year, the blue-label email posted on the Internet wrote: "In order to curb the momentum of hollowing out core capabilities, and to lay the foundation for fully embracing AIGC, from now on, the management has decided to stop creative design and planning indefinitely. Four types of outsourcing expenses related to writing, copywriting, and short-term employees.”

Blue cursor Pan Fei responded at the AIGC Summit, the dark horse of entrepreneurship:

The explosion of generative AI will greatly release the creativity of those truly gifted individuals and organizations. Gifted and talented people and teams will be liberated again in front of AGI, the difference is that this time the liberation will be more thorough.

These elements determine that this innovation is no longer an exploratory innovation, but a complete revolution in infrastructure construction and application.

You see, Sequoia, Tencent, Baidu, Blue Label, and more and more well-known companies have reached a consensus strategic agreement - AIGC!

**But, what exactly is AIGC? **

Artificial Intelligence Generated Content ？

In fact, when it comes to AIGC, there are currently two main concepts:

**The first concept: **UGC/PGC/AIGC, focusing on “artificial intelligence generated content”.

**The second concept: **AI/AGI/AIGC, focusing on “generative artificial intelligence”.

Let me talk about the first concept first, AIGC artificial intelligence generated content is not only a type of content classified from the perspective of content producers, but also a content production method. As a new productivity engine, AIGC allows us to enter the era of AIGC from PGC and UGC in the past.

Let’s talk about the second concept, AIGC is also called generative artificial intelligence, which is an important symbol of the artificial intelligence 1.0 era entering the 2.0 era. The cumulative integration of technologies such as GAN, CLIP, Transformer, Diffusion, pre-trained models, multi-modal technology, and generative algorithms has given birth to the outbreak of AIGC. The continuous iterative innovation of the algorithm, the pre-training model has caused a qualitative change in AIGC’s technical capabilities, and multi-modality has promoted AIGC’s content polygons, making AIGC more versatile and stronger in basic capabilities.

These two concepts are not completely different. The underlying essence is the same, but the emphasis is different.

So, we can make a conclusion like this:

AIGC represents the transition of AI technology from perceiving and understanding the world to generating and creating the world, and promotes the next era of artificial intelligence. In the new era, AIGC has become a new productivity engine, creating “content” that only humans could create before.

If you are still a bit unfamiliar with AIGC, you may wish to take a look at the following vivid examples

(1) Write papers with AIGC tools

Tutor: too good to meet my expectations for students

(2) Use the AIGC tool to chat with you virtually, female Internet celebrities: dating 100 boyfriends at the same time

Not only that, the company has also begun to fully embrace AIGC and began to apply it in actual work

(3) BlueFocus, a listed company embraces and applies AIGC in this way

A Change to Workflow

B Improve work efficiency

C Drives Creative Content

D Entering a new era of AIGC

Therefore, pay more attention to the essence of AIGC and be less impetuous.

Blue Cursor Pan Fei thinks:

AI empowerment started many years ago. At that time, the basic logic of AI empowerment was discriminative AI, which is different from the current generative AI.

There are many companies in the Chinese market that mix discriminative AI and generative AI. The innovation focus of the two is different. Discriminative AI is mainly about optimizing prediction and decision-making, while generative AI is mainly about creating new original content.

In the era of AIGC, there will be a group of “super individuals” and a group of “super organizations”.

In the Internet age, the logic of hiring talents is to find 3 people, pay 4 people, do the work of 5 people, and use the 996 model to create the productivity of 6-7 people, but in essence it is based on people + data + machines as the core.

In the AIGC Era, it is completely possible to realize a super individual and create an output of more than 10 people.

So please-

Fully embrace AIGC and dance with the AI era.

02 How can traditional industries embrace AIGC? 10 Typical Scenarios and Application Cases

1) AIGC+Marketing

In April 2023, BlueFocus released the AI² strategy, namely All in AI. At the same time, the mission of BlueFocus has been upgraded to: become the marketing leader of generative AI and create a brand new BlueFocus in the AI era. Based on the marketing industry methodology accumulated for more than 20 years, with the knowledge base as the underlying information support, through BlueFocus’ proprietary data training, build industry models in the directions of data analysis, consumer insights, marketing strategies, and creative concepts; create and generate Formal AI content and creative tool matrix, AI-driven creation of “new interaction, new scene, new link” experience breakthrough and innovative delivery projects.

