In the fight against the epidemic, the government, organizations, and individuals have made every effort to respond. In the past, AI, a symbol of cutting-edge technology, will also be invested in the epidemic.
This may be the first article in history to combine AI technology with the fight against the epidemic. This new type of coronary pneumonia in China, which is completely blocked, is also the first time that humans have applied AI technology to large-scale applications due to the special nature of time and space. Public health events.
At this moment, in the laboratory, hospital, public transportation hub, and community, a variety of AI technologies and products are performing their duties and racing against the epidemic.
Let’s take a look at what role AI technology has played under the epidemic; what are the implications of these effects for the AI industry, the medical industry, and even the community.
Objectively speaking, AI technology only serves as an aid in the fight against epidemics. But maybe we will ask such a question at the end of the epidemic: After this test, can we and should we do more to develop the intelligent process in the field of public health and medical research? If the answer is yes, how should we find the future in lessons and experiences?
The history of modern medicine generally believes that the Spanish pandemic in 1918 prompted humans to complete the construction of a modern public health protection system. Then from the performance of AI on the battlefield of epidemic resistance, we may start to think more.
I. Viral analysis and vaccine development
After the outbreak, one of the first announcements by major cloud computing vendors was to open AI computing power to scientific and medical institutions free of charge.
At that time, many people were curious, what is the relationship between AI computing power and epidemic resistance?
Here we will mention a certain nature of AI computing: tensor computing is used to handle unstructured data matching.
In the classical computing environment, many unstructured data processing, such as image recognition, speech synthesis, gene matching, geological information calculation, etc., cannot obtain high-efficiency calculations.
This makes AI computing require a separate computing chip and computing architecture, so that AI computing power has become a key track for chip and cloud service vendors in recent years.
The industrial accumulation of AI computing power can come in handy when the epidemic comes and needs to accelerate medical analysis capabilities.
In the field of medical analysis today, viral gene sequencing, protein target screening, and the matching of historical data of virus and drug development, all kinds of work require the support of AI computing power. At the same time, better related algorithms can greatly improve the efficiency of related detection.
These work are extremely important basic work for us to understand the nature of the virus, analyze better treatment options, and develop vaccines and targeted drugs. The value is self-evident.
Although the role of AI computing in virus analysis and vaccine development is only to shorten the matching cycle and improve the detection efficiency, and it is not possible to complete the vaccine development on our own as we imagined, but the open and abundant AI computing power is in a race against the epidemic Next, it is also very important.
As we have seen so far, all major cloud computing vendors are facing the need for epidemic resistance and have opened up AI computing power for free.
Some science and technology companies have also made genetic testing-related algorithms freely available to genetic testing agencies, epidemic prevention centers, and academia, shortening the genetic testing time for new coronaviruses.
In this anti-epidemic operation, the virus can be quickly isolated and related gene sequencing completed, and the public can see that the research information of related therapeutic drugs is coming out quickly. The industry’s accumulation of AI computing power and algorithms has played a significant role.
II, diagnostic assistance
From scientific research institutions and laboratories to the front lines of the battlefield against epidemic diseases, AI can also perform a number of key functions. Among them, the most valuable for the current anti-epidemic, one is a robot with visual recognition and voice interaction capabilities, which replaces medical staff for patient care.
The use of medical robots in the United States for the treatment of new pneumonia has also caused discussions in China not long ago.
However, medical robots need relatively mature industrial support, and it is difficult to get started quickly.
In China, AI robots have been used in hotels to perform conversions, serving as part of drug delivery and medical supplies transmission tasks. On the front line of anti-epidemics, more helpful AI capabilities are diagnostic aids. This point is currently focused on medical imaging + AI analysis capabilities.
Within a week of the outbreak, the medical imaging capabilities of many AI technology companies in China began to be stationed in major hospitals, providing doctors and patients with intelligent systems based on medical image analysis.
Although the diagnosis of new pneumonia mainly requires detection supplies such as reagents, the patient’s lung image also has relatively strong discriminable features.
Based on AI technology, traditional inspections that take hours can be compressed in seconds. This ability effectively aids reagent testing, and helps to quickly diagnose and fill medical staff shortages.
It is foreseeable that the auxiliary diagnostic capabilities based on AI medical images will gradually move towards the front line of epidemic resistance in the next few days. Major AI companies have urgently strengthened their product capabilities in this area and are also working more closely with medical research institutions.
