We are experiencing an AI gold rush. But who will eventually benefit? Is it a unicorn company that controls alchemy? Are companies still in full swing? Or is it a tech giant with a "hoe shovel?" Which country will become the richest gold mine again?
Welcome to AI Gold Rush! Today, we are experiencing a gold rush for AI. Hundreds of millions of dollars are being poured into start-up companies in all walks of life. Google, Amazon, Microsoft, and IBM all invested a total of US$2 billion in 2016 to be even hot. The rest of the company is even more fierce game with its competitors, and at the same time watch out for those new players. China attaches great importance to the AI ​​industry, and the EU has also declared that it will invest 2.2 billion US dollars to keep their place. AI is everywhere. From Google's daily search volume of 350 million to the introduction of face recognition iPhone X, to Amazon's voice assistant Alexa. The title of news about AI is overwhelming. For example, AI can assist in medical diagnosis, help banks assess customers' loan risks, predict farmer's land harvest and market conditions, and help manufacturers control product quality. And political and financial think tanks are also actively assessing the possible risks of AI in politics and security.
AI and machine learning are about to become an inseparable part of social life. Since the AI ​​gold panic hits, we can't help asking who can finally find gold? Is the bravest few? Or can everyone share a share? Will most of the gold mines be taken away by companies that have the ability to provide Shantou shovels? So who will pay for the failure of gold rush?
What is the value of AI?
To figure out the value of AI, we need to figure out who will make profits first. (1) Chip makers, (2) Platform and infrastructure providers, (3) Model and algorithm designers, (4) Enterprise solutions Solution provider, (5) vertical industrial solution provider, (6) AI user, (7) country? This article will provide a value chain framework for analyzing AI value.
The companies and companies involved in this article are representative of important roles, but they cannot be comprehensive.
1. Who owns the best AI chips and hardware?
If the growth of computing power is exponential, then the growth rate of AI demand is even faster. AI, machine learning, and big data all have extremely high demands on matrix computing capabilities, and chips are being pushed to the cusp.
Nvidia's stocks have risen nearly 1500% in the past two years precisely because the GPU chips that they produced to render games can be perfectly applied in the field of machine learning. Google also recently released their second-generation TPU. Microsoft is also brewing their brainwaves (Brainwave).
At the same time, start-up companies like Graphcore have also raised $110 million in financing and are making great strides toward the AI ​​market. Traditional chip makers, IBM, Intel, Qualcomm, and AMD are just around the corner. Even Facebook is building its own chip team.
China’s role as a newcomer is equally rampant. China’s Cambricon technology has released the first generation of cloud AI chips with independent intellectual property rights in the past week.
Being the world's leading chip designer and manufacturer, and maintaining such a position is extremely costly. This requires its strong economic strength, as well as a world-class team of silicon chips and software engineers. This makes few come from behind. Just like the California Gold Rush in that year, the companies that provided the most cost-effective gimmicks and shovels have finally made a fortune.
2. Who owns the best AI cloud platform and infrastructure?
The AI ​​competition's front has been pulled into the "cloud". As early as 2006, Amazon realized that the need to rent computers and software would increase, and AWS (Amazon Web Service) came into being. Today, the demand for AI is growing so much that many companies are beginning to transition to provide AI's Platform as a Service and Infrastructure as a Service.
Amazon is the leader in cloud services, but Microsoft, IBM, Google and Alibaba follow
The competition for platforms and infrastructure is currently only among several giants. Microsoft's Azure public and private hybrid cloud service has more than 100,000 computers. In the past few weeks, they announced that their brainwave chips have increased the performance of Bing search engines by an order of magnitude. Google Cloud is not to be outdone, and we also see Ali cloud is quietly beginning to occupy the international market.
Amazon - Microsoft - Google and IBM's protracted war will not stop. China is also quietly becoming a full-fledged player in this battlefield. Once again, people with hoe and shovel will win.
