Shulou(Shulou.com)06/03 Report--

The historian Joseph Needham likes to use his old profession, biochemistry, to interpret the sociological problems of grand narrative, and introduces an interesting term-"titration" (titration).

The so-called "titration" is to drop the reagent in the solution and observe the thing through the color change of the indicator as neutralization occurs. In Needham's eyes, the exchange of different civilizations is also a "titration" process.

This is a figurative metaphor to explore the infiltration and integration of different cultures. From a certain point of view, the matter of "AI fighting the epidemic" is also a kind of "civilized titration":

First, computing science and modern medical prevention and control system are more closely linked than ever before. AI, a group of "reagents", appears in all aspects of the anti-epidemic process, revealing the scientific and technological fundamentals of the whole society.

On the other hand, the virus is essentially a "background murmur" in human social life, and the fight against it may come to an end in stages, but it will never end, which means that science and technology and the fight against epidemic disease must be tied together for as long as "quantum entanglement" until a "neutralization effect" is produced.

Then, recording and listening to this epic of AI's anti-epidemic has become the only way for later generations to seek.

"Needham problem" in AI Anti-epidemic 1.0

Now that titration is mentioned, we might as well use it to answer a "Needham-like puzzle": why did not such a large-scale AI participate in the governance of social and public affairs before?

In the article "the first large-scale AI epidemic fight in History", the brain Polar body has interpreted the various postures of AI's combination of fighting the epidemic, and it is not difficult to see several trends:

1. Probe into the sense of participation.

The so-called "probing up" means that AI has got rid of the position of "icing on the cake", opened up the transformation from "tool man" to "think tank", and demonstrated a strategically advantageous value in large-scale epidemic prevention work. This is the first time that technology has shown a "sense of participation" in public affairs.

The most obvious manifestation is that virus screening and drug research and development, which determine the progress of treatment, are accelerated with the help of high-performance computing. The combination of medical experts, science and technology enterprises and super computing power has never been strengthened before.

Tencent, Baidu, Ali and other enterprises have also opened their own basic algorithms and 100 million-level computing resources to compete for precious time to fight the epidemic.

For example, Baidu's free and open linear time algorithm LinearFold, combined with the world's fastest RNA structure prediction website, shortened novel coronavirus's genome-wide secondary structure prediction from 55 minutes to 27 seconds, 120 times faster, saving two orders of magnitude of waiting time.

two。 The sinking of popularity.

The value enhancement of AI is not only reflected in becoming a sharp sword in the hands of scientific researchers, but also frees front-line grass-roots prevention and control personnel from a lot of repetitive work to achieve efficiency upgrading.

Take the telephone census, for example, this is an extremely simple and stressful task, and it is difficult to rely on the existing outbound call centers and volunteers for the investigation of hundreds of millions of people. Relying on AI intelligent outbound calls to automatically inquire and collect residents' epidemic situation information plays an important auxiliary role.

In the CCTV News broadcast on February 9, Baidu's free and open intelligent outbound platform attracted the attention of the public because it improved the efficiency of community epidemic prevention and control.

Similar stories can be seen everywhere in the "gridded" anti-epidemic ports, whether it is replacing the temperature gun in the hands of community personnel with intelligent temperature testing, or the standard wearing of face masks identified by AI, all showing that AI is sinking to the base of the whole society in a silent way.

3. Awareness needs to be improved.

Does the rapid deployment of AI capabilities mean that a high degree of recognition has been formed? The answer may not be satisfactory.

It can be seen that the first phase of AI epidemic prevention is mainly led by technology giants like BAT, which not only gives AI a more complete support for epidemic prevention, but it is difficult to quickly infiltrate into the scenarios and needs of thousands of industries, and the latter obviously needs to cooperate with many vertical developers and medical front-line personnel in order to maximize the potential of AI.

However, many industry personnel are not familiar with the ability, algorithm and function of AI, which limits the contribution of AI in this trial.

Summing up the harvest of AI anti-epidemic stage 1.0, we find that in such a large-scale public affairs governance, human society has ushered in a new choice, a new way.

