Well written. I don't know if I fully agree that it makes you dumber, but I do agree it makes you lazier. And I certainly agree that it changes you. Having access to easy answers makes some people believe there are easy solutions, or fail to keep asking those 2nd, 3rd, and 4th dimensional questions about the problem. AI can make people feel overconfident and not question what they are being fed as an answer. I love your thought process and love that you are challenging those who read your work. Keep that going.
I agree with you wholeheartedly. Your “confidence without interrogation” comment is spot on. There is an arrogance to blind knowledge without context. I think the better way to use AI is to keep asking questions and not be afraid to get pushback on answers you may not agree with. AI is a research tool, but it is only as good as the effort you put in (and the prompts you craft upfront). In the old days (yes, I am old) we used to have to spend hours, weeks or months to research a topic that now can be presented to you in seconds in AI. And most people assume that what they are getting is 100% factual and true, not a skimming of information that is floating out there in the internet - some true, some not, but all combined into a response to a query. You are pointing out the correct problem. Instead of taking AI at face value, it should make you question more. What are the sources, how valid are they, what biases do they add to the equation, etc. I use AI as a research tool, and as an editorial tool in my writing. To help me keep alignment with the structures (magical, character, geographical, philosophical, etc) that I put in place during my worldbuilding. I learned early on that you have to constantly validate the responses AI gives you because hallucinations are real and costly. AI will get better, but you are right in the premise that in the meantime, AI is creating an overconfidence in outcomes it produces without the real challenges to how it got there. Thank you for the dialogue. This is why I came here to substack.
Well written. I don't know if I fully agree that it makes you dumber, but I do agree it makes you lazier. And I certainly agree that it changes you. Having access to easy answers makes some people believe there are easy solutions, or fail to keep asking those 2nd, 3rd, and 4th dimensional questions about the problem. AI can make people feel overconfident and not question what they are being fed as an answer. I love your thought process and love that you are challenging those who read your work. Keep that going.
Michael thank you! I appreciate this.
I think you’re right about “lazier.”
When I say “dumber,” I don’t mean losing intelligence, more a quiet drift away from effortful thinking.
Easy answers change expectations.
They can make us stop staying with those deeper, second and third order questions you mentioned.
What concerns me most isn’t wrong answers, but confidence without interrogation.
Thanks for engaging the piece thoughtfully — that kind of dialogue is exactly why I’m writing this series.
I agree with you wholeheartedly. Your “confidence without interrogation” comment is spot on. There is an arrogance to blind knowledge without context. I think the better way to use AI is to keep asking questions and not be afraid to get pushback on answers you may not agree with. AI is a research tool, but it is only as good as the effort you put in (and the prompts you craft upfront). In the old days (yes, I am old) we used to have to spend hours, weeks or months to research a topic that now can be presented to you in seconds in AI. And most people assume that what they are getting is 100% factual and true, not a skimming of information that is floating out there in the internet - some true, some not, but all combined into a response to a query. You are pointing out the correct problem. Instead of taking AI at face value, it should make you question more. What are the sources, how valid are they, what biases do they add to the equation, etc. I use AI as a research tool, and as an editorial tool in my writing. To help me keep alignment with the structures (magical, character, geographical, philosophical, etc) that I put in place during my worldbuilding. I learned early on that you have to constantly validate the responses AI gives you because hallucinations are real and costly. AI will get better, but you are right in the premise that in the meantime, AI is creating an overconfidence in outcomes it produces without the real challenges to how it got there. Thank you for the dialogue. This is why I came here to substack.
Michael this is exactly it. Thank you for taking the time to articulate this so clearly.
I really resonate with what you said about effort, validation, and pushback.
AI becomes dangerous not because it’s wrong, but because it feels complete before understanding has actually formed.
What you describe questioning sources, biases, structure is precisely the posture I’m trying to point toward.
Not rejecting AI, but refusing to let it replace inquiry.
And yes — hallucinations aside, the bigger risk is overconfidence without a sense of how an answer came to be.
I appreciate you bringing experience and nuance into the conversation.
This kind of dialogue is exactly why I’m here as well.
Thanks for spreading awareness about AI
Just highlighting this part
When answers arrive instantly,
questions stop maturing.