The Metahuman effect blew up! But if you only think of metaverse, the pattern is a little small

Jin Lei Mingmin from Aofei Temple

Virtual? Reality?

The game technology of Epic Games makes people “confused”:

Think it’s a real brother playing with a big sword naked?

No, no, no, sometimes “seeing” may not be “real”.

The truth behind this video is as follows:

Yes, what you think is false.

However, the effect of the video display was too realistic, which attracted many netizens to marvel, and was praised 2399 on twitter. Feel this feel:

Some netizens exclaimed: for a moment, I don’t even know which side is the real world!

The technology used by this little brother is Metahuman of Epic Games.

Yes, it’s “meta” again, or it’s hard to distinguish between virtual and real… Maybe you’ve thought of the recently outrageous metaverse.

If you have such words in your mind, don’t make complaints about it.

In fact, the game technology that makes it difficult to distinguish between virtual and real is not just a tool to shape metaverse, but an important force to secretly promote the evolution of real science and technology and accelerate the development of various industries.

Yes, playing games can really change the world.

What is game technology changing?

We might as well look at this problem from an industrial perspective.

If you want to ask what is the hottest project in the technology industry in recent years, automatic driving is undoubtedly the one you can’t get around.

Game technology is promoting the accelerated development of this “Ai master”.

In the automatic driving technology, the test based on the real road environment is an important link to ensure that it can be on the road.

The industry generally believes that it needs to accumulate 17.7 billion km of test data to reach the road level. What is this concept? About 4.9 times between the earth and the sun.

Therefore, virtual simulation technology has become a necessary tool to promote autopilot landing.

The limitations of real road test can be well avoided in the virtual environment. The implementation of this process is very similar to creating scenes similar to reality in the game:

It is not difficult to see that the simulation system like the above highly restores the real scene, including geometric image, physical law, operation logic and so on.

A very important point in this process is to be realistic, and the simulation system should be “close” to the real world as much as possible.

Including changes in light, wind and snow weather, etc., so that the impact of these factors on the automatic driving sensor can be accurately measured, so as to ensure safety when on the road in reality.

Game technology, more specifically, is the game engine, which plays a vital role in this process.

Just like unreal engine 5, which has caused a sensation in the industry, it can handle the details stably and let netizens call it “virtual reality? Silly and confused”.

Therefore, if we can make good use of the advantages of game technology in building virtual simulation system and make it infinitely close to the real world, it will undoubtedly greatly promote the process of automatic driving landing.

In addition to autonomous driving, game technology also has a great impact on the film and television industry.

Also after the release of unreal engine 5, the film and television industry seems to quickly smell the subversive impact of game technology on movies.

So that in this year’s Shanghai Film Festival, the “film technology experience exhibition area” was set up for the first time, virtual production became the focus of the exhibition, and the unreal engine was invited to participate for the first time.

For another example, in the TV series full time master, the game engine technology was introduced for the first time in China, making the switching between the reality and the game scene of the male protagonist (Yang Yang) in the play a whole and very natural:

This kind of virtual production is the latest change in technology for the film and television industry, such as audio film, color film, optical synthesis, motion control, digital synthesis and so on.

Virtual production can not only realize the direct interaction between actors and virtual characters, but also the shooting team can control the scene in real time, and achieve real-time rendering through game technology such as unreal engine, which will save a lot of time and resources.

In addition to industry, game technology is also promoting the development of scientific research.

Yes, although “game” and “scientific research” seem to be two opposite words, their combination is really catalyzing the development of academia.

For example, in the school of information science and technology of Peking University, a top university in China, Professor Li Wenxin took her doctoral students to do scientific research in the glory of the king.

What they need to do is to train an optimal decision model within limited time and resources and deploy it to the game AI battle server platform.

It’s like a process of “alchemy”. Let their agents practice various “skills” through training, and then fight with other agents:

In their eyes, even game AI is the real AI:

They think   “Game” is the abstraction and imitation of “reality”. In the game, we can get a happy and successful experience close to reality, but avoid the real loss caused by mistakes and mistakes in reality.

Because of this, the game is a very good virtual environment for trial and error and iterative growth.

It is an important educational and research means to turn the real problem environment into a virtual game and iteratively optimize the problem solution through a large number of trial and error correction in the game.


In a word, the current game technology is playing a subtle role in both industry and scientific research.

This is a hot topic in the 2021 Tencent game developers conference (TGDC) held by Tencent game school, a recent game industry event.

For example, Tencent, which does not build cars, has built an automatic driving virtual simulation system using game technology   Tad SIM to improve the authenticity of autopilot simulation.

In addition to supporting scene editing and playback simulation of road acquisition data, TAD Sim can also use technology similar to game agent AI to train traffic flow AI with a large amount of road acquisition data to generate traffic scenes with high authenticity and strong interactivity.

Another example is to use Metahuman technology to generate high fidelity digital people; AI agents or platforms such as “Jue Wu” and “Enlightenment” help the development of intensive learning and strategy training

At the same time, Xia Lin, vice president of Tencent games and President of Tencent game school, also introduced the achievements after the “interweaving” of games, technology and academic circles.

For example, Tencent game school has launched scientific research and academic cooperation with 16 universities at home and abroad, and has published 30 subject achievements, 8 national patents and 5 reports.

The first master’s major in “interactive media design and technology” jointly built by Tencent and Tsinghua University has also ushered in the second academic year.

△   Xia Lin, vice president of Tencent games and Dean of Tencent game school

In Charlene’s opinion:

As a highly interdisciplinary industry, games are placed at the intersection of digital virtual and real world.