“The historical opportunity given by AIGC technology iteration is closer to marketing, closer to content, and dancing with AI is our firm and only choice.”

2) AIGC+ Design

The Rococo Group has: Rococo, Rococo, and Jellyfish Smart. Among them, Luoke can nourish 1,000 designers in the traditional company model. Like Didi, Luoke is a platform with 50,000 designers. Jellyfish Intelligence is an AIGC innovation platform, and its products are used by millions of people. people use.

"In the future, a large number of new high-energy small and micro enterprises empowered by AIGC will be born. The competition in the commercial battlefield will become: the competition of “carbon-based managers + silicon-based executive teams” VS “full carbon-based human teams”, in terms of speed and cost , scale and other aspects form asymmetric competition.”

3) AIGC+Enterprise Service

Dark Horse Apocalypse is a subsidiary of Entrepreneurial Dark Horse Group, providing digital and intelligent services for small and medium-sized enterprises across the country, and is committed to becoming the first promotion and service platform for AIGC in China. The Apocalypse large model developed by Dark Horse Tianqi is China’s first AI large model vertical to the field of science and technology innovation.

“One million companies in China apply for national high-tech qualifications every year. They need professional consultants to write application documents, and 700,000 people are engaged in this job. On average, it takes two weeks for each consultant to write materials for a company. Now through Tianqi University Model, a consultant can write the materials of two companies in one day, so as to achieve cost reduction and efficiency increase by dividing the cost by 15 and multiplying the efficiency by 20.”

4) AIGC+ software

“In the era of AIGC, large models will serve as an infrastructure to empower AI in thousands of industries. Today’s large models are similar to power plants in power generation systems. On this basis, more applications are needed, and applications are king.”

In the field of Wenshengtu and Tushengtu, Wonshine Technology launched the world’s first interactive “Tushengtu” AI painting software “Wanxing Aihua”; in the field of Wensheng Video, aimed at overseas marketing companies and teams, AIGC’s “real person” short video overseas marketing artifact “Wanxing Broadcasting Explosion”. Wondershare’s main overseas products have now been connected to Azure OpenAI services, and its Wondershare PDF element, Edraw icon, Edraw brain map, Mockitt overseas version of the ink knife and other products have integrated AIGC capabilities.

5) AIGC+Office

Undoubtedly, the pan-office field will be the “Battle of Chibi” for AIGC.

From Microsoft Office to Kingsoft WPS, as well as notion, Ali DingTalk, Tencent Qiwei, Byte Feishu in the digital age, FOMO (Fear Of Missing Out), all of them are both excited and frightened.

Who will bravely stand at the forefront of AIGC, and who will be photographed on the beach by the tide?

▲ ChatGPT is embedded in Office365, which can automatically generate presentations and forms

▲ Dingding meeting can generate meeting minutes at any time and automatically summarize meeting minutes and generate to-do items

▲ AII in One’s Notion once again integrated AI into it

6) AIGC+Finance

CICC Yao Zeyu cited three cases, namely AIGC+Bloomberg, AIGC+Morgan Stanley, AIGC+Deutsche Bank. These three examples represent applications in the field of financial information data services, applications in the field of risk control, and applications in the field of wealth management.

AIGC+Bloomberg: We know that in the field of financial information and data services in China, when using some data terminals, they used directory query to find what they wanted layer by layer. The superposition of Bloomberg and AIGC allows users to directly obtain some data, indicator libraries, existing charts, etc. they want through question-and-answer methods.

AIGC+Morgan Stanley: This mainly involves the field of wealth management. In the past, when financial advisors and investment advisors served high-net-worth clients, their service quality depended heavily on the professional level of investment advisors. Now, with the addition of AIGC, the entire The wisdom of Morgan Stanley’s bottom-level analysts, the ability of research reports, and the ability of asset allocation can be efficiently exported to investment advisors through AIGC, and the investment advisors can then be exported to customers through B2C, which greatly improves the efficiency of investment advisory services.

**AIGC+Deutsche Bank: **More applications in the field of risk control, through the extraction and processing of unstructured data, can better identify the risks of counterparties and provide early warning.

7) AIGC+Education

AI Dark Horse believes that we can focus on the layout and application of the three companies in the AIGC field, iFLYTEK, NetEase Youdao and TAL. Educational technology companies have very limited changes in education, and public schools are the core of education. We can pay attention to the exploration and practice of AIGC by universities such as Peking University, Tsinghua University, and Fudan University.