3. Who has the most cattle algorithm?
Google can be said to have rich economic power in the AI ​​field, sitting on massive data, and attracting many world-class big coffee to join. Google's search, driverlessness, speech, and intelligent reasoning are all supported by AI algorithms. Now it also includes drug development and disease diagnosis.
TensorFlow, a nobody-knowing Google machine learning framework, is also freely available to AI developers. Why do they want to do this? Jeff Dean, head of Google's brain, said that today more than 200 million organizations will benefit from machine learning.
If these companies are able to use Google's open source machine learning software, their demand for computing power will also increase significantly, and Google Cloud is optimized for TensorFlow, so it will be the best cloud platform for developing and using TensorFlow once these Companies rely on Google’s software and services, and these users will become loyal fans of Google. Again, this logic is equally valid for Amazon, Microsoft, and IBM. They are also actively and cheaply providing their AI software services.
This competition is about deep learning and cognitive algorithms that will make artificial intelligence solutions available.
Not only is the machine learning algorithm, companies are also extremely fierce in their competition in some cognitive algorithms such as speech robots, natural language, semantic analysis, computer vision, and complex core algorithms. Clarifai is a start-up company that provides image recognition systems. They provide image matching, search and other services. The company has raised nearly 40 million US dollars in the past three years. It is estimated that the computer vision market will accumulate profits of 800 million U.S. dollars from 2016 to 2025!
In such a big market, the giants naturally can't sit still. For example, IBM has about 20 APIs. These APIs can be directly used by the company and applied to artificial intelligence applications such as chat, visual, speech, language, and knowledge management. KDnuggets listed more than 50 top-level cognitive algorithms and services (https://) from giants and start-ups. These services are embedded in the AI ​​cloud, making it easier for developers to use them. Just now, Microsoft's CEO Satya Nadella issued a statement that their artificial intelligence API has more than 10 million users, of which nearly 300,000 developers use the chat robot API. I think that any start-up company will not want to compete with these "Goliaths."
Competition in this area is a positive feedback and winners will be more likely to win in later competition. They will have the ability to build more powerful teams and collect more data. Successful startups will get more investment, their research and development will produce more excellent patents, papers will be released, and R&D personnel will also have more in-depth research in a certain field, thus accelerating the subsequent results.
These people are more likely to see the direction of AI's future development, so as to guide the future of scientific research, and go ahead of other companies. These startups have determined to die. They have succeeded in becoming world-class companies, or have been acquired by other big companies, failing to disappear. Even if it is not a commercial research team, they may be successful, or in the pocket of a big company. The 14-year DeepMind is a good example. This company that has been established for two years and is focused on enhancing learning algorithms was adopted by Google. 40 million US dollars acquisition.
4. Who has the best business solution?
Enterprise software has always been monopolized by giants like Salesforce, IBM, Oracle, and SAP. They are also aware that their software will lose its market without the embedding of AI tools. But these start-up companies that have sprung up like mushrooms have survived so well in the cracks, giving them tremendous pressure.
We analyze over 200 enterprise application cases involving customer management, market management, cyber security, artificial intelligence, human resources, and automated process robots (RPA). Enterprise applications are a real kaleidoscope, and every technology point has companies offering solutions. Today, in the recruitment field, there are more than 200 companies that use AI, most of whom are freshmen. The oldest player in the field of cyber security, DarkTrace, and UPA, the leader of RPA, have tens of millions of funding. These veteran companies do not want their strength to be diluted by the new company. They hope to use the investment to make their services still lead. Saleforce has invested in Digital Genius, a company that provides customer management solutions, and Unbable, a company that does business transformation.
Enterprise artificial intelligence solutions will improve customer service and productivity.
The problems faced by these old and big companies tend to be more severe. For example, SAP, they are trying to defend their historical position in cloud services and AI. We also see that many companies are providing tools that simplify the development and use of AI. These tools make it easier to design, deploy, and manage AI. Take the training of machine learning algorithms as an example, usually 80% of the time will be spent on the data. Second, the training of the training model will also take a lot of time. A company called Pettum in Pittsburgh, United States, is committed to how to quickly optimize and deploy machine learning models, and has raised more than 10 million funds.
