CEOs, suffering from "AI anxiety"

Source: “Deep Burning” (ID: shenrancaijing), author: Wang Min, editor: Liming

In the past half a year or so, ChatGPT has exploded in popularity, leading to a large AI model. People in the industry believe that AI has entered the 2.0 era from the 1.0 era, and a truly epoch-making technological change is taking place.

When the AI wave sweeps across, no one can stay out of it. From Beijing, Shanghai, Guangzhou to small counties, AI has “pervasively penetrated”. When an opportunity bigger than the Internet appears, no one wants to miss this wave.

CEOs of companies who are always aiming at business opportunities are even more excited and anxious. They are excited that new changes mean new opportunities, and at the same time they are anxious that they will be left behind in this great change.

Image source: Generated by Unbounded AI tool

Shenran chatted with several CEOs who have embarked on this wave of AI. In the face of this wave, the pressure for entrepreneurs in the AI field lies in building their own core barriers; CEOs of small and medium-sized companies that focus on e-commerce and advertising pay more attention to details and rely on new AI tools to reduce costs and increase efficiency; The company is no longer satisfied with only being the user of the large model, but has personally stepped down to carry out the adaptation research and development of the vertical scene.

They are anxious about the choice of direction, and even more anxious about the landing scene. They hope that AI can reduce costs and increase efficiency in the short term, but in the long run, they hope to use AI as a new growth driver.

Although this new round of transformation has just begun, when the gunshots are fired on the AI battlefield, CEOs must seize the time to deploy their troops. It is not only for development, but also for survival. Going forward has become the only choice.

01 The wave of AI is sweeping, and the CEOs are all rolled up

From 2022 onwards, CEOs will gradually become anxious.

With Midjourney and Stable Diffusion igniting the technical world, AI painting became the first field to be impacted.

As an entrepreneur in the field of AI painting, Wang Dengke, the founder of 6pen, told Shenran that when Stable Diffusion announced its open source last year, he felt at a loss for a while. Knowing that you have to move fast and put in something of long-term value, even though it takes a lot of energy, is the only way to survive.

By the end of the year, ChatGPT surged, which made a group of CEOs feel shocked, happy and worried.

“As soon as ChatGPT was released, we felt the impact. We didn’t expect this technology explosion to come so fast,” said Xing Zai, CEO of StartDT, a data intelligence track company, and founder of Singularity Cloud, to Shenran. Anxiety, not knowing exactly what impact the new technology might have, but more excitement, which means there may be more opportunities.”

Yiyi, the owner of Lianmuxu Youth Snack Shop, who has been in business for nearly two years and has opened several online and offline stores, also became anxious because of AI in the first half of this year.

“When ChatGPT was first released last year, I tried it, but it was just for fun, so I wasn’t shocked,” she said, “But after the Spring Festival this year, I found that friends and company employees who started businesses together were discussing ChatGPT.” The enthusiasm of people made her feel that if she continued to remain indifferent, she would definitely be left behind by the times.

Before May of this year, new developments related to large-scale AI models at home and abroad emerged one after another, aggravating the anxiety of CEOs. However, while CEOs are anxious, they are also speeding up their actions. It has become the norm to frantically input new AI knowledge this year.

**CEOs of small and medium-sized companies, especially in the e-commerce industry, are often more anxious about the implementation of new AI tools. **

In April, driven by the surrounding environment, YiYi spent several thousand yuan to sign up for the ChatGPT community learning course, “I want to be able to control AI tools, not be controlled by tools.” YiYi’s demand for ChatGPT is from the brand From event planning to copywriting details, as well as the operation and maintenance of the community, the optimization of customer service skills, and the channel promotion of communication with consumers, the selected learning community includes brand marketing and marketing directors, new media practitioners, Small and medium-sized business owners, etc., people from various fields gather together to explore how to combine ChatGPT with their own fields.

The two young people who came out of major Internet companies have had many entrepreneurial experiences in the fields of e-commerce and Web 3. After this year, seeing the popularity of the AI wave, they also invested in the entrepreneurship of “AI service providers”. “Xiaozheng” is the brand IP and provides AI tool training services to the outside world. Its customer base is mainly small and medium-sized companies in the e-commerce industry, and some are the marketing departments of large foreign companies.

Xiaozheng told Shenran that small and medium-sized e-commerce companies are more concerned about the effect of landing and will pay great attention to details. For example, if they want AI to generate a dog leash, the company will care whether the picture drawn by AI is the same as the picture placed in the online photo . Being able to dig into the details is also because the training of small and medium-sized e-commerce companies using AI tools is more obvious. In e-commerce marketing, store decoration, etc., AI tools can be used to quickly improve efficiency.