**HKUST Xunfei:**Based on Xunfei Xunhuo Cognitive Big Model, HKUST Xunfei launches the Xinghuo Language Companion APP: an AI speaking assistant empowered by the big model, which realizes 1-on-1 oral training and portable translation. The toC business “AI learning machine” and the ToB business “smart education” will also innovatively apply the Xinghuo large model.

**Netease Youdao:**In the wave of artificial intelligence, explore the frontier of “scene first”, explore more paths of AI personalized teaching, and help more people achieve efficient learning. Among them, the application of the “Ziyue” educational model has gradually come to the fore. Youdao virtual population language coach Echo appeared in 2023 WAIC not long ago.

**TAL: After the double reduction, TAL/Xueersi did not find the second curve of “Oriental Selection” as quickly as New Oriental, but TAL/Xueersi released a self-developed mathematical model MathGPT is aimed at global mathematics enthusiasts and scientific research institutions. It focuses on problem-solving and lecture algorithms in the field of mathematics. It will launch a product-level application based on this self-developed large model within this year.

8) AIGC+ Games

9) AIGC+E-commerce

10) AIGC+Medical

**In addition, AIGC+ digital human also has a wide range of application scenarios, which deserves attention. **

360 Zhou Hongyi even believes that the digital human is the most important application entrance of the future artificial intelligence model.

Different from traditional digital humans who can only output content according to established scripts, 360AI digital humans are characterized by memory, human design and personality. In May, at the Dark Horse AIGC Summit, Zhou Hongyi shared about entrusting digital people.

03 China’s most noteworthy, 50 AIGC companies and applications

Not long ago, a company founder posted on social media:

“Recently, I chatted with two game teams. The first team cut off the original painting outsourcing team, and the second team cut down the translation outsourcing company. AI has actually begun to affect the jobs of many people.”

You may be like me, and you have heard the news that “a game company has been cut from 200 people to only 2 people”.

In June 2023, McKinsey released the research report “The Economic Potential of Generative Artificial Intelligence”, pointing out that artificial intelligence AI will replace more than half of the jobs in the world between 2030 and 2060. In other words, 50% of occupations will be replaced by AI.

Someone asked: AIGC set off a new wave of “unemployment”?

Someone replied: AIGC will definitely subvert some enterprises and employees.

Proverbs mixed with chicken soup: AI will not make you unemployed, people who can use AI will make you unemployed.

Haha, that’s the taste…

Sequoia Capital believes that: from social media to games, from advertising to architecture, from programming to graphic design, from product design to law, from marketing to sales, every industry that requires human originality will likely be subverted.

Some jobs will be completely replaced by generative AI, and some will better promote human-machine collaboration with the help of generative AI-but in general, generative AI will have a very wide range of end application markets, helping people more Create better, faster and cheaper.

Text field, code field, image field, voice field, video field, 3D field…

It is not difficult to foresee that AIGC will completely change these fields, and then restructure thousands of industries.

Share a little story:

Last Thursday, I went to Baidu company to visit and study with more than 80 black horse students.

As soon as you enter the lobby, you can see a banner hanging high above your eyes, with 11 large characters written in bold on it:

Embracing the era of large models, actions speak louder than words!

“Actions speak louder than words. Baidu is serious about AI.” A black horse student said. “It is said that this large-scale model is no less than the opportunity of the era when the PC Internet, mobile Internet, and iPhone were invented.” A person next to him continued.

I asked a few students about their industries, some of them are in SaaS, some are in Internet, there are also consumption, training, automobile, lawyer, yoga…

It is said that AI will subvert thousands of industries! In my opinion, thousands of industries are taking the initiative to reconstruct AI. It’s quite like “I’m not the only one on the mountain, I’m on the mountain when I come”.

From Internet technology giants to small and medium-sized start-ups, it hasn’t been this lively for a long time.

Generative AI is divided into application layer, model layer and base layer. The base layer and the large model are more like a giant game and an arms race. It is closely related to chips and data. Without a huge amount of capital reserves, it is difficult for small and medium-sized enterprises to really build a large model.

However, based on large models at home and abroad, small and medium-sized enterprises can also make various AIGC applications and products.

Sequoia Capital made an inventory of foreign AIGC companies and their applications: