Unfortunately, because your browser version is too low to get the best browsing experience, it is recommended to download and install Google Chrome!
At the end of last year, "a screen that changed our destiny" became the topic of screen brushing.
At that time, the supporters and opponents listed a lot of opinions, but we may admit that these discussions were based on the premise that the combination of new technology and rural life has started to trigger some changes.
If you look at this vast market in China, you will find that from agriculture to rural markets, all kinds of services, and then to the macro trend of rural labor transfer, too many of these needs can be filled by scientific and technological forces.
The relative lack of scientific and technological talents in the rural market has made intelligence to a certain extent a rigid demand different from urban society.
If we look back at 2018, we will find that from the second half of the year, technology giants have laid out the exploration of AI's entry into agriculture, and various emerging products and services of AI medical and AI education have begun to move to the new test field in the countryside.
The story of AI and the countryside is rapidly heating up in a short time. However, behind the lively layout, we will also find some clear bottlenecks that run between the rural market and the imagination of AI.
Bypassing the controversy on the screen, more AI stories are being staged in the mountains and fields. Regardless of whether 2019 can be called the first year of "AI + rural", at least this year, the field's AI story will inevitably rise by a surprising percentage.
Let's recall how, to this day, AI has accomplished the task of going to the countryside and entering the village.
Agricultural round dance of tech giants
The so-called AI agriculture is easy to understand in terms of technical logic, that is, to use the physical recognition and machine vision capabilities brought by AI, combined with data analysis technology, to re-optimize a large number of processes in agricultural production , so as to intelligently improve agricultural production efficiency and optimize Agricultural product quality.
Theoretically, this logic can grow both food and vegetables, and pigs and geese. But it is not easy to operate in practice. On the one hand, agricultural data is relatively scarce and the degree of standardization is very low. On the other hand, related technical equipment is almost blank. AI agricultural propositions are not only algorithms and data problems, but also a test of engineering capabilities and hardware manufacturing capabilities.
In the eyes of AI holders, such as technology giants represented by BAT, since it is necessary to enter the industrial AI and industrial Internet, agriculture is an option that cannot be bypassed. Its huge market potential and social value are cakes that technology companies cannot give up. In the opportunity of the full-scale launch of industrial AI in 2018, agricultural AI also began its own story.
On June 7, 2018, Alibaba Cloud released the ET Agricultural Brain at the Yunqi Conference and Shanghai Summit. Through the combination of digital file generation, intelligent agricultural data analysis, and traceability of agricultural products , AI Cloud began to bring AI solutions into agriculture.
In the following six months, both Tencent and JD announced their AI agriculture plans. It is believed that Baidu, which is good at AI, is already on the way.
Taken together, the AI agriculture proposition now has two main paths: AI farming and AI cultivation.
Speaking of breeding technology, our Chinese housekeeping skill is to raise pigs. Many people may not realize that the Chinese have written a magnificent epic on the road of large-scale and technical pig farming. It is also because of the high level of standardization of the pig raising business that it is very sensitive to new technologies. Most of the technology giants play AI + breeding, and they all start with pigs.
Ali's ET agricultural brain uses the AI camera and data analysis capabilities of machine vision to observe the growth data of pigs, so as to achieve survival of the fittest; and the voiceprint recognition and infrared temperature measurement are brought to the pig farm to pass the pig's body temperature With the sound of AI to predict the pig's physical condition, to achieve the effect of improving sow litter production and reducing mortality.
In November last year, JD.com also began to depict the romantic story of AI and pigs. In addition to accessing AI cameras and data intelligence systems, JD.com's plan also includes IoT systems, self-developed farm inspection robots, feeding robots, etc., and uses new "pig face recognition" technology.
There is reason to believe that more tech companies will start their AI pig careers next.
And AI has its own set of vegetables and vegetables. Ali's ET agricultural brain has completed cooperation cases on melon and lettuce. Last December, the "Cucumber Planting" of the Tencent AI lab team won the first place in the "AI Strategy" and the second place in the "Autonomous Greenhouse Challenge". This is also considered to be the beginning of Tencent's march to AI agriculture.
The “Growing Cucumber” exhibited by Tencent is special in that it uses the reinforcement learning algorithm to sneak the expert knowledge system into the simulator, so that the agent can effectively learn the human expert's thinking mode, and then return to the actual cultivation to increase the cucumber yield , And the cost of the sensor has been compressed, improving the practicality of the technology.
The AI planting industry is currently concentrated in orchards and greenhouses. By collecting data and intelligently identifying the plants, it is possible to judge the suitability of conditions such as fertilizer, moisture, temperature, and light, so as to make the extensive planting model intelligent. Coupled with some traceable and live broadcast Internet games, healthy + high-yield AI fruits and vegetables were born.
Being able to raise pigs and grow vegetables, I feel that AI is already very tasty in the agricultural field.
But don't be too optimistic. At present, the AI agriculture journey of the giants has just started. Today, all kinds of cases worthy of boasting are still demonstrating much greater value than commercial value.
On the one hand, agricultural data is still scarce today, and agricultural AI still needs BAT experts to go down to the field to collect data and modify parameters. On the other hand, a large number of agricultural areas and agricultural fields are vacuum zones of data. Today, AI wants to enter the agriculture, and it can only rely on some agricultural sectors with a high degree of data standardization, such as modern farms; or rely on related agriculture Collaborators in technology accumulation, such as large agricultural groups, come to complete more 1: 1 commercial experimental AI + agricultural cases.
