Modern deep-learning-based Artificial Intelligence (AI) is drastically different from the technology that preceded it over the past three decades. New AI is not merely theoretical or conceptual. Its practicality and applicability has been proven by various applications, real-life scenarios, and large-scale data sets — and even though AI developments over the past five years have met expectations, many are still worried the new AI will disappoint (much as AI has in the past). However, I believe a new AI era has just begun.
Unparalleled AI Technology in an Era without Authority
Recent years have seen numerous reports on AI, and there has been a lot of discussion about the growing popularity of AI and its impact on scholars, entrepreneurs, and investors. Generally the discussion focuses on three topics: significance, industry players, and application scenarios.
As a scientific researcher and entrepreneur, I would like to share my thoughts from two perspectives. 1) Only industry leaders will be able to precisely understand and push the leading edge of AI, and 2) Top enterprises can leverage technological foresight to gain momentum.
AI is not easy to evaluate. Judging between true and false, or differentiating between bubbles and real success is difficult. AI is a science, and hard for the general public to comprehend. Even the best AI labs in the industry find it difficult to evaluate each other’s work, as interpretations are often based on biases and presuppositions when discussing objective scientific topics.
Remember, this is an era without a clear consensus or established subject matter experts. AI is developing in leaps and bounds everywhere, and many top AI labs in the US still haven’t developed a comprehensive understanding of the technology.
Even an AI director at a global search engine may fail to fully understand all the advancements of every department in the company. But a mere decade ago only three top professors in the field were needed to determine how AI is affecting the world. They could even accurately forecast what would happen in the next three years.
Today, however, is an era with no clear AI authority. Any new technology might deliver groundbreaking social changes and generate profound economic value.
AI Helps Us Understand the Limits of Human Intelligence
AI should be able to acquire the abilities of a 3-year-old; that is, it should be able to attain the most important human characteristics. In many fields, AI has already surpassed human performance. Machine image recognition surpassed that of ordinary people nearly four years ago, marking a significant tipping point for AI. Machines outperform humans in many of the most basic identification functions. For example, in terms of facial comparisons, machines outperform humans when identifying both familiar and unfamiliar faces. Machines can find your face among 1.3 billion others faster than your own mother. Machines might even know your significant other better than you — and in professional fields, machine identification may outperform experts with years of training.
Current AI helps us recognize the boundaries of our intellectual abilities.
Facial Recognition Applications Have Global Significance
Currently, China is pioneering the application of advanced AI in vertical scenarios, solving not only Chinese problems, but problems around the world. The significance of China’s AI application scenarios will be felt around the world.
For instance, thanks to YITU’s AI applications in law enforcement, the technologies police previously hoped to use have become available in the last two years. China alone now possesses technologies capable of the advanced functions that were once only possible in science-fiction movies.
Despite the rapid development and popularity of AI, industry insiders such as BAT (Baidu, Alibaba, and Tencent), entrepreneurs, and investors still underestimate the great possibilities of cutting-edge AI technology.
AI Technology Dramatically Accelerates Business
Impossibilities of past decades are now a reality. Industry-level experimental data gives us insights into the far future. Today we say AI is popular because the overall results give us confidence.
Businesses continue to unlock AI application scenarios. As the technology is commercialized, the integration of AI and business operations will accelerate. The current level of intelligence on display in some applications is beyond our imagination, which has failed to keep up with the speed of AI advancements.
Meanwhile, technology’s societal impact is an oft-discussed topic, which will only become more sensitive as AI surpasses human intelligence. The growth of artificial intelligence will change the division of labor tremendously, which is a problem confronting every modern nation. Many existential concepts need rethinking as AI changes the way we conceive of ourselves. Our concepts and models of the social division of labor and economics will undergo dramatic changes over the next two decades.
Technological Foresight Is Paramount to AI Developments
AI by itself is not a product. Scientists have no special advantages in AI development; but instead, a technological understanding of business scenarios is the key. YITU’s industrial deployment prioritizes the understanding of people, followed by the understanding of products and technology. We will continue learning, and aim to understand our customers (such as police officers or doctors) better than themselves.
Technological foresight allows us to understand user data locations and its commercial value, so we can accurately predict further development. The gap between the top players in the industry may grow — especially between the very best of them — because development speed differs greatly from those who believe and those who don’t.
Absent historical precedent or clear AI authority, we should work hand-in-hand with partners on this shared journey into the unknown. AI can even revolutionize all other technologies, including scientific disciplines like human-machine interaction, searching, robotics, microprocessor design, medicine, and pharmaceuticals. The past five years have seen explosive AI developments, which have only intensified our plans for the future. AI will soon enable multi-dimensional technologies to build on application scenarios, bringing us more dynamic experiences. From a multi-technological perspective, machine vision, hearing, semantic understanding, and motion control will quickly progress in the next few years. Terminal intelligence will combine with chips to penetrate the last barriers of user experience and interaction. The leap from the Internet of Things to the Internet of Intelligence will make intelligence ubiquitous.
Enterprises in a better position enjoy the advantage of time. Cumulative effects brought on by technological leadership and more pluralist AI technologies empower AI to be disruptive across industries. Such advantages not only determine the gap or positional relationship between the first and second players in an industry, but also enable leaders from AI-advanced industries to disrupt industries undervaluing AI.
We believe the potential of AI is unparalleled because we have seen it, and YITU is embracing this unprecedented era together with our partners.