**Industry leading companies or companies with a certain size are more anxious about the direction, how to combine the company’s existing business with AI, or use AI to develop new businesses. **

Hou Zhenyu, vice president of Baidu Group, once said that when Wenxin Yiyan was just released, most of the customers who came to Baidu in March were led by the top leaders and CEOs of various companies. .

Xing Zai said that in the second half of last year, Singularity Cloud had already held a strategy meeting and began to mobilize its forces to follow up on the large model. After ChatGPT came out, the effect was really amazing. After this year, the company organized another extremely Guest night talks, synchronizing the cognition of new changes in AI to establish a new direction.

Wu Qiang, CEO of Tianrun Rongtong, an intelligent customer service company, previously said that AI has activated the minds of customers and mobilized their interests. If companies do not keep up with the direction of customers’ attention, they will be eliminated by the market.

“The company’s founders and executives immediately paid attention to changes in technology iteration and market demand, and quickly promoted the research and development and scene mining of large-scale model-related fields,” Tian Fengzhan, chief scientist of Tianrun Rongtong, told Shenran that after the release of ChatGPT, Quick internal research and testing found that it has excellent performance in semantic understanding and reasoning calculations, just like the iPhone opened up a new experience of using smart phones. At the same time, customers have begun to enthusiastically seek services, “For large models, whoever uses them quickly and well can take the lead in the competition of peers.”

02 Save money, make money, all rely on AI

Since May, there has been a slight decrease in the new developments of large models of major manufacturers compared with the previous two or three months, and the market sentiment has also cooled down. After nearly half a year of exploration, the direction of the CEOs is gradually becoming clear.

When the AI painting industry was impacted by strong players, the commercialization of 6pen was affected. Wang Dengke said, “Last year 6pen was still making money, but by the end of the year, although users were growing, revenue was declining because many products in the market were free, and 6pen failed to find a point of differentiated charging.”

Wang Dengke said, “We don’t just want to make AI a core tool, but we hope to create more valuable content through it. For example, we are trying a ‘story universe’ project recently, which is a combination of human and AI. A platform to create stories together.”

He believes that in the field of AI painting, 99% of companies have no technical barriers. But this does not mean that the company has no barriers. This barrier may exist in product capabilities and the establishment of user communities. Product capability refers to whether you can find commercial scenarios that can be implemented, while user community refers to whether you can establish connections on the platform and generate content. These contents are not scattered and unrelated like AI paintings. Rather, it has storytelling and coherent content.

** Some CEOs of small and medium-sized companies with stable businesses have already realized the use of AI to reduce costs and increase efficiency. **

“It’s no exaggeration, at least 100,000 yuan in labor costs can be saved in one year,” YiYi said. Lianmuxu is rooted in the health food field and develops through multiple channels, but the daily operating costs are relatively high. With ChatGPT, it is equivalent to having a free GPT employee, which is equivalent to having “live broadcast operation + Xiaohongshu operation + community operation + private domain operation + copywriting editing” at the same time.

In the past more than a month, YiYi found that the solutions obtained through lazy and wish-like questions are often just general discussions. If you learn to ask questions by feeding data, you will gradually get customized solutions and ideas. By this time, some It is no longer necessary to find special interns to complete the work related to basic operations. The company’s original 10 positions have now been reduced to 6.

Xiaozheng told Shenran that it is precisely because of the remarkable effect after the training that the small team of two people has served nearly 20 companies in the two or three months after the Spring Festival this year, and one company charges about 30,000-50,000 yuan. The revenue is about 500,000.

In the process of providing training for enterprises, Xiaozheng found that just providing training cannot meet the needs of many customers. Therefore, Xiaozheng hopes to provide model capabilities for these companies in the next step, by using the interface of the existing general large models in the market , pre-debugging to make the model more in line with the needs of the enterprise, so that the technical difficulty is not too great, and the cost is not particularly high. Xiaozheng’s current training business can completely divert new business in the future.

**Many companies have begun to pursue the application of AI large models, or have put AI large models on the agenda as a new revenue-generating business. **

Hou Zhenyu mentioned earlier that by May, the rank of the person who found Wenxin Yiyan would be lower than in March, and he would no longer be the top leader of the company, but the person in charge of business or technology in each company. Started to really hit the ground running.

