简体
CHINA'S LITHIUM PIONEER
ENGINEERED TO EMPOWER
Provide comprehensive batterysolutions for full scenarios
In an environment of “survival of the fittest,” AI companies should steadfastly uphold their core values.
Time：
2019-03-07
Building a car is a money-burning endeavor, especially when it comes to self-driving vehicles. Before mass production can begin, numerous technical challenges still need to be addressed. The driving scenarios are far too complex, and…
Automobile manufacturing is a capital-intensive endeavor, particularly in the case of autonomous vehicles. Achieving mass production still requires overcoming numerous technical challenges. The driving scenarios are exceedingly complex, and current algorithms are far from achieving full autonomy.
Although machines are increasingly performing well on specific tasks, they clearly do not yet possess general common-sense reasoning abilities. They have not reached the same level as humans; their capabilities are limited to ever-improving object recognition—and nothing more.
At its core, this machine remains under our control. Computational “errors” can lead to fatalities—just as people have lost their lives in the deployment of autonomous driving and medical automation.
Last year, an Uber self-driving vehicle struck a woman who was crossing the street. Subsequent investigation revealed that the autonomous-driving software initially misidentified the woman as an unknown object, then mistakenly classified her as a car, and only finally recognized her as a bicycle. The woman ultimately died.
The head of Waymo and a leading figure in the autonomous-driving industry has personally acknowledged that self-driving cars will enjoy widespread adoption on roads for decades to come; meanwhile, Silicon Valley co-founder Steve Wozniak has stated quite bluntly: “Self-driving cars are simply impossible to achieve in the short term—I simply don’t believe ‘autonomous driving’ can ever be realized.”
Capital flows in quickly but also out just as fast, and we still have a long way to go before true realization is achieved.
Artificial intelligence itself is not a bubble, but AI startups are.
Today, in the burgeoning field of artificial intelligence, some startups are already valued at several billion dollars, signaling that the industry has long since surpassed the early days of the internet era.
Some industry insiders believe that four major bubbles are currently affecting the development of artificial intelligence:
The Tech Bubble: Artificial intelligence is an interdisciplinary, application-oriented field with a high professional threshold, yet truly valuable knowledge is exceedingly scarce.
Capital bubble: There’s too much capital and too many concepts, the chasm is glaringly obvious, yet everyone is still waiting and watching.
The AI Bubble: Today, much of what is labeled as “artificial intelligence” is little more than pseudo-intelligence. Most companies still do not know how to leverage intelligence and data to serve customers and create value.
Valuation Bubble: It’s not difficult to estimate the value of hundreds of millions of companies; the real challenge lies in identifying sustainable profit drivers. At present, most AI applications remain at the stage of technological tools, still some way from achieving platformization and productization, and have yet to generate revenue.
Given the high professional barriers to entry, there are very few individuals who possess both technical expertise and industry knowledge in the field of artificial intelligence, which contributes to generally high salaries in this sector.
Recently, Hu Yu, Chairman of the Board of Trustees of the University of Science and Technology, stated in an interview that the annual salary for a master’s degree in this industry is 400,000 yuan, while a doctoral degree can command as much as 800,000 yuan. To qualify, one must have at least two years of experience in the industry and a total work history exceeding five years. With exceptional performance, annual compensation can even reach 2 million yuan. For former directors of research institutes, the salary level can soar to between 5 and 6 million yuan. In fact, some employees at our company are already earning more than I do as the university president.
Capital’s expectations for artificial intelligence are excessively high, inevitably driving up the value of AI talent—and ultimately passing these increased costs on to AI companies. According to Hu Yu, some firms are now considering how to reduce their AI workforce due to the unsustainable cost burden.
Where there is investment, there must be returns. Hu Yu notes that, at present, AI companies are still not profitable. Everyone is talking about how to make money—or even how to make less—but in reality, very few companies can actually generate sustainable profits from AI. This, in fact, is a bubble.
AI companies are forging ahead amid a capital bubble. At present, homogenization is rampant in the AI sector: many firms lack genuine innovation and merely follow trends blindly. The only way forward for AI companies is to deflate the bubble and pinpoint real pain points.
In the overarching environment of “survival of the fittest,” AI companies should remain grounded, build up substantial capabilities over time, and relentlessly pursue innovation. By leveraging data and technology, they can explore more effective commercialization pathways, better serve users, and ultimately realize their own value.
Prev：
Tel:+86-27-83258996 ext: 8015
Email: export@whforte.com
Email: marketing@whforte.com
Add: Forte IndustrialPark, Wujiashan Taiwan Business InvestmentZone, Wuhan,430040, China