I must have come across this wording (the title of this post) somewhere in my reading in the recent past, but it got stuck in my brain.
What a succinct way to describe how automated systems can eventually take over our lives?
It has already happened via social media over the past decade or so, but it is now scaling rapidly via the deployment of autonomous AI (Artificial Intelligence) agents.
AI is firmly taking hold of our lives. Even people who reject that this is another fad are being influenced (sometimes unconsciously) by AI as it starts to pervade all facets of life, all around the world.
I am not talking here about the potential dangers of AI taking over autonomous weapons in the battlefields of the future, or AI causing bio-terrorism and cataclysmic climate changes. While all these are possible, only very foolish governments would let AI agents roam free when it comes to existential threats. Humanity may not be under immediate threats of extinction, notwithstanding the statements of the OpenAI safety researcher who quit a few weeks ago and went public about the extinction of humanity itself. No one can laugh it off, the threat is real. With the lack of guardrails on AI development it is only a matter of time before some perceived threat actually becomes a real one about which the whole world would feel completely endangered.
My post is however on what we usually consider as mundane activities.
Can my beliefs and behaviours be influenced by AI? Of course, yes.
AI can make me believe what it is communicating as the real thing which I should look at carefully (!). The persuasion becomes automated, and no longer based on what close friends communicate. AI takes precedence over friends and family, as it sounds rather real and very factual. Any number of queries that I have launched on Claude specifically have been returned with accurate responses (since I already knew the correct answers), and I keep testing Claude. There was a time when it apologised after I challenged its wrong answer. I am sure it learnt from that experience, as when I tested it on the same query using a different wording on a different Claude account, it did not give the same wrong answer. The problem with AI is that it gives what appears to be convincing answers to almost anything under the Sun, so you tend to believe them more than non-AI humans who you have trusted all your life! This automation of persuasive ability is very real, and often dangerous.
What about the structured decision-making steps that most of us are used to on any major issue in life? Those steps have been tested by us on numerous occasions all through our life. But now we are surrendering those thinking and thought-provoking steps to AI, which we believe to be more brainy than us! Yes, AI can gather together all the worldly wisdom and thoughts in under 10 seconds or even less and present its analysis and conclusions without us belabouring through our personal thoughts. Sure, it is a good productivity tool. But unless we already have thought about the issue, and know more or less the contours of the possible answer, we would not really benefit from AI’s responses. What good are we if we cannot ourselves figure out a possible solution to a problem?
I treasure my autonomy and critical thinking ability, and so should you. These are skills which have been honed over the years, and I have no intention of surrendering these skills at the altar of AI. My queries would be more like “can you explain why global bond prices are going down, and why interest rates are going up, give me an easily understandable explanation in layman terms, and also highlight what could lead to more instability in global financial and commodity markets in the coming three months?”. I can ask Google itself which would use its Gemini AI to answer. Or I can ask ChatGPT. I always found that long queries with constraints and guardrails in the query itself makes AI think and come up with a long thoughtful response (it takes a few more seconds if you can check for yourself!). I also ask “do you think you can take over my brain’s thinking on complex issues facing my life and run my life for me, managing my finances, and in the process do you think you can do a better job than what I have been doing till now; have you provided similar advice to other humans?”.
I don’t want to be persuaded by AI on any topic. I do not allow AI to advise me on the content of my blog posts, as I do not trust AI to do as good a job! Yes, that’s what I think, and I have no intention of changing my view on this matter. There are areas where AI can help, and we need to figure that out very carefully. So far I have only seen common people using AI as advanced search agents. I would not ask an AI agent to make my flight booking, write my emails or manage my calendar. All these and more are done best by me as I keep constantly juggling my priorities. I don’t want to trust AI on all these personal matters.
When it comes to emotional persuasion involving feelings which appeal to the heart rather than the mind, my sense is that AI would fail (I have not tested this), though I see news coverage on how AI agents are providing “love” to anyone not experiencing love and feeling rather lonely. But this is nothing but fake love feeling to which a normal human being with normal feelings would not succumb. But the world is changing, and I wouldn’t be surprised if AI Love Agents eventually proliferate.
Social persuasion is the most dangerous thing that AI agents are capable of causing in a moment all across the world. Youngsters are very prone to social persuasion by AI. This is a serious danger which needs to be curtailed. Peer pressure can never be the reason to surrender to AI.
A lot more can be written on this rather interesting and relevant topic.
Have a good week ahead,
Cheers,
Vijay Srinivasan
4th October 2026