CB Insights, a well-known research institution in the United States, recently released a heavy report "Top AI Trends To Watch In 2018", which provides an in-depth study of the current development of the AI ​​industry and gives 2018 The 13 leading edge trends in the AI ​​field are the most noteworthy.
The CB Insights report mentioned that Chinese T-shirt manufacturer Tianyuan Garment Co., Ltd. and the Arkansas State Government signed a memorandum of understanding to launch 400 “workers†in the new clothing factory in Arkansas. It is worth mentioning that these 400 are sewing robots developed by Georgia start-up company SoftWear AutomaTIon. In this cooperation, the complicated work is all done by robots, and the human workers are only responsible for high-end work such as robot maintenance and operation.
The wave of machine automation driven by AI was once thought to impact labor-intensive manufacturing, resulting in the replacement of large-scale blue-collar jobs.
But what is easily overlooked is the emergence of new types of work: robot nannies. With the automation of highly automated manufacturing, warehousing and other robots, it is also necessary to maintain the robots. For example, in the Amazon warehouse, more than 100,000 robots have been put into use, which has created new jobs for thousands of people. In Japan, by 2025, more than 80% of aged care will be done by robots instead of nursing staff. .
2. Machine learning is omnipresent and omnipotentCB Insights points out that in 2018, artificial intelligence is everywhere, or rather, machine learning will be everywhere. This technology is almost "omnivorous" and will create endless possibilities in 2018.
In 2018, the UK's IntelligentX is expected to launch the world's first AI brewing beer; Russia's DeepFish is committed to using neural networks to identify fish in radar images; Sweden's Hoofstep is raising wind investment funds, plans to deepen the horses Behavior analysis.
3. The potential broke out, China has become a global leader in AIIn the field of AI, China is trying to surpass the United States and other Western countries, and the government has invested a lot of energy and money. In 2017, China's artificial intelligence startups accounted for 48% of the global total, and for the first time surpassed the United States to top the list. According to CBInsights, the United States will continue to lead the world in terms of the number of artificial intelligence startups and total equity transactions in 2018, but it is gradually losing its global trading dominance.
China's achievements in the AI ​​field are mainly due to the development of two major areas of face recognition and smart chips. The former benefited from the government's strong support in recent years, and the latter is directly on the strong US chip. challenge.
In terms of facial recognition, the unicorn Megvii has achieved great results. The company's investor Alibaba Group (through Ant Financial) and Foxconn jointly launched the "City Brain" project in Hangzhou, China in 2016, using artificial intelligence to analyze and monitor camera data.
In terms of smart chips, in July 2017, the Chinese government stated that it will reach a cooperation with the United States in 2020 and become a world leader by 2030. Chinese company Cambricon is committed to producing 1 billion processing units in the next three years and is developing chips specifically for deep learning.
In addition, in addition to domestic research and development, China's major technology giants such as Tencent, Baidu and JD. It is reported that Baidu and Jingdong have invested in ZestFinance recently, and Tencent has invested in ObEN. In 2018, this will also become a powerful chip for the Chinese chip competition.
4. The future of national defense turns to AIFuture wars will depend on unprecedented intelligence. The drone is just the beginning. With the increasing convergence of traditional defense, surveillance, and cybersecurity reconnaissance, the need for algorithm-based AI.
Artificial intelligence has a natural advantage in the field of defense. Because cyber attacks are constantly evolving, it is often necessary to face previously unknown types of malware during the defense process. Artificial intelligence, by its powerful large-scale computing power, stands out and quickly screens millions of events to detect signs of anomalies, risks and future threats.
According to the CB Insights report, in 2017, about 34 companies were IPOs and joined the market of big companies such as Cyber ​​eason, CrowdStrike, Cylance and Tanium. Among them, each company's valuation is more than 900 million US dollars.
5. Voice interaction competes globallyThe CB Insights report shows that voice interaction warfare currently coexists in English-speaking countries: Amazon Echo and Google Home dominate the smart home speaker market.
It is worth mentioning that although Amazon has a leading position in speech computing, it is lagging behind in language support. It hopes that users around the world can interact in English, German or Japanese.
Google Home is available in English, German, French and Japanese. Apple's HomePod is currently only available in English.
In addition, in the Chinese market, Amazon and Google are not very playful. The Chinese market is still raging, and the market-leading competition will continue. Alibaba's previous report mentioned that since its official release in July 2017, its Chinese version of the Tmall Elf has sold more than 1 million sets.
