Tech industry leaders warned on Monday that global market enthusiasm for the potential of artificial intelligence may be overheated, overlooking that the technology is still immature; compared with the problem of AI developing too fast, cybersecurity is a more urgent short-term issue.

Shou Zi Chew, CEO of TikTok, said at an event in Singapore that humans still retain advantages in many creative fields. He said humans still outperform AI in the nuanced processing of visual recognition and in long-term memory recall. At a separate event in Singapore, Microsoft executive Judson Althoff expressed a similar view and urged rational thinking.

Leading technology companies have recently adjusted their tone collectively. Despite having invested hundreds of billions of dollars in AI research and development and infrastructure, companies are beginning to lower market expectations and no longer hype AI’s huge short-term gains and potential threats.

At a conference held by DBS Bank, Chew said: “For one thing, I think the technology is not there yet. The human brain can accomplish many things that models still cannot achieve.”

Companies are now trying to strike a balance: on one hand, promoting and selling AI products, and on the other, not wanting the risks hidden in this transformative technology to scare away customers and the public.

Microsoft global commercial business chief Judson Althoff said at a Microsoft event in the same city: “I want to distinguish hype from reality. There is no path by which unlimited cognitive computing power replaces human value pursuits.”

Althoff urged the industry to earn the same trust that people place in plumbers or handymen who come to their homes for repairs. Over the past few weeks, executives from Anthropic and OpenAI joined other business leaders at an AI safety meeting convened by U.S. President Trump. Echoing their views, Althoff said: “Observability within organizations and at every layer of the technology stack is crucial.”

Chew, who leads TikTok, one of the world’s most influential social platforms, emphasized the need to fully recognize the cybersecurity risks posed by AI. He said such short-term risks are “real” at a time when companies are racing to deploy AI.

Source: Sina Technology