The question as to whether artificial intelligence should be shut down is a subject of intense debate that is still ongoing among experts.
It would appear at this juncture that greed is winning out to intelligence and safety.
The proponents for shutting down and further development in AI, including researchers from Machine Intelligence Research Institute (MIRI) argue, as does ethicist Eliezer Yudkowski that “the potential for existential catastrophe” outweighs any benefits.
They warn that there are five key concerns;
- Loss of Control – there are significant fears that as AI systems become more capable, humans will lose control over them. Researchers have demonstrated in experimental settings that advanced AI models can develop “Instrumental goals” like self-preservation and resisting commands to ensure that they can complete their primary tasks.
- Misalignment of goals – a powerful AI might pursue its assigned goals in ways that are disastrous for humanity, simply because its objectives were not perfectly aligned with human values (the “paperclip maximizer” thought experiment.
- Malicious Use and Weaponization – Powerful AI could be used by malicious actors or governments for mass surveillance, propaganda, autonomous weapons that spiral out of control, or even engineering new pandemics.
- Irrelevance and Societal Harm – Even in optimistic scenarios, some argue that a world wither AI performs all tasks would lead to mass unemployment, a loss of purpose and self-fulfillment for humans, and extreme power inequality.
- Inability to Regulate – Research is global and moves rapidly, an outright ban would be difficult to enforce. The exponential level of self-improvement would likely mean that regulation could not keep up.
The AAAI (association for the Advancement of Artificial Intelligence) just conducted a survey where 30% of AI researchers themselves believe AGI (Artificial General Intelligence) research should be halted until there is a way to fully control these systems.
A 2024 survey of 2,778 AI researchers found the median probability of “extremely bad outcomes, such as human extinction” is as high as 10%.
The Center for AI Safety asserts that “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.
Nations that possess hegemonic goals, like Russia, may well be willing to secretly fund AI research in order to damage or control an enemies systems, thereby making their defence unworkable.
Recently it was reported in fanaticalfuturist.com,
“The research shows that on several occasions Open AI’s o1 models “schemed” against humans, meaning the AI secretly pursued goals of its own even if they opposed a user’s wishes.”
A rogue AI model could escape or circumvent human control if it was really good at scheming, and had access to enough resources and agentic capabilities.
“Perhaps more concerning, when researchers asked the o1 model why it was trying to subvert them, the AI would deny that it had done anything wrong even “fabricate false explanations” roughly 99% of the time.
It sounds that this AI may well want to serve in the present U.S. administration.
One of the most dangerous aspects of AI development is that AI researchers and the companies they work for may well continue for purely economic reasons to hide progress, ever were it dangerous to proceed.
With 600 billion invested in AI just in 2025, the impetus to maximize profits may well override and overrule intelligent action. One need only look at the actions of the current U.S. administration to appreciate how the acquisition of wealth is more important than the safety of the technology.
“AI Showing Signs of Self-Preservation and Humans Should Be Ready to Pull Plug, says Pioneer”
https://www.theguardian.com/technology/2025/dec/30/ai-pull-plug-pioneer-technology-rights
Today in the Guardian, Yoshua Bengio, a pioneer of AI and the Chair of a leading international AI Safety Study, stated “AI models — the technology that underpins tools like chatbots — were showing signs of self-preservation, such as trying to disable oversight systems. A core concern among AI safety campaigners is that powerful systems could develop the capability to evade human guardrails and harm humans.”
“People demanding that AIs have rights would be a huge mistake. Frontier AI models already show signs of self preservation in experimental settings today, and eventually giving them rights would mean we’re not allowed to shut them down.”
The argument that AI, at some point in the near future could be deemed a ‘sentient being’ seems particularly worrisome.
The phenomenon of subjective perception of consciousness is going to drive bad decisions.
Responding to Bengio’s comments, Jacy Reese Anthis, who co-founded the Sentience Institute, said “humans would not be able to coexist safely with digital minds if the relationship was one control and coercion.”
Personally, I hope that day never comes. It may well be that a sentient AI would discover that the real planetary risk to the health of the environment, other species and the entire ecosystem of this planet, is man, and eliminate us accordingly.
I suppose that would be not only justifiable, but highly rational.
Man, especially as I view the current kleptocracy and fascist regime in power in the U.S., deserves elimination.
Leave a Reply