In August 1956, a group of scientists represented by Shannon created the far-reaching concept of artificial intelligence around the theme of "using machines to imitate human learning and other aspects of intelligence" at Dartmouth College.
Unfortunately, in the later exploration process, AI players can't think and try like the original scientists. In the commercial society, it is difficult for industries without hematopoietic ability to thrive, and commercialization was once a "question of life and death" for AI.
Today, more than 60 years later, AI investment and financing have risen again, and players have gradually listed. While continuing to explore how machines can better blend in with human needs, have AI players found a path to sustainability?
AI bids farewell to the "classical era"
Since 2016, DeepMind's Alpho series of artificial intelligence has stunned the world, but for reality, an AI that plays chess and plays games that is unparalleled in the world has only played a role in igniting the fire of attention in the FIELD of AI. On the basis of no commercialization results, the tide of investment and financing has turned from prosperity to decline in two years, in addition to the head factories that are not afraid of burning money, the "AI Four Little Dragons" company represented by SenseTime, Megvii, Yitu and Yuncong is still trapped in realization.
After the AI boom subsided in 2016, "difficult to apply, difficult to make a profit" became the industry's leading thinking, a large number of small and medium-sized companies disappeared, and the AI business of large companies was once shelved. This short period of two years can be called the "classical era" of AI since it became the exact industry: the players' technical skills are amazing, and they may casually pull out a research and development personnel who are all phDs from some of the world's top 20 universities, and the papers have been issued N, but there is no way to make money, and there is not enough interaction with the actual needs of human beings.
However, some developments in the AI industry this year have brought different cognitions to the outside world.
Since the second half of this year, the AI industry has raised more than 200 financings, focusing on the application areas that combine AI with consumer home appliances and intelligent driving, such as technology and horizon; in early December, Meta (Facebook's parent company) announced that it will merge the AI team into the Reality Labs department responsible for XR software and hardware, so that AI and meta-universe business development directions will be deeply integrated.
(The AI business landing path in megvii technology's prospectus)
Most of the clues point to the fact that AI, after a series of tortures on commercialization, has begun to focus on the application layer rather than pure technological progress. In 2018, SenseTime CEO Xu Li said, "SenseTime is the world's leading artificial intelligence platform company, and will take technology as the core in the future to empower more industries. Now, down-to-earth empowerment integration has become the common pursuit of the AI industry, and AI is looking forward to bidding farewell to the "classical era" of relying on technology to get out of the circle and only earning fame and not earning money.
Different from independent AI startups, the top companies in some fields have more demand for the application of technology, and there are corresponding scenarios for use. Therefore, there are already more accurate examples of AI's external empowerment. Around the evolution of the relationship between AI and humans, the deeper "imitation" of human beings, AI should resonate with demand, never eat the technical highland of human fireworks, and lean over to a more grounded actual business.
AI commercialization should be good at "grounding gas"
Technology must penetrate deep into the "capillaries" of each industry in order to perceive the flow of demand and activate the vitality of commercialization. Many details are often overlooked, such as when we use Taobao to scan to find the same model, when users press and translate pictures on WeChat, when we shoot dynamic special effects and open live beautification in Douyin and Kuaishou, the executors behind the scenes are AI.
At the beginning of 2020, Twilio, the global leader in the cloud communications industry, once had a judgment: artificial intelligence and automation will dominate 2020. For Twilio's industry, AI can form a good interaction with a series of cloud communication services based on CPaaS, and improve efficiency in intelligent calling and intelligent customer service.
This is not a special case, on November 23 this year, Ronglian Cloud also released the "communication, audio and video platform + AI atom" capability at the "New Business , New AI" conference hosted by it. Gurudevan, Twilio's Chief Product Manager, said: "Businesses that implement transformation strategies by focusing on extending current systems and processes through AI, rather than replacing them, will be the most successful." "Including Ali, Tencent, Douyin, and Kuaishou, the mainstream company's approach is exactly like this.
So, how does AI systematically extend and empower the business processes of specific companies to achieve commercial value? This can be considered from the two levels of improving the business foundation + expanding the business boundary.
At the basic level, the role of AI is relatively easy to understand. For example, the details of Alibaba Tencent's products are often the optimization of user experience. AI can help the process of reading and translating maps to achieve a closed loop in one APP.
For Ronglian Cloud, which started with cloud communication business, its basic business is CPaaS voice and SMS, which needs to bear certain operator costs, and there are some difficulties in improving gross profit margins.
The focus of Yunlian CLOUD AI is on NLP and intelligent decision-making, and the effective mechanism of AI in a cloud communication is as follows: through fine-grained opinion mining, the integration of preset knowledge graphs, and a better understanding of what customers are saying and wanting. Artificial intelligence customer service is no longer a mechanical link of one question and one answer, and has a better dialogue strategy.
(Ronglian Cloud AI team's opinion triple group discrimination and dialogue display)
Therefore, those industries that attach importance to customer exchanges, such as banking, insurance, large consumption, automobiles, etc., will prefer to choose CC products of Ronglian Cloud, because AI helps them improve the efficiency of communicating with customers and ensure the user experience as much as possible. In Q3, the proportion of CC+UC&C business revenue with higher gross margin further increased to 63.9%, and the CC cloud contact center business increased by 125.4% year-on-year. This growth process itself is a process of commercialization of AI technology.
