The following are important things to note about the report:
- The rapid progress of natural language processing benchmarks GLUE and AI2 inference challenges can answer multiple-choice questions at the primary level.
- The number of published papers in China has increased. The number of papers published by authors related to the government has increased by 400% compared with 2017, and the number of corporate AI papers has increased by 73%. In contrast, the growth of published AI papers in the United States mainly comes from such as Google. Nvidia andMicrosoftAnd other technology giants.
According to the report, the number of research papers published is the most in Europe, followed by China. According to Elsevier's report released this week, the number of papers published in China will lead the world in less than five years.
- Artificial intelligence is a global industry, Scopasdatabase83% of the published papers are from outside the United States.
- The annual AI conference NeurlPS, ICML and CVPR have thousands of participants.
- In terms of AI-related patents, the United States continues to lead; AI startup financing has increased by 450%, compared with 200% in other areas.
- Currently, more than half of the AI member partnerships are non-profit organizations such as ACLU and the UNDevelopmentPlanning Department, etc.
—TensorFlow remains the most popular machine learning framework.
As more and more companies want to apply artificial intelligence, the number of AI and machine learning mentioned by major public companies in their telecom reports has also risen. This indicator reflects the way companies change their way of talking about artificial intelligence. According to a report released this week by Tata Consulting, 46% of organizations have begun to apply some form of artificial intelligence, but applications are not the same as successful applications, and smart companies are not just discussing AI, they are also looking for ways to Successfully deploy AI within the enterprise.
Although there are still companies that use AI to blow big bangs, today's air pledges are no longer effective. Former Baidu AI chief scientist and Google Brain co-founder Wu Enda recently said that irrational AI hype around AGI and killer robots is no longer as common as it used to be, which is gratifying. Understanding the AI's can and cannot be done helps reduce the public's fear of AI.
At the moment, there may still be a significant number of startups and companies that claim to be artificial intelligence companies trying to prove their worth in the AI space. But more and more, claiming that AI is far from enough, strength is the key. Companies must have the ability to demonstrate that artificial intelligence can bring a virtuous cycle to gain a competitive advantage.
The most sensible companies are developing intelligent machine systems while not forgetting to build trust, quickly transform consumer sentiment, and keep in mind the diversity of employees and the diversity of values.