This year, it is rare to see investors providing help in times of need. It is more like icing on the cake. In the field of AI, it is becoming increasingly difficult to obtain financing, and the path for entrepreneurs is becoming increasingly difficult.
However, there are some new phenomena that make us discover that the threshold for AI entrepreneurship is actually lowering. Super self-employed people are blooming everywhere in the wave of large models. They exist in every imaginable scene, such as AI smart plug-ins, AI cameras, etc. More self-employed people take advantage of the convenience of large models to start small-scale entrepreneurial activities.
What can be seen is that entrepreneurship in the field of AI is developing towards strong application and implementation-oriented development. This is also the same trend observed by Jishi. In AI entrepreneurship, many entrepreneurs in the AI industry and Jishi said that the prerequisite for starting a business is to land it first.
For Chen Zheqian, the founder of Yizhi Intelligence, who started a business in the field of human-computer interaction as early as 2017, feasibility is one of the main reasons why their products are welcomed by many brands.
When chatting with Shi Shi about the brand’s trends, Chen Zheqian also mentioned that brands are increasingly focusing on results and are strongly ROI-oriented. Correspondingly, their AI outbound calls and the newly launched digital people this year have also made corresponding changes. Changes, more and more AI applications are appearing in marketing scenarios.
01 Starting a business needs to be implemented
Let’s face it: Last year, the concept of the metaverse was very popular. Is AI this year just a hyped concept like last year?
Chen Zheqian: Actually, it is not that AI suddenly appeared today. On the contrary, it has actually experienced several waves. From my perspective, I have experienced three waves of AI boom.
In 2012, face recognition technology was very popular at that time, and people discovered that artificial intelligence was worth studying.
Then the emergence of AlphaGo in 2016 allowed people to discover the greater potential of AI. At that time, human-computer interaction became a craze. It was also during that period that various smart speakers and other human-computer interaction products quickly emerged. However, the technology was not yet mature at that time and was not widely used.
Yizhi Intelligence was boosted by this wave, because AlphaGo can be regarded as the starting point for the real craze of AI in the field of entrepreneurship.
So the big models started at the beginning of this year are the third wave in my opinion. Of course, if placed within the entire AI industry, it should be considered the fifth wave.
Jishi: From 2016 to now, what do you think is the most critical factor in AI entrepreneurship?
Chen Zheqian: My biggest feeling is that there are certain technical barriers to doing AI. This is not something everyone can do . If a high school student wants to study AI, it is actually impossible.
But AI is not so lofty that it cannot be reached. Some large companies will be happy to prove the superiority of AI and shape the concept of AI changing human life. However, at least judging from their financial reports, the income, financing amount and cash burning speed are obviously not equal .
Large companies still have the financial resources to supplement AI, but for entrepreneurs who only focus on AI, this is not a company that can feed itself or support itself, and can truly create commercial value for society.
I have always believed that AI entrepreneurship must be implemented. What I have always told my colleagues is that we are actually a business company. We just use AI technology to create a product. If we want to find a way to sell the product, the profit generated must be enough to support subsequent research and development and Sale.
AI has been a craze from time to time in recent years. With the craze, some investors will be willing to invest, so some people with technology will start a business and play with some concepts before the product is released. .
However, judging from this year’s situation, the probability of investors investing in the AI concept is very low. Everyone is cautious and begins to pay attention to the practical application of AI. They will not provide help when the time is right, but will choose to add icing on the cake. Therefore, for entrepreneurs who want to obtain financing, landing is even more important.
Ji Shi: So there is actually no need to badmouth AI. The most important thing is to have something in hand.
Chen Zheqian: Yes, I think a very important point is that why some people pessimize AI or turn AI into a big banner is not necessary. AI is actually a technology.
If you want to hype a technology very high, but if you don’t generate revenue, then the technology, no matter how awesome it is, will not have much value. In addition, technology has certain limitations. For example, large models allow today’s image generation to be very good, but large models are very slow in real human-computer dialogue. For example, if you use it to make some dialogues, it will definitely generate it sentence by sentence, and you have to wait for two or three seconds.
If in a phone call scenario, people would prefer AI to reply within 500 milliseconds, rather than having to wait two seconds to give an answer.
Technology will definitely become more and more powerful in the future. But as AI entrepreneurs, we need to find PMF and find truly commercializable scenarios under the current technical level.
02 Human-computer interaction – a rising AI scene
Jishi: How did you come up with the idea of making AI outbound calls?
Chen Zheqian: When I started my business, my goal was to create intelligent dialogues. Because the dialogue system plays a bit like a brain, it must have a carrier to be of value. Some people use dialogue systems in speakers, televisions, or cars.
In 2017, based on the overall level of the smart conversation industry, smart speakers, smart driving, and smart customer service were still just gimmicks. The robots at that time were not smart enough to make you think they were real people.
As I said before, entrepreneurship must consider feasibility, so I chose smart outbound calling because the phone itself is a very good carrier. Secondly, smart outbound calling only lasts a few minutes, and robots can make it in these few minutes. In order to achieve “disguising the real thing from the fake”.
