This article by”
AI the front“Original, original
The value of AI must be on offense, not defense


Author | Geekbang CEO Huo Taiwen


Edit | Emily

AI Front Line introduction: On January 13, 2018, AICon Global Conference on Artificial Intelligence and Machine Learning technology was held in Beijing International Convention Center with the theme of “Facilitating the landing of ARTIFICIAL Intelligence”. Huo Taiwen, CEO of Geekbang Technology, the organizer, delivered a keynote speech titled “Industry Trends and Insights of ARTIFICIAL Intelligence” at the opening of the conference. The content covers industry application analysis, talent situation, government policy, code trends and industry development. AI front went live and brought you the full text of the speech as soon as it was over.”


As the moon is constant, as the sun rises.

Since the academic conference held in Dartmouth in 1956, artificial intelligence has come a long way. I think there have been countless ups and downs in the process, and there are also many untold stories. But at this moment, artificial intelligence this “old man” but once again with a new look to return to the public field of vision, standing on the top of the tide.

Whether it is mobile Internet, which emerged 10 years ago, or cloud computing, which started hot five years ago, I think their influence on human society is far less than that of ARTIFICIAL intelligence, because THERE are really too many industries that AI can penetrate and influence.

It is said that the development of artificial intelligence will usher in a new inflection point with the continuous consolidation of computing and data capabilities. Some people say that artificial intelligence is just like the impact of electric light on human beings in the industrial Revolution, which will be profound and far-reaching.

Looking back over the past year, AI has received unprecedented attention at the national level, in the media, in industry and academia. Remember Ng’s famous words: a hundred years ago, electricity could bring huge transportation, communication and agricultural networks to many enterprises and industries. Today, artificial intelligence will bring the same big changes to many enterprises. His words have been lingering in my ears, and also made me understand the meaning and value of artificial intelligence more intuitively.

Today I’d like to take the opportunity to share with you some of the industry trends I’ve seen over the past year and our insights.

First, let’s take a look at the trends of ai and machine learning-related terms searched on Baidu over the past five years.

From the trend, it is not hard to see the current attention of the whole society to AI and machine learning. So it’s fair to say that 2017 was the first year of artificial intelligence. I had a chat with an old friend two days ago. He said that nowadays Internet companies are embarrassed to call themselves Internet companies if they don’t talk about artificial intelligence when they go out. From our perspective, ai has been in the news almost every day for the past year.

Apple, Google, Microsoft, amazon, Facebook, Tencent and alibaba, among the world’s TOP20 companies by market capitalization, have All released their AI strategies and basically put forward “AI First” or “All in AI” vision plans.

AWS, Azure, Ali Cloud, Tencent Cloud, IBM Cloud, Oracle cloud these cloud enterprises have also opened their artificial intelligence capabilities on cloud computing. In April this year, Ma Huateng used a few words to clarify the relationship between cloud computing and artificial intelligence in terms of open capabilities. He said that cloud is the infrastructure for digital upgrading, while artificial intelligence is the cutting-edge products grown out of cloud. “Cloud + artificial intelligence” may be equivalent to “electricity + computer” in the future. Think deeply along this metaphor, you will find that it is indeed very appropriate, isn’t it because of the computer, just have the Internet era?

Turning to China, we can also feel the ai boom. Let’s start with a few numbers on hiring. This is a data report from dragnet.

In terms of the number of ai posts released, the number increased by 184.78% in 2016 and 25.85% in 2017 compared with 2015. In terms of the number of deliveries, the total number of deliveries increased by 127.16% in 2016 and 184.20% in 2017 compared with 2015.

From the perspective of education statistics of AI talents entering the industry, master degree is the main force of AI talents.

However, this does not mean that only those with a master’s degree or above can enter the FIELD of ARTIFICIAL intelligence. According to Data from Lago.com, the second largest number of AI talents are those with a bachelor’s degree or below, after 65.7% of those with a master’s degree.

In terms of talent distribution, Beijing is far ahead of other first-tier and second-tier cities. The capital’s unique technological atmosphere and policy support have encouraged more AI practitioners to conduct AI research and development in Beijing.

As for the requirement of working experience in ARTIFICIAL intelligence positions, there is little difference between 2016 and 2017, and the majority of people with 3-5 years of working experience are still in the minority, while those with more than 5 years of experience are still in the minority.

So how much do AI workers get paid? Most of the news I’ve seen this year is that AI engineers are highly paid and there’s a huge imbalance between supply and demand. However, from the pull check data, we can conclude that the hiring demand for AI technical talents in 2017 was not as strong as expected. As you can see from the chart, the growth of recruitment demand in 2016 was quite strong, but the growth of talent delivery in 2017 was quite strong. What does that mean? In a word, the demand of the market will precede the response of the talent, which is the jargon of recruitment. I don’t know if you understand, the demand of the market will precede the response of the talent.