If the medical staff are using their knowledge and life to resist the progress of the disease, then AI is devoted to the work of constructing and improving the prevention and control efficiency of the whole society, giving birth to the combat effectiveness of long-term combat.

What preparations should be made for anti-epidemic 2.0:AI?

There has been a little rest in the high winds and torrential rains of the epidemic prevention war, and those who have not been hit by the virus in this fight are about to join in the real life of resuming production.

Behind this, there will be a series of intertwined, complex and long-term practical problems-orderly rework, resumption of work and study, production prevention and control. This also indicates that the epidemic prevention work has entered the 2.0 stage of preventing secondary spread and backflow.

What does AI need to do well in this long-term "arms race"?

In order not to talk on paper, we might as well take Baidu, the technological development representative of China's AI industry, as an anchor to see what roles platform enterprises need to play if AI wants to accelerate in "epidemic Prevention 2.0".

1. The preacher of technology.

In the previous stage, we saw the value output of AI in the core region. How to deepen the AI enabling scene and play a greater significance, we need to activate AI developers, researchers and users from all walks of life to jointly explore the special scenarios in the anti-epidemic needs and realize the value multiplication of AI. This requires platform enterprises to empower developers with an open attitude and jointly use AI to solve those deepening subdivision scenarios.

Baidu, which has the largest open ecology of AI in China, has launched an initiative to actively mobilize developers and launched the "War Plague" defense program for AI developers, providing free technical services and online AI training courses to enterprises, institutions and developers in need during the epidemic.

The AI mask recognition model mentioned earlier was developed by engineers related to Baidu brain's EasyDL customization training and service platform, and the staff also led the students to train the model in detail during the live broadcast class on Feb. 12.

At the same time, the model is also free and open source on Baidu Flying Propeller, and developers and enterprises are able to quickly build personalized monitoring programs. Presumably in the future public places, we will have the opportunity to see that the use of AI to efficiently detect whether masks are worn correctly will become a favorable weapon to ensure traffic efficiency and reduce the risk of human cross-infection.

two。 The flag bearer of the economy.

In the face of the needs of many industries and parties, it is obviously not realistic for AI to protect the orderly "restart" of the society, and it is obviously not realistic to rely on the hands-on efforts of several major science and technology enterprises. How to quickly achieve the release of AI capabilities, cloud services become the best port.

At present, we see that Baidu Smart Cloud has created a number of AI products and solutions to provide technical support for rework and production in various industries.

For example, it provides comprehensive technical support for the national primary and secondary school network cloud platform to help primary and secondary school students to "suspend classes without school suspension"; Baidu Hi enterprise intelligent telecommuting platform is also open to the public to help enterprises quickly restore their production capacity through telecommuting coordination functions such as high-definition voice telephone, enterprise cloud disk encrypted transmission, web video conference, etc.

In labor-intensive factories, Baidu also uses cloud services to innovate with partners to solve the problem of workers being unable to return to the factory. For example, the "surface defect visual inspection equipment", an intelligent automatic testing equipment jointly built with micro-billion Zhi Chuang, can accurately detect more than 30 kinds of defects on 6 sides of parts in one second to ensure that the factory can start work smoothly in the case of understaffing.

In addition, Baidu Intelligent Cloud has also launched related services to reduce costs and increase efficiency for physical enterprises that have already started.

For example, the combination of big data, Internet of things technology and artificial intelligence analysis and forecasting algorithm to achieve electricity tariff optimization, reduce personnel daily energy copying costs, through the rational use of energy to speed up the response of market competitiveness.

In short, it is urgent to use AI to defend the economic war in the next epidemic prevention cycle.

3. The defender of mankind.

In the front line of epidemic prevention, we see that self-driving and robots play an important role in safety protection instead of manual operation and prevention of cross-infection.

So in the face of larger-scale personnel mobility and higher operating pressure in the 2.0 stage of epidemic prevention, can unmanned go further?

Baidu self-driving platform Apollo teamed up with partners to fire the first shot in machine combat.

During the epidemic, self-driving operating vehicle companies aiming at anti-epidemic scenarios can get free support from the self-driving cloud service matrix provided by Baidu Apollo platform.