III, intelligent temperature measurement
With the arrival of the return tide, public places such as airports, stations, and highway junctions have become key points for preventing and controlling epidemics. And the long line of temperature measurement in these places has also become helpless in special periods. But this long wait will also cause large-scale crowd gathering, which obviously also brings risks to epidemic prevention.
In recent days, you may find that intelligent temperature measurement without waiting or removing masks is enabled in many places, and people can pass without feeling.
In similar systems, AI is an integral part.
First, the AI needs to lock the face without removing the mask to accurately match the detected person with the detection data;
Secondly, it is necessary to identify and track the body shape, and compare it with the body temperature threshold transmitted by the sensor to provide early warning to those with abnormal body temperature.
With infrared and visible light sensors, AI temperature measurement has significantly improved the passing efficiency of public places.
Judging from the related products currently used in various places, a single device can guarantee a detection pass rate of 10-20 people per second, and it has clearly equaled the normal flow efficiency of pedestrians at the railway station, airport, and subway.
In public places, AI’s protection against epidemic situations is also manifested in the field of public security.
For example, an AI camera can use face recognition technology to determine the trajectory of a person’s public places. This technology has not only improved the level of public safety in recent years, but has also repeatedly performed wonders in this epidemic prevention.
For example, the history of infection, which is only ten seconds, and the parties are completely unknown, can be traced back to avoid the large-scale fission transmission of the virus in unexpected circumstances.
The combination of public health safety protection and AI technology is redefining the balance between safety and efficiency. It is believed that the application of AI during the epidemic will change the long-term construction of the public health security system thereafter.
IV, Smart phones:
Another unseen area of prevention and control took place over the phone.If you have experienced going out during the epidemic, then you may have received a call from the AI_ It will ask you about your going out, the time and place of travel, and your physical condition after returning to your place of residence. The outbound call systems come together and become the data base for basic-level epidemic prevention and control.
Facing the huge population scale and the scale of travel during the Spring Festival, the census and epidemic prevention notices of communities and grass-roots residents have become a huge challenge in the actual implementation Uninterrupted calls are obviously not appropriate.
So calling this repetitive labor becomes a task that the grassroots can’t afford and must complete. In this case, the relatively complete intelligent customer service system has become the key to breaking the game.
The AI customer service system based on intelligent calling and voice interaction can be transformed into an intelligent inspector in a short time, completing personnel inspection, return visits, notifications and other matters, which is hundreds of times more efficient than manual calls.
Furthermore, some smart phone systems can also complete relatively complex epidemic prevention inspections and return visits, such as randomly surveying users and surveying their living conditions to form sampling statistics; for example, ontinuous follow-up visits to key groups to form a focus Prevention and control system.
At the same time, medical management units and grassroots units at all levels can also develop their own more targeted smart phone systems based on the ability of smart phones that are currently generally free and open, making AI a part of social care and social bonds in special times.
Looking back at the combination of AI and anti-epidemic work, it can be seen that, unlike other technologies, AI can enter the core of anti-epidemic work at all levels.
For example, it is not difficult to find that the Internet can only solve the role of information transmission. Although this role is irreplaceable and very important, it cannot directly accelerate virus analysis and vaccine development like AI.
The core working layers in all fields can be covered by AI. This is a prerequisite knowledge that can help the society to form AI technology under the anti-epidemic examination.
At the same time, we must also see that the basic ability of AI technology is to improve industrial efficiency and replace part of redundant labor, not human labor.
In other words, AI is only an aid in the fight against epidemics, but it is a very important aid and accelerator. In general, we can see that in the epidemic prevention and control scenario, AI can play a role in three conditions: Labor efficiency is slow, and there is an urgent need to improve efficiency. Such as intelligent body temperature detection in public areas.Massive repetitive labor. Such as AI telephone surveys and notifications. Fuzzy data is difficult to accomplish with classic computing models. Such as AI virus gene sequencing.
These three AI working characteristics have actually been repeated many times. The reason I want to emphasize it here. I hope to call on AI developers from all walks of life, as well as developers and manufacturers in the fields of medical products, genes, and robotics, to re-emphasize the basic capabilities of AI, to find special scenarios among many anti-epidemic needs, and to exert their open capabilities and wisdom.