Compared with China's vast farmland and pastures, BAT experts are obviously not enough.
Therefore, today's AI agriculture is still in the stage where the technology giants do a good job, set a good example, and then attract powerful partners to promote it together. Large-scale agricultural transformation is far from enough. Only when a fool-like product and an AI agricultural solution that can be sold on the terminal are formed; after the industry middle layer between the tech giants and farmers has established a clear industrial chain, the phrase "AI changes agriculture" has the confidence to be said.
In any case, when we see data experts and algorithm engineers in the buildings of Beijing, Shanghai, Guangzhou, and Guangzhou, and squatting in the fields to observe the growth of crops, we can still see a tense story.
New role of AI in rural society
In addition to industrial empowerment, AI technologies for social services are becoming more and more popular. AI medical, AI education, AI finance, and AI government services are all becoming new technology tracks.
However, it is necessary to pay attention to such a logic. Unlike mobile Internet integration of services, AI + social services solve the problem of unmannedness. By replicating and redistributing machine learning from human experience, AI can replace the work of some professionals to some extent. For example, AI voice interaction replaces teachers, machine vision equipment replaces doctors for medical image recognition, and so on.
Such capabilities may only be seen as substitutes and efficiency tools in regions where talent is saturated. In rural areas, the problems of presence and absence are likely to be solved.
The provision of medical, educational and other services based on AI capabilities in rural areas continues to grow today. For example, Ali based on its smart speaker Tmall Genie launched the “Tmall Genie Station” program last year. The program provides more educational resources for children in rural areas by establishing a Tmall Elf Station Library. In areas where early education talents are relatively scarce and teaching aids are scarce, access to smart speakers is a solution.
In the medical field, more cases of AI entering the village can be observed. For example, last month, many media reported on the trip to the countryside by the Baidu Lingyi team's AI fundus screening machine.
Through the training of visual recognition algorithms, Baidu has created an AI-based fundus screening device that can effectively identify early fundus lesions such as "sugar nets". In contrast, there are not many doctors in rural and township areas with accurate fundus screening capabilities. Complex ophthalmic diseases can only go to major cities such as provincial capitals, and early diseases are more likely to be ignored.
With the addition of AI, this long-standing problem will hopefully be resolved. Because AI devices replace not only machines, but also the judgment and recognition capabilities of the doctors behind the machines. This is invaluable for rural areas.
Similar cases currently occur mainly in medical image recognition and laboratory testing. There is reason to believe that in the near future, AI will help bring remote clinics and even remote surgery.
As with the introduction of AI into agriculture, the problems of AI rural medical care and AI rural education are still at the initial stage. Similar cases today are more in the category of corporate public interest. If we do not solve the problem of commercialization and promotion, then we will always see that AI has given love again, not that AI has really changed the lives of most Chinese groups.
Rural labor and AI infrastructure
Another combination of rural areas and AI is not what AI has helped the countryside, but the other way round-the labor cost advantage in rural areas is becoming a kind of fuel for the development of AI.
Last year, many media started to report on keywords like AI Village and AI Rural Factory. This type of rural factory business model is that AI requires a large amount of training data, which is mainly based on picture data. The data indicates that this almost thresholdless job, which is relatively easy to work compared to going out to work, has an extremely high repetition rate, and is almost indispensable, is continuously moved down until it is transferred to the village to start work.
So someone said with a smile, you found that mobile phones and tablets can recognize flowers, birds, fish, insects, etc., it feels very tall, I don't know that it was taught by your second aunt mother in your hometown.
When the AI village appeared, some people thought it was ridiculous, and some people lamented that "there is no intelligence without human labor." However, from the perspective of industrial structure, a large number of necessary tasks in the development of the AI industry are indeed common sense and can be outsourced. And this kind of work will spontaneously find a production place with low labor costs. Then the vast rural labor force has naturally become the primary choice.
Objectively speaking, the combination of rural labor cost advantages and AI will not die out soon. In this kind of outsourcing work, image recognition is still the main body today. However, with the deepening of the data and AI industries, various data-related tasks will flock to the outsourcing market, such as data cleaning, data sorting, and processing of data sets in vertical industries.
For AI, these tasks are essential, and for rural areas, they mean that they can be done relatively decently in front of a computer without having to stay away from home.
It must be noted that if such AI villages and AI rural factories do not actively seek the gradual upgrading of their data processing capabilities, and rely only on common sense + image sorting to maintain their livelihoods, the market will soon be under a lot of competition. Dry out completely. Only by entering the vertical industry, practicing relatively stronger data operation capabilities, and possessing data confidentiality capabilities that can persuade the market, can we obtain long-term survival rights in this emerging outsourcing market.
Anyway, if we think that it is cool to find algorithm developers and security engineers through hackathons and algorithm competitions, then it seems that there is nothing worth talking about finding rural aunts doing data cleaning through outsourcing. of. In fact, both follow the same law of value: meet the needs of emerging industries and give play to their own market positioning advantages.
It is not difficult to see that the story of AI entering the village is undergoing a vigorous growth process today.
Giants are occupying places, agricultural enterprises are iterating on their own; public welfare is glowing, and entrepreneurs are beginning to turn around; algorithm engineers have entered the shed, and villagers in the village are educating AI systems across the ocean.
The change has just begun, but the change has begun. These tension pictures may be where the Chinese AI is deeply affected today.
How to adapt to such changes? Perhaps a good solution is to think about the industrial needs and market needs of AI from the perspective of individual occupations, and then take patience slowly. No food can be seen in spring, but it must be sown in spring.