The phrase “outsourcing of thought” has been appearing frequently in media and social media lately. It seems to carry a warning message: in the age of artificial intelligence, we are delegating our thinking to machines. But let me ask a question. When we use these terms, are we truly sure about what we are doing?
Why Are We Drawn to New Terms?
When a new term emerges that seems familiar or somehow appealing, most of us might respond in this way. Rather than examining its precise meaning, origin, timeliness, or relevance, we uncritically accept it, thinking, “Is this a new trend? If so, shouldn’t I know this much?”
Why does this happen? Perhaps it stems from two emotions within us: the intellectual satisfaction of understanding a new concept, and the anxiety of not wanting to be seen as someone who doesn’t know even this much. So, before even questioning whether the term accurately describes recent phenomena or whether it’s even a viable concept in the first place, we accept it and use it more often than anyone else, fueling our intellectual vanity, without a doubt.
So, let’s examine this recent buzzword before us. Does the expression “outsourcing of thinking” truly capture the phenomena we’re experiencing, or is it just another buzzword with a catchy sound?
Is “Outsourcing” Really Outsourcing?
First, let’s consider the word “outsourcing.” When a company outsources marketing or software development, there’s a clear structure. There’s a client who commissions the work and a vendor who delivers it. The client holds the core strategy and direction, while the vendor executes according to those specifications. Once the work is complete, the client rigorously inspects the deliverable for quality before making a final decision.
So, are we, who entrust our thinking to AI, truly in the position of the client, outsourcing our thinking to AI?
After asking questions to an AI chatbot, do we meticulously examine the accuracy and appropriateness of the answers it provides? Do we check whether it aligns with our original intent, whether the facts are correct, whether the logic holds together?
Perhaps we are simply accepting the “plausible answer” uncritically, satisfied with it?
If it’s the latter, is this really outsourcing? When we accept results without exercising critical judgment, without retaining the core decision-making power—isn’t that something else entirely?
Does AI Actually “Think”?
Here’s the second question. How do you define the word “thinking”? Are current AIs truly capable of thinking?
Consider how current large language models (LLMs) work. They calculate which words are more likely to appear next in a given context, based on patterns learned from massive amounts of data. They do this through mathematical tools such as ‘linear algebra’ and the computational power of state-of-the-art GPUs.
Can we call this “processing” as “thinking”?
What does it mean for humans to think? Isn’t it a process of understanding a problem, weighing various possibilities, and making judgments based on our experiences and knowledge? Can a machine that doesn’t “understand” the meaning of its output, but instead generates an optimal approximation based solely on statistical patterns, truly be called a thinking system?
Omnipotent AI that grasps the meaning and essence of everything humans think and then proposes solutions beyond human reach ultimately exists only in movies and science fiction. At least for now.
The LLM (Large-Scale Language Model), which we so admire and use today, is not a machine that “thinks” at all, as we define it.
So What Are We Actually Doing?
When we bring these questions together, an interesting picture comes out.
The term “outsourcing” inherently presupposes a clear ability to order and inspect. However, in many cases, we simply ask meaningless questions to AI models I without even knowing what to ask, and then accept the plausible answers they give us without any verification.
The term “thinking” presupposes understanding and judgment. However, LLM does not engage in the process we typically call “thinking.”
So what does “outsourcing of thought” actually mean?
A situation where someone lacking the ability or will to think asks a non-thinking entity to think for them? Can this truly be called “outsourcing”?
Let’s consider this. If you were a project manager at a company, would you choose an outsourcing vendor who simply boasts of their capabilities without understanding the implications of the RFP you gave?
What’s the Real Problem?
Perhaps the very phrase “outsourcing of thinking” is evidence of the problem we face.
We have already surrendered much of our attention and judgment to YouTube algorithms, social media feeds, and recommendation systems. Then, when a seemingly plausible but internally contradictory phrase like “outsourcing of thought” emerges, what do we do? Rather than examining its appropriateness, we simply accept it and spread it.
Isn’t this the real problem?
Are we perhaps using a term coined to criticize a phenomenon, only to recreate the very problem we were trying to criticize?
The Questions We Need to Ask
So what should we be asking ourselves?
Instead of asking, “AI is bad,” ask, “How am I using AI?” Instead of asking, “Outsourcing of thinking is the problem,” ask, “Am I really thinking?” Instead of asking, “Buzzwords distort reality,” ask, “Do I truly understand and use these words?”
If we truly believe this is the problem, shouldn’t we first examine the language we use? Instead of simplifying complex realities with fancy buzzwords, perhaps we need to step back and ask questions.
That might be the only way to ensure we haven’t actually given up on “thinking.”