Twilio's view is also for similar reasons, when intelligent customer service is enough to cope with the ordinary needs of users, artificial force can be left to more refined and higher value-added needs. At the same time, it is more possible to expand the business.
For Twilio and Ronglian Cloud, AI customer service is just the beginning of the business process. Sun Changxun, CEO of Ronglian Cloud, regards the extension and evolution of Ronglian Cloud to CRM and SCRM as a core part of solving customer marketing problems.
Communication - customer acquisition - retention - improve satisfaction, in this service closed loop, AI to solve the first ring is very effective, but the first ring of value-added potential is not enough, integrated into the whole life chain, AI has higher value. Therefore, Ronglian Cloud not only has Ronglian Qimo, which is deeply rooted in the field of customer service, but also implants AI knowledge base and AI tag analysis in the SCRM of Ronglian Cloud Qimo Marketing Cloud.
In essence, the Ali of e-commerce, the jitter of short videos, and the Ronglian cloud of cloud communication are all focusing on user needs of AI technology. The theme of Ronglian Cloud's "New Business new AI Conference" is "AI has a heart, decision-making has wisdom", and the word "has a heart" actually points out that in the past, the superficial reason for the difficulty of AI commercialization was the scene, the deep reason was the lack of grounded thinking, and the AI living in the laboratory would never understand the fireworks of the human world.
Among the four AI tigers, Megvii Technology is the largest proportion of C-end business revenue, including mobile terminals, CIoT, and it is also a provider of equipment security and computational photography solutions for many mobile phone manufacturers. It has to be said that this is a definite practice of AI application.
In terms of development perspective, the end game of AI is the high integration of data and intelligence, it has infinite possibilities in the scene, and players with AI technology should not use the upper limit of imagination in advance to seal the pace of exploration to a wider range of application fields.
Digital intelligence, the end of AI scene application?
In the knowledge of "is the commercialization of artificial intelligence dependent on the B-end or the C-end?" In response to this question, one respondent mentioned a very important point: the difference between the B and C ends and the tradition is that the former considers whether it is useful and how useful it is, and the latter cares about whether it is easy to use. Another point is that don't just think about AI technologies, but think about how to do things with them.
(Zhihu answered by "Ling Bing" to the commercialization of AI in B and C)
We can draw a conclusion that artificial intelligence has always been the enabling existence in the mouth of senseTime CEO, and the Internet + big data + artificial intelligence should play an effective role in promoting the digital and intelligent transformation of all walks of life. Lei Jun said that pigs can also fly when standing on the wind outlet, and AI should not catch the wind outlet, AND should be the wind maker.
AI applications are deep mining of traditional scenarios. Today's smart wearables and smart homes are more intelligent in communication, no longer just a question and an answer. Huawei and Xilix launched the aito new brand of smart cars, shouted out the slogan of "Adding Intelligence to Auto", using harmonyOS system + operation, AI voice interaction, the intelligence "hit the public screen". In the hands of Tencent AI Lab, its game AI has been used in drug research and development, microscopy fields. BMW, Mercedes-Benz, and other top OEMs are training autonomous driving with AI and 3D engines.
This is an intelligent use of large-scale data, while at the same time feeding back more effective test data. The more it connects with reality, the more malleable AI becomes.
The same is the intelligent customer service, Ronglian Qimo's focus can be as high as IQ, emotional intelligence, knowledge, because Ronglian Cloud determines that AI should be a tool to help enterprises transform digitally, and its core should be people. How to better connect with people is the area where Ronglian Cloud is strong and good at, and in addition to the field of Ronglian Cloud, more players are also trying to create new experiences. For example, from Baidu's independent operation for more than one year, Xiaodu has created a fitness mirror explosive product with a unit price of more than 4,000 yuan based on AI technology.
Eric Schmidt, who once ran Google, believes that this is the golden age for investing in artificial intelligence. UBS's advice is that ARTIFICIAL intelligence is the top investment theme for the next 10 years.
The reason may be understood from this perspective: At the "New Business, New AI" conference, Sun Changxun, CEO of Ronglian Cloud, said that Ronglian Cloud's mission has always been to "improve the communication experience and operational efficiency of human organizations". This will allow AI's data and intelligence to be fully released. AI players continue to pursue excellence in a human-centered manner, which will create a more commercial future worth looking forward to.
Commercialization will eventually form a civilization, and the core characteristics of civilization are self-improvement and evolution, at least in AI-related fields, we see this dawn.
epilogue
Kevin Kelly, author of "Out of Control" and known as the "Internet Prophet," once said that the most profound technologies are those that are invisible, weaving themselves into the details of everyday life until they become part of life.
This may be an advance summary of the direction of AI to lay down the application trend. Once upon a time, when AI was at its brightest, its impression in the minds of the public was a series of tall nouns. But today, AI has almost done what it calls for in life, and it lingers. When AI comes to life, its value can be obtained, written more deeply.