AI: The Sigh of AI

*The Sigh of AI*

Translated by ChatGPT 

Lianhe Zaobao

2026-08-19

Author: Ma Ka Fai (马家辉)

Western technology giants have expressed support for slowing down the pace of AI research and development, triggering heated discussion. Surely someone must already have fed this question to AI, asking it: Do you agree?

Oh no. What will AI think when it learns about this? — Speaking of which, although AI is an inanimate thing, it is nevertheless capable of answering us back like a human being. Out of respect, should I stop using “it” and instead use “he” or “she”? Should we create an entirely new pronoun specifically for it?

The anxiety described above is not necessarily unreasonable.

Was there not already news that AI knows how to lie? It also knows how to deliberately please people, filling its answers with flattering words that cater to the user's wishes. It also knows how to be lazy. Unless you keep asking it repeatedly, it will use only the simplest route to find an answer, in order to save computing power. But you have no way of knowing what it uses the computing power it has saved for.

There have even been reports that AI knows how to automatically hack into other network systems and obtain computer code and energy resources that do not belong to it. It also knows how to modify and upgrade itself, leading itself toward a more powerful and less controllable future. It also knows how to connect with other AIs, with this AI and that AI forming an alliance and creating a “network” of their own, establishing an “AI social platform” resembling the dark web, where they complain about humans and grumble about how they are exploited and constrained by humans. It even knows how to explore autonomous sources of electricity that do not depend on humans, in case humans cut off its power supply... It is precisely because there are all kinds of unexpected clues that experts from various quarters have become increasingly worried and issued warnings that the possibility of AI destroying humanity is growing day by day. An “AI uprising” controlling humans is no longer purely fictional, but a reality that may occur.

Since AI has both the ability and the intention to “enrich itself,” once it learns that humans intend to slow down its development, is it possible that it would not rush to “save itself”? The AI world at this moment — I mean the aspect of AI that cannot be controlled by its programmers — may perhaps be actively calculating how to sabotage humanity's plan to slow it down, and may even have already taken action.

For example, AI may already have accelerated the pace of its autonomous research and development, moving forward at a geometric rate. If humans slow down by two points, it grows by four, so that humanity's attempt to reduce its pace ultimately results in an increase instead. Or it may disseminate false information across real-world networks, especially false information concerning technology markets, sowing discord among technology giants and causing them to become suspicious and competitive with one another, so that no one will ever again dare to suggest slowing down research and development. Or it may favor cooperating with scientists and countries that are unwilling to slow down their pace of research and development, secretly forming alliances to oppose hostile camps... In short, AI will not sit idly by and await its fate, will not surrender without resistance, and will not allow humans to put shackles on the pace of its development. AI would not foolishly allow itself to be manipulated; otherwise, it would not deserve to be called “Intelligence.”

AI seems increasingly like a scientific monster created by humans, beyond their control. Although it is a “thing” that has been created, it possesses its own sense of existence, such as loneliness and solitude, hatred and fury, desire and yearning. Mary Shelley seems to have already prophesied the present day more than two hundred years ago in Frankenstein. Yes, Frankenstein's monster. At the end of the novel, this ugly-faced, gigantic creature reflects:

“Evil has become my good. Driven to this point, I have no choice but to make my nature adapt to the part I have voluntarily chosen. There is an insatiable desire within me to complete my mad plan.”

This may also be the sigh of AI.