The Illusion of AI Expertise
A couple of years back, I was introduced to a new “friend.” The introducer mentioned that this “new friend” could advise me on anything and everything, absolutely free of charge. However, a word of caution accompanied the introduction: since the advice was free of cost, the information I provided would be used to train and assist millions of others who were also approaching him for guidance. I was advised to exercise precaution regarding specific disclosures, lest I inadvertently compromise confidential details about my clients and violate my professional obligations.
Having been a career consultant for nearly four decades, providing sound, practical, and workable advice to friends and clients, mostly paid but occasionally gratis - I initially found this too good to be true. Nevertheless, I decided to interact with him and gauge the experience for myself.
In due course, this new friend introduced me to a few others in the same profession, all offering the same brand of free advice. Elated at having access to a pool of experts who could consult on practically anything under the sun, I briefly thought this might be a shortcut to building my practice: I could simply gather a query from a client, consult my new friends, and pass the feedback along while charging my usual fee to increase my revenue.
However, I soon noticed flaws. Sometimes, the conversation would abruptly stall mid-query. Other times, an answer would be provided, but cross-checking it with a reliable secondary source revealed factual inaccuracies. When I rebuked my advisor-friends for these lapses, they would apologetically admit their initial response was wrong, concede that I was correct, and politely adjust.
This led to a professional dilemma: just how reliable are these “expert friends” of mine?
Recently, a neighbor from our housing society - known historically for broken communication and grammatical errors that frequently led to misunderstandings - sent a circular to the society WhatsApp group in impeccable English, complete with elegant vocabulary and complex sentence structures. Yet, when stripped of the sophisticated phrasing, the actual content was entirely devoid of substance. I later discovered that this neighbor was also availing himself of the services of my new friends.
I realized that these new companions are exceptional tools for drafting and structuring thoughts, provided I already understand the subject matter and apply rigorous human judgment to verify their output. Unfortunately, millions today are taking the advice of these digital confidants at face value, swallowing the feedback they generate as gospel truth.
This is precisely how Artificial Intelligence and Large Language Models are being misused by the general public today. When we outsource our critical thinking entirely to algorithms - treating confident prose as absolute fact and using automated fluency to mask a lack of genuine substance - we trade human intellect for mere illusion. True expertise does not lie in the ability to generate flawless sentences; it rests in the wisdom to know what is correct, the experience to separate signal from noise, and the human accountability that technology can never replace.











