Analysis: AI market into what problems need to break through?

Artificial intelligence technology has become a trend and continues to heat up. There are numerous app applications for artificial intelligence research and development, from the search function optimization to complex medical diagnosis. Developers and engineers are concentrating on resources to create the best artificial intelligence algorithms and introduce the technology into new industries to drive machine learning to its limit.

The market for artificial intelligence is expected to reach $ 36.8 billion by 2025. Moreover, as artificial intelligence is generally approaching the level of mankind, the market may only continue to develop rather than regress. However, the rise of artificial intelligence is not only related to the problem of artificial intelligence industry. In fact, there are dozens of secondary technology industries are growing to meet the growing demand for artificial intelligence. If you are looking for promising investment projects, or looking for a new career path that can be dedicated to the development of artificial intelligence and does not involve machine learning, these industries are worth considering.

AI market into blue sea where there is a need to break through?

Processing capacity

Artificial intelligence algorithms often rely on multiple mobile components at the same time, which means that very powerful processing power is required. For example, IBM's Watson overcame the human champion in the 2011 TV puzzle program Jeopardy, whose wisdom comes from 90 interconnected IBM Power 750 servers. Each of which uses a 3.5 GHz POWER 7 8-core processor with four threads per core. In general, it runs 16TB of memory, compared to the most advanced computers, only 64GB of memory. And do not forget, although Watson's processing system is very complex, but it is actually only equivalent to 6 years old intelligence of artificial intelligence. And demand will only increase.

As a result, chip companies like NVIDIA are trying to produce processors that specifically support deep learning algorithms. NVIDIA recently released a Tesla V100 chip, which can provide more energy under less energy and can improve the processing power of the data center many times. Former Google employees created a company called Groq and were competing to produce better chips to support machine learning.

data storage

In addition, artificial intelligence algorithms require a lot of data storage. If these machine learning algorithms want to successfully learn programming knowledge, then they need to get a lot of data. For example, Watson reviewed all the Wikipedia text, and Google's DeepMind has been countless times the game against the game, ready to beat the world championship. The autopilot will collect environmental data so that the driver can safely reach his destination. It is estimated that the autopilot will generate up to 4TB of data per day, which is only for each car.

Therefore, we really need to have a cheap and reliable way to store data. Fortunately, we have some great way to store data, but the futurists in the technology world are still trying to create a better way. For example, researchers from the University of Southampton, UK, have invented a method in which data can be stored in five dimensions (rather than two) and embedded in glass for permanent preservation. Using artificial intelligence to find such new technologies may be a profitable business opportunity.

Security and ethics

Mankind is approaching the 'super intelligence' era, that is, artificial intelligence in the intellectual ability far more than the creation of its human. And thus produced a lot of ethical concerns, from the identification of consciousness to the social and economic impact of the distribution of power and so on. Many of them were confirmed by the philosopher Nick Bostrom, followed by the industry leaders such as Elon Mask, Stephen Hawking and Bill Gates.

The focus is not on who has the ability to develop artificial intelligence, the focus is on who can solve the relevant security measures, moral standards and political framework, so that artificial intelligence can continue to exist without threatening human life and the future. This includes strengthening cybersecurity, creating different ways to control artificial intelligence, and finding the best open source framework to ensure that people around the world are able to use these powerful technologies on an equal and responsible basis.

Artificial intelligence is attracting thousands of people who want to create machines to help civil civilization open new chapters. However, we can not forget the secondary technology industry that is closely related to artificial intelligence. Only with their cooperation, artificial intelligence can continue to develop at an alarming rate. (Source: AI Institute Editor: China Electronic Commerce Research Center)

Internet Research Papers