“Metaverse + artificial intelligence”, who can achieve who?
This year, the concept of “Metaverse” was born, and its popularity is no less than that of the original “blockchain”, from everything can be blockchain, to everything can be Metaverse. If it is understandable that a large number of Internet companies are constantly deploying the concept of “Metaverse”, it is a bit confusing that clothing, food and even real estate are engaged in the Metaverse.
Not to mention whether the Metaverse concept is being overconsumed, can it really help so many industries develop?How can it be possible, then the first one should be the “artificial intelligence technology” as one of its underlying supports, including computer vision, machine learning, natural language processing and intelligent speech, etc. These technologies are currently in the embarrassing bottleneck of commercialization. , can the Metaverse let them get out of the bottleneck?
AI “runs through” the Metaverse
When it comes to the Metaverse, in fact, quite a number of film and television animation enthusiasts are not unfamiliar. The term Metaverse was born in the science fiction novel “Avalanche” in 1992. In 1999, a Hollywood blockbuster “The Matrix” was the most popular. The concept of “Metaverse” was first turned into a work; in 2002, another Japanese manga called “Sword Art Online” further enlarged the relevant elements; it was not until “Ready Player One” was passed in 2018 that “Metaverse” really entered the public eye. .
However, the specific definition of the Metaverse is still unclear, but even so, it still cannot stop the market’s enthusiasm for the Metaverse, and many people even believe that the Metaverse will become the final form of the Internet.
The New Media Research Center of Tsinghua University divides the technologies that support the development of the Metaverse into two categories: back-end infrastructure and underlying architecture. The former includes Internet of Things technology, interactive technology and video game technology, while the latter includes network and computing technology, artificial intelligence technology and blockchain technology.
At first glance, it seems that the concept of the Metaverse is born for video game technology and artificial intelligence technology, one is its best application end, and the other is the inevitable technical end of the existence of the Metaverse.But in the field of application, the Metaverse is obviously still too early, so can today’s artificial intelligence technology support the Metaverse?
Some people have divided the Metaverse into eight elements, namely: identity, social interaction, high immersion, low latency, diversity, anytime, anywhere, economic system, and civilization. One direction of the intervention of artificial intelligence technology is to enhance the high immersion and diversity of the entire “Metaverse” through technologies such as computer vision, machine learning, natural language processing, and intelligent speech. Not only that, artificial intelligence will run through the entire ecological chain of the Metaverse, from content production, distribution to application, to accelerate content production, enhance content presentation, and improve content distribution and terminal application efficiency.
The reality is that the current application scenarios of these technologies are only simple application scenarios such as license plate number recognition, face recognition, content recommendation, real-time translation, and intelligent customer service. This also highlights a contradiction faced by the development of artificial intelligence technology, that is, the slow implementation of commercialization, resulting in technology and enterprises not making money, making technology research and development and implementation once again bottlenecked.
Breaking this bottleneck, can the Metaverse become that sword?
Can the “Metaverse” make the underlying technology truly rejuvenate?
“Artificial intelligence” does not make money, just like “the meat buns beat the dogs, there is no return”, it has become a well-known law in the industry. Before the price of pork fell sharply, some people even said that artificial intelligence is not as valuable as raising pigs, and the prospects are very promising. Good but the “money” scene is not good.
Take a look at the domestic “AI four little dragons”, Shangtang, Megvii, Yitu, and Yuncong are the answers. For many years, these four companies have been firmly occupying the top four shares of the domestic computer vision application market, and they have achieved technology implementation in retail, security, manufacturing, government affairs, medical and other industries, but the issue of making money still seems to have nothing to do with them. Half-hair relationship.
Since last year, the “Four Little Dragons” have started the road to IPO listing and financing. In August this year, it was reported that SenseTime had submitted a listing application to the Hong Kong Stock Exchange, and SenseTime had become the last company among the Four Little Dragons to submit a listing application. , but their prospectus is more “ugly” than the other.
None of the four companies made a profit, and the loss area continued to expand. Judging from the four-year loss range, it can be seen from their prospectus that Megvii Technology lost a total of 14.25 billion yuan from 2017 to the first three quarters of 2020; Yitu Technology lost 7.286 billion yuan in three and a half years; Yuncong Technology lost 2.2 billion yuan from 2017 to the first half of 2020; SenseTime lost 24.3 billion yuan in three and a half years from 2018 to the first half of 2021, one more than the last three.
However, their only way to “make money” is financing. According to Tianyancha, SenseTime has raised a total of 9 rounds, with a cumulative financing of more than 2.6 billion US dollars; Megvii Technology has raised more than 1.2 billion US dollars in 7 rounds; Yuncong The 11 rounds of financing for science and technology exceeded 3.4 billion yuan; the 10 rounds of financing for Yitu technology exceeded 2 billion yuan.
This is also the reason why they have to seek listing at present, but we must know that under the framework of artificial intelligence technology, computer vision technology is already one of the most mature technologies, but it still faces huge losses, “burning money” in research and development, and reducing gross profit margins. “There is no problem solving”, and they all claim that they will not consider profit for the time being. How difficult will other artificial intelligence subdivision technology fields be?
Another example is Baidu. This company is a leader in the fields of natural language processing and intelligent speech. Baidu Natural Language Processing (NLP) is one of Baidu’s oldest basic technology departments, but the technology brought by Baidu It’s just a layer of “value halo”, but it’s definitely not making money.