If these companies can quickly prove that their solutions can truly meet the needs of businesses, their future will be infinitely better. However, just like the software gold rush, there are only a small number of start-up companies that can come out laughing in every field. If these companies are seen to have enough potential or pose a certain threat to big companies, they will be purchased.
5. Who has the best vertical solution?
AI is leading a competition on vertical industrial solutions. In the medical, financial, agricultural, legal, industrial and other fields, a large number of start-up companies are active. They are eyeing those foods that originally belonged to the old company's mouth.
These start-up companies can be more competitive if they have access to the following resources: (1) large brand-specific data sets, (2) expertise in a specific area, (3) A talented AI development and application team, and (4) sufficient funds, made them grow rapidly. In general, start-up companies that can be regarded as successful have generally complete listing plans, high business efficiencies, and high investment returns.
New industry AI solutions will bring strength to the organization and will also damage the organization.
For example, ZestFinance is committed to the credit rating field and raised 30 million US dollars. They claim they have the best data scientists in the world. They certainly have the best scientists in the world. They use 30 million to dig people! For newly arrived new companies, it is very difficult to squeeze into this market. They need to have enough economic strength to support it. As another example, Affirm, which provided loans for consumers and raised over 70 million U.S. dollars, these companies will quickly form a barrier to protect their competitive position. They have enough data to attract enough customers to obtain more. The profitability and financial flywheel effect will be able to operate at a rapid rate.
6. Which companies can demonstrate the value of AI?
Major companies are actively looking for AI solutions that can improve their services. They will never allow those start-ups to share their benefits with them. They will not even be willing to give up their advantages. Now, the innovation competition between companies is extremely fierce. Big companies have their own think-tanks. They invest in startups and grow their own startups to make themselves invincible in this innovation contest.
Large companies have an absolute advantage in this competition because they have a lot of data. The data is just like the relationship between gasoline and cars for AI or machine learning. So who will bring the advantages of AI to better use, are those insurance companies that have huge historical data? Or are those investment banks and financial companies that understand the behavior of all customers? Or is it a search engine company that knows all the search history of Internet users?
Companies are well valued from artificial intelligence to enhance customer service, increase productivity, and improve products and services.
Small and large companies have different positions, and they have different values ​​from the AI ​​market. According to Garner's forecast, by 2022, business activities triggered by AI are expected to reach 3.9 trillion US dollars. It is not an exaggeration to say that thousands of organizations need AI to solve problems for them. Companies can better meet customer needs through AI, save expenses, lower prices, and sell better products and services. AI will make big companies more powerful, and the price will be those small companies. But these big companies need a certain foresight, strong execution, and they can bear the failure of innovation. Their success will become inevitable.
7. Which country will be the largest beneficiary of AI?
Countries are also competing for dominance of AI. China is not afraid of this battlefield. China's investment in the development of AI technology and the growth of startup companies can be quite large. The relaxed management environment in China, especially the control of data usage, has made China a world leader in security and face recognition.
In the very near future, a Chinese police officer successfully identified the suspect from a concert of 50,000 spectators. China Business Technology is committed to large-scale face analysis and image analysis, and raised 60 million US dollars to become the world's most valuable startup company.
According to the China market report, their mobile market is three times that of the United States, and mobile payments are 50 times that of the United States. This is a huge data advantage. The EU’s strict control over data privacy may be one of the constraints of European AI development. The effectiveness of investing US$2.2 billion remains to be seen.
Will these be the artificial intelligence's sovereign winners?