With the significant changes taking place in the global game market, the trend of high-quality, globalization and diversification of the game industry will be more prominent.

After reading this, some people may wonder why a game company should invest in cross domain technology R & D and application.

But in fact, if we dig into the history of game and technology development, it is not difficult to understand why Tencent attaches so much importance to it.

Games have always been the best “experimental field” of cutting-edge technology

Games are the “experimental field” of cutting-edge technology, which is never false.

The most powerful proof is the development of hot AI technology.

Perhaps when it comes to AI, many people will think of IBM’s “dark blue” and Google’s alphago, which have defeated humans in playing chess.

In fact, chess is the main form of video games in the earliest stage.

As early as 1950, less than five years after the birth of general-purpose computer, human beings had played a game with artificial intelligence.

At that time, computers were more used in laboratories. If they could be used to do something else, people would feel very novel.

Play chess on the computer or fight against the computer?

This may have long been out of the public imagination. Therefore, when brain Bertie appeared at the Canadian national exhibition, it attracted a lot of people’s attention.

This is a game similar to tic tac toe chess. Players can play chess through a nine palace operation board. The computer will “think” for a moment, and the chess situation will be displayed on the large screen at the same time. The difficulty of the game can also be adjusted.

According to the definition of machine intelligence by Alan Turing, the “father of artificial intelligence”:

If a machine can talk with human beings (through telex equipment) and cannot be identified as a machine, it is said to be intelligent.

Then, brain Bertie may be what artificial intelligence looked like in its earliest days.

This game based form is actually the way many scientists explore AI.

In 1944, von Neumann first proposed the minimax algorithm of two people playing chess in his book game theory and economic behavior;

In 1947, Turing wrote the first chess program;

In 1950, Shannon published the article “computer chess program”, which started the theoretical research of computer chess.

In 1951, Turing’s good friend Christopher Strachey wrote the world’s first visual game checkers.

Unlike brain Bertie, checkers is a game similar to checkers, with a higher degree of difficulty; But the same thing is that they are man-machine combat.

Later, Arthur Samuel, a pioneer of artificial intelligence who popularized the word “machine learning” in 1959, also wrote a checkers game on a commercial computer of IBM.

At this time, the checkers game has been very different, because the machine is no longer simply settled according to the program, but can fight through autonomous learning. Samuel also set up a reward mechanism in the program.

Here, we have seen the shadow of reinforcement learning.

In the following decades, scientists in the field of AI seemed to be playing chess. They tried checkers, chess and go.

Ken Thompson, the father of UNIX, once developed a chess computer Belle, which is also the world’s first chess computer that can reach the “master” level.

Moreover, it also attracted attention at the national level. At that time, Belle wanted to go to the Soviet Union for competition, but was detained by the U.S. government at the customs because of “trying to deliver advanced weapons to the Soviet Union”.

In fact, the US Soviet hegemony was also extended to computer chess games.

It can be seen that AI has always stood at the forefront of scientific and technological development since its birth, and video games have always been an important scene for AI to reflect its strength.

The game is not only a “testing ground” of cutting-edge technology, but also an important channel for the general public to touch cutting-edge technology.

In astronomy, biology, medicine and other research fields, there are always games.

For example, novel coronavirus pneumonia has been involved in the fight against Foldit.

It combines protein structure prediction with the idea of spatial puzzle game. Players need to constantly adjust the side chain, skeleton and other structures of protein to make the final structure reach the required score.

During the novel coronavirus pneumonia outbreak last year, scientists in order to quickly identify the structure of the virus antibody, introduced the latest research results to the game, allowing game player to try to design new structure for research reference.

Moreover, this is not the first time Foldit has contributed to scientific research.

Previously, the reverse transcriptase structure of AIDS was cracked by game player in this game. The structure that scientists failed to predict for more than a decade was solved by players in ten days.

Nature also named the game in 2010.

Coincidentally, there is also a scientific research project in DLC’s wedding battle of guns, love and tentacles in ownerless land 3.

Developers designed a simple DNA classification game on an arcade in the game. Bricks of different colors represent different nucleotides.

Players only need to make the bricks in a certain line have the same color within the rules to pass the customs. The process of arrangement is actually to correctly classify the DNA.

The resulting data can help Mcgill University draw a map of human intestinal microbiology, which plays an important role in clinical research of diabetes, depression and other diseases.

Not only that, many sophisticated technologies also embrace the public through games.

For example, SpaceX, which has repeatedly let civilians experience space travel, launched a space simulation program last year.

It allows people to experience the process of spacecraft docking with the space station in the game.

Moreover, the restoration of space state is very realistic. A little operation will cause force changes, which is also a test for players.

Obviously, video games, originally regarded as entertaining the public, have played many unexpected roles in the development of decades.

Hasabis, founder of deepmind, once said:

Game AI is a stepping stone to general AI. The real reason why we study these games is that it is a very convenient testing ground for developing algorithms.

This sentence may now be extended to VR, AR, simulation and other scientific fields.

Nowadays, when we talk about games, the words such as autonomous driving, AI and Metahuman are no longer strange.

This is why at a game developer conference of Tencent, we will see the production, learning and research achievements of many universities, covering areas such as in-depth learning, urban structured modeling and so on.

We can feel that this is not the end of the combined development of games and cutting-edge technology. At present, games, as an art and entertainment form with the most intensive application of cutting-edge technology, not only promote other industries, but also catalyze the technological evolution and release more diversified values.

How many cutting-edge technologies will advance and land through games in the future?

Let’s wait and see.