When we studied the outbound call usage scenarios, we found that consumer brands need this AI outbound call function more. As a result, we have also created value for consumer brand customers. Our own revenue has also grown very fast, and our current market share is also good.
Jishi: Why do you value human-computer interaction AI more?
Chen Zheqian: Human-computer interaction can bring companies closer to users, and users will have higher recognition of corporate brands.
The iteration of technology itself is a process of shortening distance. In the past, you could watch movies from five to ten meters away, watch TV from two to three meters away, and watch your mobile phone within half a meter. After Apple’s Vision Pro, the distance between the glasses and the eyes may be only a few centimeters.
Hardware interactive devices are actually getting closer to people’s eyes. In the same way, the marketing method used to be the media, such as newspapers, which could only allow one-way interaction and no interaction. In the self-media era, you can click on comments and interact with officials using text.
Now, for example, you can watch celebrity live broadcasts, and if you post a barrage, you may be replied by the celebrity. AI outbound calls can make you more interactive with the brand. The reason why private domains are so popular is because it brings users closer to themselves.
In fact, we have been using technological means to bring enterprises closer to users, thereby improving users’ commercial value, or in other words, increasing users’ LTV.
Jishi: In the process of making outbound calls, have the brand’s requirements for your AI capabilities changed?
Chen Zheqian: There must be. What I can clearly feel is that this year, for example, many brands during 618 failed to complete their KPIs.
Brands now increasingly hope to directly see conversions, with ROI as a strong orientation, rather than planting seeds and waiting for them to sprout slowly. In the past, we would do some phone planting, but now brands want to see faster conversions. For example, after making a phone call, you can directly see whether the user has placed an order.
So now, the brand is not asking how much AI can help it reduce costs, but is more concerned about how much it can help it increase revenue. This year we launched the digital human tool, and we also paid attention to some changes in the brand.
At present, I feel that brands are actually paying attention to this kind of large model, and digital human live broadcasting. It not only helps brands save the cost of live broadcasting, it is actually a revenue-generating tool.
The reason is that with the same budget, you may only be able to build one live broadcast room in the past, but after joining Digital Live today, you still have the original budget, but you can build more live broadcast rooms, and the traffic that each live broadcast room may bring is one live broadcast room. The difference from ten live broadcast rooms.
03 Super individuals will definitely appear in AI entrepreneurship
See the truth: Are there any new positions created within the company?
Chen Zheqian: Yes, we have a new position in the digital field called digital human live broadcast operation . The task is to control the digital human anchor and let the digital human do a good job in live broadcasting activities.
What is different from traditional live broadcast operations is that traditional live broadcast operations review the situation with the host. Now it is him who reviews the situation with the digital people. He studies what other skills and reply mechanisms the digital people can use in the live broadcast room that can be upgraded.
In fact, there are many new positions on the market, such as AI training engineers. In addition, what I also want to express is that in addition to those entrepreneurs in the AI field who need to obtain financing, there is also an increasing number of “grassroots” AI entrepreneurs who segment themselves based on the large models of major manufacturers. Small businesses in small fields also exist.
From the perspective of a super individual, starting a business in the AI industry will become easier, not harder. This also means that based on AI technology, more super individuals, or one-person companies, can be born in the future.
Jianshi: It’s obvious that you are a technology optimist.
Chen Zheqian: Because you have to believe that technology will eventually bring ultimate power. Just like when we think back to more than ten years ago, when we did not have Baidu and Amap, we may not be able to imagine how everyone traveled without maps at that time. Ten years from now, we may not be able to imagine how we live our lives now.
Technology will definitely change a lot, but we can’t imagine it yet.
Jishi: So for the future trend of AI, what challenges will entrepreneurs face?
Chen Zheqian: I think there are three.
First, cash flow is the biggest challenge. Because cash flow depends on the first investor. Companies that rely on investor financing today rather than making money themselves will have an increasingly difficult time operating in the future.
The second channel of cash flow is customers, but today customers’ own revenue is still less than expected. Many external places will definitely take measures to reduce costs. This means that the challenge that still exists in the future is how to support themselves!
Don’t count on external blood transfusions. When external factors are so difficult, your own products, internal working environment, services, research and development, efficiency, etc. all need to be fully mobilized and considered. You must not only ensure that you can survive, but also ensure that your products are still good. high quality.
Second, peer competition will definitely become increasingly fierce in the future. Market competition, shopping malls are like battlefields. It’s just that as the economy goes down, competition becomes more intense. Everyone wants to survive, and no one wants to be the loser.
Start-up companies are inherently difficult. Every start-up company wants to survive, and if it wants to survive, it must fight for hard power. But some companies are not so ethical, and they will engage in price wars if they can’t compete.
Then merchants will fall into a dilemma of choice in the face of price and quality. So the challenge for everyone is: how do you stand out in such market competition and make users want to remember you.
Third, we must pay attention to internal management. In the past, some extensive management could be done, but in the current situation, refined management must be done. Salary is one aspect, it must balance the psychology of employees. Secondly, employees must be mentally motivated so that while receiving their wages, they can also feel that they have gained something and grown in the company.