Recently, Britain, Germany, France, and Japan have announced their AI strategies. For example, French Prime Minister Macarons announced that the French government will invest 185 million U.S. dollars in the next five years to support the establishment of the AI ​​ecology, including the collection of large amounts of public data. Google DeepMind, Samsung has set up a laboratory in Paris, France, and Fujitsu has also set up a Paris R&D center. The United Kingdom has just announced that it will invest 140 million U.S. dollars to support AI ecological construction, including 1,000 doctoral student grants.
However, if a country begins to invest in the construction of the AI ​​ecosystem, who will be the ultimate beneficiaries? Did doctoral students trained in the United Kingdom and France finally get poached by Google? At the same time, these machine learning engineers will have a six-figure annual salary, and most of the economic value will be taken away by the United States. His shareholders, or holders of US Treasury bonds, will be the final winners.
AI is bound to increase the efficiency and wealth of companies and countries. However, when 30%-40% of the work is replaced by machines, how will these assets be redistributed? The economists will jump out again this time, and the lessons learned from the increase of technology automation in history are still there. We will create jobs or deprive it. The focus of public debate often revolves around a statement by Geoffrey Hinton, who said that when a machine can diagnose diseases through images, the radiology doctors should be laid off.
But we look back at the Chinese market. Chinese radiologists read more than 140 million lung CT maps each year, and AI liberates them from reading large numbers of lung cancer images. Work has not diminished. On the contrary, the demand for efficient and accurate diagnostic personnel has increased. However, it cannot be denied that when wealth finally accumulates in a small number of companies or countries that possess technology and data, unrest may be issued. Those countries with relatively backward technologies will suffer. Because AI will only continue to upgrade countries with high-end technological strength, the disadvantaged countries will only fall further behind, and this will return to the concept of positive feedback mentioned earlier.
to sum up:
Through the analysis of AI layout, we are indeed in a golden age of AI. How to migrate the economic value of AI, summarize the following aspects:
1. The global tech giants will be the gold rush and spade – providing tools for anyone who wants to participate in gold rush. Google - Amazon - Microsoft and IBM are scrambling for AI's top spot. They provide chips, clouds, algorithms, and services in one go. China’s Alibaba and Baidu are now coming from behind. Small startups should not be underestimated.
2. Clusters of startups are committed to the development of cognitive algorithms, enterprise solutions, and vertical industrial solutions. These companies want to survive, they need to have monopolized data, professional domain knowledge, strong economic strength and can attract top talent. The real winners will be those companies that can truly solve the actual problems, and those companies that have lost their speculation will have no place. Buyouts may become obstacles for startups, or they may be a stepping stone to success. Many companies may eventually disappear in people's eyes.
3. The companies are already ready to go. Some people say that the output value of AI will reach trillions of dollars, and these big companies will have to share such a cake. AI can improve customer experience, increase productivity, increase automation, and increase the competitiveness of products and services. Most of this cake comes from companies that have a large amount of valuable data, companies with a large customer base, and companies with a wide range of industries. Large companies will become more and more powerful, but this requires that large companies have sufficient leadership and wit. Companies such as Google, Facebook, Apple, and Amazon themselves provide AI services, and of course they will have a large customer base. Traditional industries from the retail, medical and media industries, practitioners began to worry about their jobs, because the AI ​​algorithm will likely become their direct competitors.
4. Competition between AI countries is also ongoing. China's intention to become an AI leader in 2030 has not been deliberately concealed. China has the advantage of social structure. Many European countries are trying to persuade their governments to reduce their investment in AI, because they feel that the talents they have cultivated will in the end be robbed by these AI giants, such as Chinese and American AI companies.
All in all, the AI ​​gold rush will help companies and countries that can control AI tools, technology, data, engineers, and have a certain amount of industry, customers, and capital. These companies and countries will become the biggest beneficiaries of this gold rush.
However, there will also be a small part of the most brave startup companies that can find gold mines that have been forgotten in the corner. Like all gold rushes, most start-up companies will become victims. Many people feel that they have not felt any impact of this gold rush.
SHENZHEN CHONDEKUAI TECHNOLOGY CO.LTD , https://www.szsiheyi.com