The current situation is that we can see that major AI companies and cloud computing vendors are rapidly invoking AI capabilities to develop related products, and are putting them into the battlefield against epidemic diseases for free. However, the coverage capabilities of only a few leading companies are obviously not enough. A wide range of developers must be involved to maximize the value of AI technology and enter the 2.0 stage of AI epidemic prevention.
At present, AI platform companies have also begun to invest in anti-epidemic battlefields with single AI capabilities. At the same time, they focus on empowering developers, letting everyone use the open and technical models to find anti-epidemic scenarios, solve long-tail needs and improve The overall efficiency of the epidemic.
This process may require the joint efforts and efficient communication of Internet companies, AI developers, and medical and scientific researchers.
Today, many AI developers are eager to try, but they are not clear about the needs, data and standards of medical-related scenarios. The greater participation of medical workers in this design is also an important contribution to the battlefield against epidemics.
I have no intention here to praise the contribution of AI technology and industry in epidemic prevention work. After all, it is far from praise and conclusion.Moreover, AI is definitely not the protagonist on the battlefield of epidemic resistance, but its auxiliary functions in many key areas can still let us see some future possibilities.We certainly hope that one day, AI can play a leading role in public health protection. After all, the more AI pays, it means less sacrifice for medical workers.
This is the first time that AI is moving towards anti-epidemic on a large scale. Although major AI companies have demonstrated sufficient response speed and social responsibility, they still need to see the entire social system and medical system. The tolerance and use of AI technology is relatively basic. . such as:
1. The ability of AI is not widely recognized by users
This time AI is rapidly moving towards the battlefield of epidemic resistance, mainly relying on the cloud + AI base has formed a high degree of completion, which basically has the characteristics of being available at any time.However, the AI capabilities provided to the front line of epidemic resistance are relatively single, lack sufficient scene coverage, and many of the capabilities can only be used on a small scale in first-tier cities. improve.
We can think of AI as a combination of a large number of basic capabilities and basic algorithms, which can be assembled at any time to solidify into specific products and platforms.
However, it can be seen in the fight against epidemic that front-line medical, medical research, and technology industry developers are not familiar with these capabilities and need to spend a lot of time communicating with each other. This has also led to the inability of many good AI capabilities to be quickly promoted across the country and settled in key regions.
2.The software and hardware industry chain is relatively weak
AI has a good performance on the battlefield of epidemic resistance, but it can also be seen that these performances mainly come from the software level.
It is difficult for us to see AI robots and IoT hardware rapidly moving towards the front line of epidemic resistance. First-line medical personnel also do not have medical robots that can be remotely controlled and equipped with cameras, microphones, stethoscopes and other equipment.From software AI to the future of AI robots, a well-structured AI + IoT industry chain is needed. The weak hardware ecosystem has made the intelligent efficiency of many industries in China as expected. This is also a problem that must be faced and solved in the next technology cycle.
3.The overall level of intelligence in public health needs to be improved
AI has done a lot to fight the epidemic. But maybe we still have to ask, logically, can AI make more contributions to the epidemic prevention and epidemic prevention?
The answer is obviously Yes.
For example, there are AI startups in the United States that use social network data to predict new pneumonia. Can this be a possibility in the future? Let AI warn us.
Another example is the intelligent management of public health systems and the intelligent deployment of emergency materials. Complete solutions have emerged in these areas. However, because these capabilities require long-term preparations, they have not been seen during the epidemic.
For example, in a recent hotspot event, some organizations suggested that the deployment of supplies was not timely because there were too few staff. Can AI be deployed unmanned? This is not uncommon in the intelligent industry and intelligent logistics scenarios, but it has not become a living force on the battlefield against epidemics.
The cruel and terrible epidemic is believed to have been deeply appreciated by everyone. There is no doubt that we will overcome the epidemic and that it will happen soon.
But at the same time we should also consider, what have we left behind after the epidemic?
When the epidemic dissipates, is it necessary for us to raise these possibilities:
For example, around the intelligent technology, comprehensively upgrade the public health and epidemic prevention technology system;
For example, increase investment in intelligent medical care, especially intelligent long-term investment in vaccine development, virus research, and new drug development;
For example, should we introduce and improve the hardware industry chain, and let AI robots replace medical workers and become the front-line fighters of clinical infections?
Maybe these problems are not only related to AI technology, medical industry, and the epidemic. But reflection and action are always timely.