6. White-collar work is challenged by AICB Insights believes that AI will enter the decision-making stage more generally, which will pose a threat to white-collar work.
An EAAS market map shows that lawyers, consultants, financial consultants and other professional work will also face the invasion of AI, but in these areas, AI mainly plays an auxiliary and improving role, such as improving work efficiency, and it is more reasonable price charges. And commercialization, which will affect the pay structure of external law firms on an hourly basis.
What kind of white-collar work is temporarily safe? CB Insights' view is that areas such as education and healthcare that require high levels of emotional intelligence are currently at the lowest risk of automation.
7. AI terminal trends appear, edge computing becomes the next big fieldArtificial intelligence is not limited to powerful supercomputers and large devices, it is also becoming part of smartphones and wearables and devices. CB Insights said that AI development is entering the "end" era, and terminals such as mobile phones, automobiles, and wearables will increasingly welcome AI blessings. The marginal application of artificial intelligence is far more than this, and it has its own appearance in many popular fields such as smart home and automatic driving.
8. Capsule network appearsThe CB Insights report shows that neural networks have an infinite number of architectures. One of the most popular types of deep learning recently is called convoluTIonal neural networks. A new architecture, the capsule network, has now been developed that will exceed the convolutional neural network at multiple fronts.
Although the convolutional neural network has been successful for a long time, scholars generally believe that it still has defects, which may lead to a security gap. Based on this, Geoffrey Hinton, one of the pioneers of deep learning, published a research paper in 2017, introducing the concept of “capsule networkâ€. Compared with CNN, CapsNet makes up for a lot of defects, such as data volume, accuracy, training data diversity, etc., and better performance.
9. The battle for high-paying AI talents is intensifyingCB Insight cited the report as saying that there are currently about 300,000 researchers in the field of artificial intelligence, including some students. According to the data, the demand for AI professionals in global enterprises is likely to exceed 1 million. The talent competition of global artificial intelligence companies will intensify, and the compensation of artificial intelligence experts will reach a new high in 2018.
In the United States, Glassdoor's search for "artificial intelligence" shows that there are currently more than 32,000 jobs listed, several of which have reached six digits. The company is very willing to pay a high salary for smart AI experts.
10. Machine learning capital carnival ended2017 is the peak of carnival for machine learning. Investors invested more than $15.2 billion in machine learning startups in various industries, an increase of 141% over 2016. In the past year, the US incubator has absorbed more than 300 machine learning startups, a three-fold increase over 2016.
Machine learning will become the new normal in 2018, but AI startups will also greet the inspection period, with conceptual hype, no product and business model, and may be shuffled. Investors will be more cautious about investing in AI companies, and financing will become more difficult.
11. The pattern of corporate AI giants began to take shapeWith technology giants such as Google, Amazon, Salesforce and Microsoft improving their corporate AI capabilities, small businesses are hard to sustain.
Google released Cloud AutoML. Customers can bring their own data to train algorithms to suit their specific needs. Amazon began selling artificial intelligence as a service and "Amazon Artificial Intelligence" under its AWS. The goal of Amazon's AI is to serve large companies.
12. AI+ medical welcomes springIn the United States, regulators are considering approval of AI for clinical use. AI can provide earlier accuracy judgments in diagnosis, which can lead to the discovery and treatment of many conditions at an early stage. AI “watching movies†will also have advantages over human doctors and can be quickly popularized. In addition, AI entrepreneurship in healthcare is also entering the hottest stage.
Recently, Yingrui Pharmaceuticals and Biopharmaceutical Company AstraZeneca announced a partnership with Ali Alibaba subsidiary Ali Health to develop AI-assisted screening and diagnostic applications in China. GE and Nvidia have also teamed up to provide deep learning capabilities for GE's medical imaging devices.
13. AI enters DIY stageIn 2018, AI's technology and products will become more popular, and it is possible to create personalized AI in DIY. You don't need a Ph.D. in computer science or mathematics to build your own AI. In the open source software library, hundreds of APIs and SDKs can work, and giant companies such as Google will develop more and more simple packages, and barriers are lower than ever.
Google launched a project called "ArTIficial Intelligence Yourself". Its first product is the Raspberry Pi's speech recognition suite - users can send any voice they want to their personal voice assistant.
Ningbo Autrends International Trade Co.,Ltd. , https://www.supervapebar.com