As we all know, technology cannot be directly profitable. It needs to finally land on the scene or product. Baidu smart speaker is the carrier of Baidu natural language processing and intelligent voice technology. Is Baidu smart speaker easy to use? This answer can be seen from the market share in general. According to data from Luotu Technology, in the first half of 2021, Baidu, Tmall Genie, Xiaomi, and Huawei’s top four brands have a share of 95.8%, of which Baidu’s share has reached 32%, ranking first in the market. Anyone who knows something about the smart speaker industry knows that Baidu smart speakers have topped the industry for many years, but so far smart speakers are still a “losing money”. Smart speaker products are also telling the market that even if they have technology and products, they will never make money.
This situation makes one wonder if “technical companies” really can’t make money? Therefore, compared to the Metaverse that requires artificial intelligence technology as the underlying technical support, it is better to say that artificial intelligence technology may need to continue to amplify the technical story through the concept of the Metaverse and attract the attention of capital.
If the C-side does not pay, and the B-side is difficult to pay, will the Metaverse follow the old path?
Compared with AR, VR or MR, a product that is more in line with the concept of the Metaverse may be a “brain computer”, which also caused a huge “storm” in the early years.
Like the concept of the Metaverse, the product of the brain machine was also born in science fiction works, but it gradually lost its voice as it slowly entered the market. Although the big world of the Metaverse is still far from fashion, it is time to discuss commercialization, and the problems faced by the “brain machine”, the best “entrance and carrier” of the Metaverse, may eventually become the Metaverse. problem of the universe.
First of all, the C-side does not pay. At first, many technology companies showed great interest in brain computers. For example, as early as 2019, iFLYTEK, its wholly-owned subsidiary, Xunfei Yunchuang, jointly invested and established a company called Guangzhou South China Brain Control Intelligent Technology. Co., Ltd. is a brain-computer interface company with a registered capital of 41.7687 million yuan.
Since 2019, DeepBlue Technology has also reported its intention to enter the field of brain-computer games. Today, two years have passed, but no sound has been heard.
Speaking of which, do you still remember a domestic variety show called “Extreme Challenge”? A “idea drone” has been shown above, which controls the take-off, landing and flight of the drone through the head ring of the brain. I thought that such a cool product would definitely have a good market performance. However, Yunrui’s “idea drone” is still on sale, but through its official website and operating materials, it seems that the main business of the enterprise has shifted to In the field of medical health, the market is not individual users, but various hospitals, scientific research institutes, physical examination centers and health care units.
In addition, a unicorn company called BrainCo on Harvard campus has also launched two brain computer products, namely the Focus Fusi headband and the Focus Fit meditation headband, which are mainly used to improve attention. A similar product appeared and was partially applied to students, but it attracted a lot of abuse, and the ending can be imagined.
Secondly, it is difficult for the B-end to pay for orders. In addition to the C-end market, the B-end market is also optimistic. At this stage, the most famous brain-computer company in the world is Musk’s NEURALINK. Today, NEURALINK’s technology is still more in the experimental stage, and the commercialization part is still progressing slowly.Domestically, Naolu Technology is a leading enterprise in the domestic EEG field. Compared with its personal consumer-grade sleep products, it has had several iterations, but its enterprise-grade products have only one real-time monitoring and protection operation applied to security and safety production management. The helmet for the mental state of the personnel has to be said that this is far from the expectations of users for the application of brain-computer technology. It is also because the vast majority of enterprise-level users have no knowledge at all, and do not know what this type of product can bring to them. Even though the concept of the Metaverse is hot now, it is not an exaggeration to say that more than 90% of the participants are blindly following the trend.
At present, the main audience of the B-end market is still scientific research and medical institutions, and the scene is still mainly in the laboratory. The main reason is the technical level. For example, the application of brain-computer in the medical field has brought new hope to the whole human being, but the bandwidth problem of the brain-computer interface has always prevented the application from achieving breakthroughs. In addition, the high cost makes it difficult for companies to afford it. For example, Musk once said frankly, “At this stage, the technology of brain-computer products is not yet mature, and the small size makes the cost of the equipment very high.”
The 2017 Sino-US Joint Brain Computer Report pointed out that it takes an average of 7.4 years to double the number of neurons that can be recorded at the same time. It takes 2100 years to record 1 million neurons at the same time. All neurons (5 to 10 billion) in 2225. This means that without solving the problem of bandwidth, the role of brain computers in the medical field will be greatly limited.
In addition to brain-computer products, including artificial intelligence companies such as SenseTime, Megvii, Yitu, and Yuncong, are their technologies worthless? Obviously not, it’s just that either you can’t find a suitable landing direction, or the suitable field has not grown up. For example, in the field of intelligent driving, computer vision is one of the key technologies to ensure intelligent driving, but at this stage the entire intelligent driving industry is still in the research and development stage, so it still takes a lot of time for computer vision companies to achieve commercialization.
Whether the development of the Metaverse, as well as the application of technology and commercialization, will encounter these problems, the answer is yes. Maybe 3 or 5 years later the Metaverse will really be born, but for now, at least for artificial intelligence companies, the new stories brought out by the Metaverse are far greater than its actual value. Artificial intelligence technology achieves the Metaverse? Perhaps it is the Metaverse that achieves artificial intelligence technology.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/if-the-c-side-does-not-pay-but-the-b-side-is-difficult-to-pay-can-the-metaverse-really-help-artificial-intelligence-technology-learn-to-make-money/ Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.