Where there are people, there are rivers and lakes. The "technical battles" of giants are the appearance of modern rivers and lakes. All enterprises firmly grasp that science and technology is the key to the primary productivity and constantly compete for the comprehensive strength of platform, technology, scene, ecology and other aspects. You sing and I'm on the stage.

In our cognition, genetic science and artificial intelligence are "Wulin experts" in technology. In a series of press releases, they are usually accompanied by words such as "Immortality" and "rejuvenation", which makes them feel proud of the Jianghu.

Compare the two carefully - artificial intelligence is at a critical moment and has more or less penetrated into people's lives, but genetic science always "only hears its voice, not its people", like the flower of kaolin, sealed in scientists' laboratories, and few people have access to its core technology.

Who can make great achievements in the vast world? Artificial intelligence and genetic science are in close combat. Is it friendly cooperation or competition? In this century's eye-catching "Huashan sword debate", Yan Xuan, an industry analyst of intelligent relativity (ID: aixdlun), opened a brain hole.

From the flowers of kaolin to the flowers blooming everywhere

At present, the industrial map of artificial intelligence has been very rich, and the shadow of AI can be seen in the fields of finance, education, medical treatment, public security and so on. From this point of view, although genetic science is the predecessor of AI, the speed of "fame in the Jianghu" is not as fast as that of the younger generation.

The application of gene science usually focuses on gene detection. There are two application scenarios for genetic testing, one is health check, such as cancer prediction, but this is generally the high-end medical service sought by the rich, so that it has become famous for World War I. It is Anngelina Jolie who has detected the risk of breast cancer through gene sequencing. The other is a commercial "fortune telling" genetic test, such as whether your child has singing talent through genetic test. The son of boxing champion Zou Shiming has done such a test.

But this does not mean that AI is slightly better. The field of life science is limited to barriers such as research environment and technical conditions. In previous years, it has developed slowly, but in recent years, the high technology of life science may enter an explosive period. For example, gene editing is well known as the Star Technology CRISPR / cas9. Last year, the international research team led by mitalipov completed the study of the first human embryo gene modification in the United States.

The imagination space in the field of life science will be very large. Of course, this kind of technology will also encounter the same ethical constraints as AI. From this point of view, AI or genetic science "fight alone" will become a hero in the Jianghu. However, the combination of the two experts seems to be more attractive considering their specialties and some defects. Especially in the medical field where both want to attack, there is an opportunity to show their skills.

1. AI and gene complement each other in disease diagnosis

The correlation between a gene and disease is based on statistics. For example, when studying the relationship between cancer and human genes, we should first count the gene data of a large number of cancer patients, then compare them, and then screen out similar genes, that is, select the genes that may cause cancer.

But this result is easy to be overturned, first of all, because the statistics are the results of previous cumulative data. With the increasing number of cancer statistical samples, people will find more similar gene loci, and the probability of suffering from cancer will change accordingly.

Secondly, human suffering from a certain disease is also related to other factors. For example, the prevalence of lung cancer is strongly related to air pollution. However, genetic testing can not touch the actual situation of patients.

In other words, even if a large number of data are obtained from gene testing, the direction of disease is still limited. In the face of such a situation, it may be helpful for us to join AI to deeply learn the functions of to C-side database.

Take the AI death prediction system recently developed by Stanford University in the United States as an example. This system counts nearly 2 million medical data of patients who die within 12 months, then calculates the weight and intensity of each information by using big data through deep neural network, and compares it with existing patients to generate the probability score of the patient's death within 3-12 months. In this way, the accuracy of disease diagnosis can be guaranteed to the greatest extent.

In addition, AI can supervise patients' actions through Chatbot, scene recognition and other functions, which is more conducive to assist gene diagnosis.

2. Divine assistance of pharmaceutical industry

Professor Peter Coveney of University College London, UK, said: "through the patient's genome sequence, we can infer the shape of the enzyme, construct an accurate three-dimensional protein structure, screen matching drugs, tell the results to the attending doctor and give the optimal prescription."

