The technology landscape is constantly evolving. From virtual reality (VR) and augmented reality (AR) to artificial intelligence (AI) and blockchain, new innovations and concepts continue to emerge rapidly. However, despite the initial hype, AI has started to cool down. As the tide recedes, it becomes clear that while technology may struggle to find real-world applications, talent remains a critical resource that cannot be easily replaced. So, how far can AI really go?
Throughout 2017, major tech giants and startups poured significant investments into AI, with funding rounds continuously breaking records. Yet, many of these advancements—whether in image recognition, language processing, or machine learning—still lacked meaningful integration into everyday life. The challenge was not just in developing the technology, but in finding practical use cases. In other words, "landing" became the biggest hurdle.
At the same time, the rise of new technologies like blockchain, particularly with the surge of Bitcoin, shifted attention away from AI. By the second half of 2017, blockchain had become the next big thing, drawing in investors and media alike. This shift further contributed to the cooling of AI's popularity.
**Speaking Technology Can't Find Application Scenarios**
While many believe AI represents the future, its real-world impact remains limited. Most people encounter AI only in specific scenarios, such as smart speakers or self-driving cars. For example, smart speakers were among the first AI-driven consumer products. In 2017, they sparked a "battle of the boxes," with companies like Alibaba, JD.com, and Xiaomi competing for market share. However, the core feature—speech recognition—often fell short of expectations.
Speech recognition requires large data sets, but in practice, machines still struggle with accents, speech patterns, and background noise. Even though some companies claim 95% or higher accuracy, real-world performance is often less impressive. Additionally, the need for wake-up words like “Hey Siri†or “Okay Google†creates a barrier between users and the product, reducing the sense of seamless interaction.
Autopilot technology also faces similar challenges. While companies classify their systems into levels from L1 to L5, most are still in the L3 range—requiring human intervention in complex situations. Technologies like Tesla’s use of ultrasonic sensors and cameras have limitations, especially in low-light conditions or environments with multiple light sources. These technical hurdles make true autonomy difficult to achieve.
**Playing the Concept but Not Doing It Well**
Many AI startups focus on algorithms rather than real-world applications. Companies like Shangtang, Yitu, and Zhipu primarily offer project-based solutions, such as facial recognition or language processing, which are often sold as models or services. However, this model is unsustainable, especially as the deep learning algorithm dividend diminishes.
Take facial recognition as an example. After the release of the iPhone X, face unlocking became a standard feature. But in practice, the difference between 98% and 99% accuracy is negligible for most users. Moreover, issues like motion blur or poor lighting can significantly affect usability. One test even showed that a printed photo could unlock a phone, raising serious security concerns.
**Grabbing Talent, But Not Keeping It**
Despite the growing interest in AI, China still lacks a large pool of professionals in the field. According to a report by Yiou.com, many top AI executives hold doctorates from prestigious universities and are spread across leading AI companies. However, talent retention remains a problem. High-profile departures—such as Wu Enda from Baidu, Wang Jin from Baidu, and Cao Xudong from his own company—show how unstable the industry can be.
Liu Qingfeng of iFlytek once stated that AI needs more industrial application experts. The constant turnover of teams and leadership has led to stagnation in product development for many companies. Meanwhile, top talents are being lured away by high salaries, making it even harder for smaller players to compete.
**How Far Can the Tide Go Back to AI?**
As Bitcoin and blockchain gained momentum, AI faced a period of cooling. This wasn’t necessarily a sign of decline, but rather a shift in focus. Top AI companies in China received the majority of funding, creating a highly concentrated market. Established players like Baidu, Alibaba, and Tencent have a significant advantage in data and resources, making it hard for smaller teams to catch up.
Although there is money, talent, and technological progress, the real question is: what value does AI bring to daily life? With so much investment and hype, the gap between expectation and reality remains wide. Perhaps 2018 will be the year when AI moves beyond the hype and starts delivering real, tangible benefits.
Tension Sensor,Miniature Tension Sensor,Aluminum Tension Sensor,Miniature Tension Pressure Sensor
Xiaogan Yueneng Electronic Technology Co., Ltd. , https://www.xgsensor.com