There are still some data that I haven’t put out, you can read the AI front article, we will have this analysis. I talked to a couple of headhunter friends and came up with these conclusions:

Talent is of course important, but it is a great waste of talent if we do not give it a fertile ground to grow freely. As we can see from the data, so many AI technical talents choose Beijing and the capital. One of the important reasons is that Beijing’s policies guarantee the growth of talents in the field of artificial intelligence.

SoftBank chairman Masayoshi Son once said that ai will directly determine national competitiveness in the future. Putin was even more direct, saying that whoever becomes the leader in artificial intelligence will rule the whole world. And the Chinese government is definitely one of the first countries in the world to pay attention to the development of ARTIFICIAL intelligence.

Since 2015, The State Council and relevant ministries and commissions have issued a number of ai-related guidelines and action plans, which were even mentioned in the report to the 19th CPC National Congress. On December 14, 2017, the Ministry of Industry and Information Technology released the “Three-year Action Plan for promoting the Development of the Next Generation of ARTIFICIAL Intelligence Industry (2018-2020)”, which is more specific. It mentioned the key fields and technologies, and also listed the three-year goals. In short, it hopes to accelerate the development of the artificial intelligence industry. And be a world leader. In July this year, The State Council officially issued the “New Generation of ARTIFICIAL Intelligence Development Plan”, also clearly pointed out the three-step strategic goal of the development of the new generation of artificial intelligence, which is basically the national planning blueprint in the field of artificial intelligence.

As you can see, during the three years from 2015 to 2017, China has issued a number of documents to promote the development of artificial intelligence industry, especially in 2017, many documents directly took ARTIFICIAL intelligence as the title of the document, which shows the importance and strategic importance of artificial intelligence. Taking the recently released action plan as an example, we can see that the Ministry of Industry and Information Technology directly lists the action plan and action targets for the next three years, and also points out the direction of the key areas of artificial intelligence development. Generally speaking, the idea is clear and the direction is clear. The next step is to implement it.

The MIT Sloan Management Review is one of the top 10 general management journals in the world and a research tool for experts at the world’s top business schools. He recently surveyed executives and managers of more than 3,000 companies around the world and analyzed their use of AI.

As you can see from this chart, 75% of decision makers believe that AI can expand business boundaries and 84% believe that AI can help them maintain competitive advantages.

I talked with my friends before that many people think that artificial intelligence can help them save costs. I think this is right, but the direction is wrong. The value of AI must be on offense, not defense. This survey shows why.

But the ideal is full, the reality is very skinny. At this stage, now, now. The use of AI in the enterprise is just beginning. From the data above, we can see that only 5% of enterprises have large-scale application of AI in their products or business. But 80% of businesses haven’t even started yet. This also confirms that the change in hiring demand we saw earlier was not as dramatic as we expected in 2017. But I also believe that this round of AI will spread to the enterprise much faster than the last round of cloud computing. So I often say internally that AI is just getting started.

Let’s look at some analysis at the code level. GitHub previously released their 2017 numbers.

At the language level, we’ve seen Python grow dramatically over the past year. GitHub ranks programming languages based on the number of PR submissions in the last 12 months. In 2016, JavaScript continues to dominate the charts.

Interestingly, Python has surpassed Java to become the second most popular language in 2017, with a 40% increase in PR last year, thanks to the growth of artificial intelligence over the past year.

GitHub also counted the hottest project topics in 2017, and we were surprised to find that machine learning and deep learning topped the list as well. In addition, there are API, plug-ins, iOS, database, Bot, games, Library and other keywords.

GitHub also listed the most forked open source projects of the past year. As we can see, TensorFlow leads the list with far more forks than Bootstrap. If you look at some of the projects in the back most of them are front-end, they’re all Web. The fact that AI is able to bypass these projects with extremely rich application scenarios is a testament to developers’ interest in it. Of course, it also illustrates the importance of open source in the field of artificial intelligence.

Having said that at the code level, we’re moving up to the industry level.

PitchBook is a data analytics service that focuses on private equity, PE/VC/M&A. According to PitchBook, ai and machine learning received more than $10.8 billion in venture capital funding globally in 2017. Over the past decade, venture capital investment in the sector has grown dramatically: from less than $500m in 2010 to $5.7bn in 2016, 2017 nearly doubled from 2016.

Data show that Chinese companies accounted for four of the top five investment events in 2017.

In my opinion, the development of a field is inseparable from these three elements: first, the maturity of technology, second, the promotion of talents, and third, the development of enterprises.

These three factors influence each other and together form the basis of ecology. I have talked so much, summed up can be said that the application of artificial intelligence is just beginning, but its future is also very close. The impact of ARTIFICIAL intelligence on the industry, the society and the country is unprecedented. This is not hype. Because artificial intelligence can be combined with any situation we are in: the moments you are checking, the TV, fridge, washing machine you are using at home, the bus you are taking to the conference or the small car you are driving

AI is everywhere, I believe.

For more content, you can follow AI Front, ID: AI-front, reply “AI”, “TF”, “big Data” to get AI Front series PDF mini-book and skill Map.