For example, Zhikai, a Baidu Apollo ecological partner, was able to quickly deploy unmanned cleaning and disinfection vehicles / unmanned distribution vehicles to 16 key COVID-19 patient treatment hospitals across the country.

In addition, Apollo will also provide a series of software, services, hardware and landing financial support for the research and development of unmanned equipment, so that people who go to work and on the street feel a little more at ease.

The Paddle of Technology and the other side of Humanities: on the basic cultivation of AI Enterprises

From the cruel background of the epidemic, we also saw the courage and responsibility of those heroes and enterprises.

A question worth pondering is why Baidu can respond quickly and play these three roles, exporting its technical capabilities in both epidemic prevention and economic warfare. What are the characteristics worth digging into?

I think there are three essential points:

1. Technical belief.

In the contact with Baidu people, it is not difficult to feel the respect of Baidu culture for scientists and technical experts from various details, which makes Baidu represent the top strength of China's AI industry on the technology dome of AI. At the same time, Baidu is also in awe of the needs of developers and users. Many partners have told us that after they have ideas and needs, Baidu's technicians act quickly, cooperate selflessly and support them at any time. This pursuit of Hainan's hundred rivers and technological beliefs enables Baidu to gather vertical models of many subdivided scenes, so it can also reach the front line of anti-epidemic as quickly as possible.

two。 Big data accumulation.

We know that in order for AI to play an accurate and efficient role in grass-roots practice, in addition to model performance, we also need to use relevant data sets to adjust in order to ensure availability.

The accumulation and timely transfer of big data has become the basic guarantee for Baidu AI to play a rapid role in fighting the epidemic.

For example, to use AI to track the history of epidemiology, you need to call big data in transportation, payment, travel and other aspects to analyze, and then establish an infection model.

In the anti-epidemic process, Baidu map migration big data platform continues to upgrade, users can intuitively and clearly grasp the national hot routes, popular destinations, popular departure places. This is closely related to his space-time big data and his ability of analytical technology.

3. Industrialization vision.

Another difference of Baidu AI in epidemic resistance lies in the satisfaction and synthesis of many complex scenes.

On the one hand, this stems from the multi-domain coverage of Baidu's own business scenarios, such as search, map, input method, translation, self-driving, intelligent voice and so on. These services have been supported by Baidu AI technology and have become a window of value for rapid release to the whole society.

At the same time, with the help of Flying Propeller, Apollo, Baidu brain and other technology open platform, Baidu industry users can also quickly find the required AI capabilities.

So far, Baidu brain has achieved a variety of AI capabilities, such as open voice, semantics, images, video, knowledge graph, and so on. At the hardware level, it provides powerful computing power of millions of tops through the new AI computing architecture. This also makes the whole society to complete a Baidu as the center, software and hardware coordination of the "AI Big Bang".

From the case-level AI to the real industrial intelligence triggered by Qianhang Wanye, Baidu has led intelligent technology to a new industrial cycle, which has led to a clear distinction between Baidu and other technology companies.

It may also explain why AI can have an unprecedented titration with human society at this time, this moment and this scene, and will affect the long-term future.

Finally, I would like to talk to you about the relationship between plague and human civilization.

In many reports, from time to time, people compare the COVID-19 epidemic with the 1918 pandemic.

If there is any similarity between the two, one is that human beings must accept the reality of a long war against the virus, just as 100 years later, mankind has not completely eliminated the influenza virus, a new pandemic is almost inevitable. As one flu expert said: only hear the clock tick, but we do not know the time.

The second is to re-examine the technical tools in hand. Just as 1918 was the first great confrontation between modern science and the forces of nature, it also established a modern medical system for the whole society. in order to prevent further misfortune, in addition to relying on faith to face the challenge, let the "titration of technology" again and again, to achieve the reconciliation of the whole society and disaster, and then usher in the next prosperity.

Before the end of the battle, the epidemic is not just about destruction, death and rebirth, it is also a story of science and exploration, a story of how humans should look at and apply technology to their way of thinking.

To be sure, Baidu will still promote AI, so that the game time between humans and viruses can be earlier, easier, and win faster. At this point, Baidu's continuous hard work and little by little progress in China's AI industry is a blessing for everyone.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.