Combining gene sequencing technology with the computing power of AI may be able to explore a new path for the pharmaceutical industry. Taking AIDS treatment medicine for example, scientists say that the more than 50 simulation models they have built now will be equipped with 5000 processor computers to calculate for 12-18 hours without stopping, and do a lot of data analysis to calculate the results, so as to give the sorting of drugs. In the pharmaceutical industry, AI is comprehensively improving human efficiency.

3. Personal health management terminal

Gene chip has always been a "magic weapon" in the field of life science. What breakthrough will be achieved by combining it with AI? Health management.

Gene chip can quickly and accurately identify a person's biological information. Unlike AI, which can only recognize images, intelligent terminal + gene chip may create a human GIS intelligent device. This rich multi-dimensional information chimera can be used to interpret personal health and identify personal identity information. It is equivalent to having an intelligent, exclusive and visible "genetic ID card".

Such equipment can also quickly detect the pollution and harm of contaminated microorganisms or organic compounds to the environment, human body, animals and plants. After improvement, gene chip can express different digital information by using different biological states. Based on gene chip and gene algorithm, it can also be used to manufacture biological computer.

Jump out of the AI "humanoid" circle

On the one hand, in genetic science, the significance of interpreting human gene program language lies in interpreting the program code of the earth's biosphere. We learned in textbooks that genes are codons that compile amino acids. Except for a few differences, the genetic codes of known organisms on earth are very close. Therefore, according to evolution theory, genetic code should appear very early in life history. This is of great significance to the study of the birth of species. In the future, we do not rule out the possibility of "man-made" earth biosphere.

On the other hand, the robots, computers and other artificial intelligence we mentioned at present are unconscious. Even if we constantly try to make AI more like people, what machines do is mainly cognitive. What they do is a pattern recognition. Listen to the sound and look at the images. The advanced functions of the human brain still cannot be transplanted to AI. There has been no major theoretical breakthrough in AI since deep learning, which also proves that it is difficult to create non life with consciousness and emotion.

American philosopher John Hiller once put forward the experimental theory of "Chinese Room" to refute the view that computers and other artificial intelligence can really think. In this thinking experiment, as long as the computer has an appropriate program, in theory, it can be said that the computer has its cognitive state and can carry out understanding activities like human beings.

Based on this theory, we may be able to open a relatively large brain hole, that is, to jump out of the circle that artificial intelligence must appear as a bionic machine and "human like AI", but to use animal thinking to simulate animal programs, which may be a space with great potential.

Kirsti Albert, from the University of New England in Australia, once marveled at the intelligence of ants. "Ants are much smarter than we think - we call ants superindividuals because they form many networks similar to the human brain." indeed, compared with the human social system, the social system of ant colony is quite sophisticated.

If we regard this "artificial" biosphere as an experimental factory of biological evolution, we may be able to summarize a set of program language of a species different from human beings. It should be explained that this animal specific program has always existed, such as the social behavior of ants, but no animal will summarize it by itself in the process of evolution. When human beings can look at the whole life history, it is not impossible to summarize the program language of a species, which means that human beings can create "animal intelligence" based on the new program language.

Instead of allowing AI to continuously accumulate and learn Artificial label data, it is better to build a knowledge map of animal behavior and carry out systematic training. Once the training is successful, the benefits will be permanent for people.

Most importantly, when genetic science makes a breakthrough, human beings will have the opportunity to obtain non formulaic variable fluctuations (i.e. emotional fluctuations) of animals and plants. Then, implanting such fluctuations into AI as a database may solve the problems of insufficient animal corpus and unclear semantics in our daily life. If animals have true feelings, we may realize the emotional communication between human and animals through gene technology + AI technology.

In any case, AI and genetic science will blossom everywhere sooner or later, and human exploration of science and technology and nature will always move forward like the arrow of time.