AI : LLMs and the Death of Critical Thinking

Are We Letting AI Do Too Much?

Every day, more of us are relying on generative AI in our jobs, and even personal lives. Whether it’s for writing emails, drafting reports, or even generating entire business strategies. Tools like ChatGPT, Copilot, and MidJourney have made it easier than ever to produce high-quality content at speed. But while AI can get the ball rolling, I think it’s fair to say that too many people stop at step one, mistaking AI-generated output for a final product.

The problem isn’t that AI is designed to replace our thinking—it’s that we’re letting it. AI is an accelerator, a launchpad for ideas. But if we treat it as the destination rather than the starting point, we stop flexing our creative intelligence, we leave our experience at the door and the result is page-filling but lacklustre.

Why do so many people stop at step one? The answer is simple: it’s easy.?

AI provides structured, coherent, and seemingly polished output in seconds. It removes the blank page struggle, the deep thinking required to craft - and structure - arguments, and the messy iteration process that fuels true insight. But this shortcut comes at a cost: the loss of differentiation, of originality, and of the deeply human perspective that makes ideas truly compelling.

What Happens When We Don’t Think for Ourselves?

1. Less Challenge, Less Depth AI aggregates from vast amounts of training data, averaging out insights rather than pushing the boundaries of thought. This means that if we simply accept AI’s first response, we risk producing content that is safe, predictable, and ultimately uninspiring. I’ve had interesting experiences when I’ve used an LLM to give me guidance on how to approach something, and then needed to replicate it a few days later and got a slightly different answer. But which one was right? Possibly both, but which was better? If we take the first answer, what are we missing? Often the problem is simply that we don’t know what we don’t know.
2. The Junior Dilemma Someone recently pointed out to me that AI is an extremely effective way of covering the grunt work used to train young professionals. In the past, entry-level workers honed their skills by writing reports, analyzing data, and summarizing research. Learning from their senior colleagues about better approaches by putting in the hard basic work. Now, AI can do all of this instantly—so where does that leave juniors? If the foundation of learning disappears, how do they build expertise? Where will we be in 10-20 years time when our visionaries who came up through the ranks have retired and moved on?

There’s an interesting point of view here that I enjoyed: https://www.duperrin.com/english/2024/11/22/ai-automation-jobs-juniors-skills/

1. Loss of Human Differentiation The other big issue I see is that businesses thrive on unique perspectives and strategic vision. Leaders are hired for their experience, insights, and ability to see patterns others miss. But AI doesn’t think—it regurgitates and repackages. Relying on AI to shape strategy risks creating generic, undifferentiated businesses that lack a true competitive edge. Yes, you can craft that marketing campaign strategy with ChatGPT but guess what? So can your competition. And they do. If you have any doubts about how effective the human factor is, go and watch your favourite social media channel and pay attention to how much is AI generated vs human. Or sites like Quora or Reddit - I bet you’re like me and you find yourself scrolling past the AI content because it's just a bit dull. Guess what - what's obvious to everyone in copywriting, is also true in code generation. And even more so in business strategy but it might not be quite as obvious until it's too late.

There’s a great article here on the advent of AI-generated slop: https://reutersinstitute.politics.ox.ac.uk/news/ai-generated-slop-quietly-conquering-internet-it-threat-journalism-or-problem-will-fix-itself

1. The Decline of Problem-Solving Skills AI offers easy answers, but the real value of human intelligence lies in asking better questions. If we let AI think for us, we risk losing the ability to break down complex problems, weigh different perspectives, and craft novel solutions.

I recently spoke with Decoding Data Science at AI Everything in Dubai on the gaps between where we are now and the bold proclamation of AGI (and there’s a link to more on this here: https://www.dhirubhai.net/pulse/artificial-general-intelligence-dream-distance-andrew-dunbar-svbhf/ ) One of the biggest gaps is around reasoning - the understanding of the solutions to problems. Largely what our day-to-day AI solutions like LLMs use at the moment is based on statistical prediction. So the AI doesn’t really understand the solution it is presenting, just that it's most likely the right answer. We are trying to solve this with neurosymbolic and causal inference models but we aren’t there yet. This is one of the biggest stumbling blocks - if we can build AI that truly understands what it is solving and the solution, then we can look at applying those insights across functional areas and achieve the broad General Intelligence we strive for.

How to Retain the Human Edge?

AI isn’t the problem— sadly the problem is us. Humans are fundamentally lazy - its a blessing and a curse. There’s a famous quote by Bill Gates “I will always choose a lazy person to do a difficult job because a lazy person will find an easy way to do it” and AI is the epitome of this. The key is to treat AI as a tool, not a replacement for our thinking. Here are a few ideas on how to make sure AI sparks creativity rather than stifles it, along with prompts and ways of thinking to support it:

1. Go Beyond the First Draft: The first answer AI provides is never the best. Use it as raw material, not a finished product. Rip the output apart - rewrite it, challenge it. Where do you disagree? Argue with your LLM. Find relevant other thoughts on this. Prompt: I’ve rewritten the content as follows [insert content]Review and tell me what's missing, what's inaccurate and how it could be improved. Rate each area along with suggestions for improvements.
2. Use AI to Challenge Your Thinking: Instead of asking AI to generate a generic response, prompt it to explore different angles. Prompt: What assumptions is this based on and what is the risk that they are incorrect? What’s the counterargument?
3. Layer Human Experience on Top: AI lacks lived subjective experience, intuition, and strategic judgment. That’s where you come in. Ask yourself: How does this align with my past experience? Where does my intuition disagree? Suggest where you disagree and see how it responds - you probably aren’t the first, and you don’t need to agree with everything. I regularly remove output I disagree with as long as I feel strong in my convictions.
4. Understand the value of the Junior: The really exciting part about this new Age of AI is that we are moving into an Age of Play. The barriers of entry to almost all fields are now lowered across the board. Whereas before the craft itself in areas such as copywriting, code development, and even more complex areas such as clinical assessment could be a significant barrier to overcome, a junior can now present a very credible contribution in many fields.
5. Humans as refiners, reinforcers : the biggest successes we are seeing at the moment are in the fields of human reinforced learning as we train our data models, in the refinement of AI generated content via a human editor. Humans are and always will be our secret sauce when talking to other humans - as good as personalisation gets, we need that delightful surprise, the challenging serendipity that comes from engaging with other humans.

AI as a Catalyst for Bold Thinking

We are now seeing the shift in focus from execution to ideation, with that comes faster iteration, broader exploration, and certainly much easier access to information. It’s a tool that accelerates creativity—but only if we use it to push our own thinking more than we might have before, not replace it entirely..

This week, I’ve been rapidly prototyping AI-enabled ideas, chasing their real business applications—work that once took weeks now takes hours. From AI driven transcription analysis to real time data storage of insights - I am fired up everyday with what is possible. There is no doubt that AI speeds and enables that execution, but its human insight that directs the journey. That’s why junior talent remains vital. Interns and graduates bring fresh energy, challenge assumptions, and keep seasoned teams innovative. If we stop fostering talent, we stagnate.

The leaders who thrive in this new landscape won’t be those who blindly trust AI, but those who interrogate, refine, and build upon it. The best ideas will always come from human minds. AI can open doors—but we must walk through them. So roll up your sleeves, engage with the technology, and bring the next generation with you. That’s how success is built.