Xing Zai said that Singularity Cloud’s current strategy in AI is implemented. On the one hand, it needs to improve internal efficiency. For example, Code Review (code review) used to be done manually, but now the efficiency of combining machine review codes is higher than that of pure manual labor. In addition, the marketing department uses AI to provide ideas for copywriting and pictures, which is also a typical case of using AI to improve efficiency.

On the other hand, it is a product upgrade. In the field of large models, the current progress of the open source community is very prosperous. In terms of AI model exploration, Singularity Cloud focuses more on the direction of open source and single-card specialization models, hoping to provide privatized deployment in specific scenarios and directions. “The data in hand can be directly imported into a single-card specialized model without uploading the data to a general-purpose large model such as ChatGPT. This is the advantage of a data intelligence enterprise.”

After a series of tests since the end of last year, Tian Fengzhan mentioned that by fully integrating customer application scenarios, AI technology, and large language model technology, Tianrun Rongtong released a large language model vertical in the field of customer contact in April this year. Industry solutions to help customers reduce costs and improve efficiency.

Ye Jun, the president of Dingding, previously told Shenran that for some SaaS companies, it is also a very good industry chain business to make exclusive models for enterprises.

03 AI revolution has just begun

“This AI innovation is not a short-lived trend, nor is it a false proposition, but a deterministic thing, an epoch-making change.” A person with a technical background understands the specific aspects of this change, he said , “In the next three years, there will definitely be major changes in the industry, and it is still in the early stages.”

At present, the development of AI is still in a hazy period in most fields.

**In many small and medium-sized enterprises in the e-commerce industry, the role of AI tools has already appeared, and in the short term, the focus is still on reducing costs and increasing efficiency. **

“We are currently experimenting with the MVP version, but the implementation of AI requires long-term operation and maintenance, and it may be difficult to see the effect in terms of income in the short term.” YiYi said to Shenran. “In the short term, AI must focus on helping us save money. After reducing costs and increasing efficiency, we can use human brains to do more valuable things.”

Now YiYi is implementing the ChatGPT practical operation manual summarized by professionals in various fields, striving to have mature practical experience and reproducible model solutions in the vertical field of e-commerce + AI, and gradually turn the liberated manpower into productivity. It’s just that when AI can become a new business segment of e-commerce companies and bring revenue to the company, it is still actively exploring.

Wang Dengke believes that the impact of the AI wave is still more positive than negative. “It has heated up the entire industry. Although there are more competitors in the industry, the entry of a large number of users means more new opportunities.”

However, domestic exploration of large models is still in its early stages. Even large models such as Baidu’s Wenxin Yiyan and Ali’s Tongyi Qianwen are still in the internal testing stage. And these companies that have begun to explore small and medium-sized models have not really reaped the first pot of gold. **The industry is still waiting for AI to achieve the tipping point of large-scale revenue generation. **

Everyone in the AI revolution needs to follow the trend. However, middle-level executives or low-level employees are more involved in the practical operation of AI tools. YiYi also found that when communicating with peers, middle-level or low-level employees in general companies will pay more attention to the combination with this wave of AI trends. On the other hand, I also hope that there will be more bargaining chips for promotion and salary increase.

And CEOs are more involved in decision-making. CEOs in the AI field are anxious about how to build their own industry barriers in an ever-changing environment. Bosses in non-AI industries, although they have stable business and room for selection, can find suitable people to do this business, but how to combine AI with existing business is also a challenge. How to quickly sort out and make decisions is also a test for them.

Xiaozheng told Shenran that small and medium-sized companies tend to promote the popularization of AI from top to bottom. They already have mature business segments and are not too worried about being replaced by AI. Instead, they have greater pressure on peer competition. “Business owners will also come to us to inquire about the new progress of our peers in the implementation of AI.” Xiaozheng said that when people in the circle are accelerating the implementation of AI, those who walk behind will definitely feel anxious.

YiYi told Shenran that the idea of many entrepreneurial bosses is, “Why is the next employee artificial?” CEOs hope to have more and more AI employees to replace artificial labor. A senior director with a salary of 10,000 yuan does not need to pay social security and is on call 24 hours a day. As a boss, it is difficult to refuse.

But in any case, when AI is surging, CEOs must also run forward at top speed in order to reap the dividends in this wave of platform opportunities. The race has just begun, and the next competition will be more intense and cruel.

*The title picture is from Pexels. At the request of the interviewees, Xiqiao, Mumu, Lin Lin, and Li Yang were all pseudonyms.