The hilarious thing is that I work with some who are still in their anti-tech reaction. I therefore get quite a lot of sparring practice from humans. That said, I totally used Claude (or one of them) to act in character to push back against my presentation of the use of AI in schools to some higher ups who had misgivings; I asked it to give me strategies and scripts, all of which I then condensed to three phrases which I carried in my head into the said presentation.
It worked.
I stayed in control and had pivots as needed.
Result = everyone on the same page, food for thought fed in a calm way, no explosions and no defensiveness.
That is a great example. I find that, for my own use, it makes sense to use it as a sparring partner when I am most unclear on the subject, my goals and objectives, or the intended output.
We could argue that students - having less domain expertise and experience producing the objectives we seek — benefit the most from the sparring partner approach.
However, sparring is tiring! So, for adults I advocate that we use it in specific moments where it is most necessary. And for students, the same - we can’t ask them to spar every time - it just wouldn’t be fair. But we can generate opportunities for the occasional “sparring round” with GenAI as a means of pushing their thinking.
Right? It did go all go a bit meta. Asking AI to spar with me to help me to spar with humans about using AI with younger humans🤣🤯.
Great point about the exhaustion. Was talking today with a colleague about the need, at a certain point, for the human in the loop to step in and just help if the pupil / AI chat gets too much for the kid.(This being hypothetical though: we haven’t gone all out with AI tutors in class, yet. All tbc next academic year…)
We don’t grow from nods. We grow from sharp, unexpected hits that force us to anchor our logic and refine our stance.
Your framing reminds us that inviting challenge is an act of trust, not threat. Too many treat AI and even human collaborators as mirrors to reinforce what they already believe.
It’s a clear reminder that using AI, or working with any thinking partner, takes courage. Courage to face blind spots, rewire bias, and embrace stronger outcomes.
Real progress lives in that tension.
Thank you for sparring in public. That’s where the real proof of thinking happens
“Inviting challenge is an act of trust, not threat.” Beautiful!
It takes a special person to do so - I can’t say I do it all the time (or even enough). But giving young people the option, opportunity, or choice to do so (with and without AI) feels like an imperative for the future.
Thank you for your thoughtful reflection. Wonderfully said.
Hitting the nail on the head once more Mike! That's what I meant in my last piece on that disastrous MIT study: "Imagine what we could have learned if that impressive EEG setup had been applied to a well-designed task"
It is kinda cool I am using AI this way without realizing it was a better way. Instinctively, I knew that AI as a creator was not only terrible at the job, but terrible for me. So, I have used it test me and spar with me. For example, I created a prompt which turned the LLM into a Socratic guide (or rather, sparring partner) and it made me think about an idea in a different way.
Recently, I had to learn about a new service at work, and I asked it to test my knowledge by asking me questions. As a result, I actually remember prime elements of the service much better.
AI is not a “partner.” It does not understand anything, including what it means to be a partner. It simply makes predictions. Stop anthropomorphizing AI.
No, I don’t view it as a partner - which is precisely the point. Toaster ovens don’t produce language, fetch data, or pattern out ideas onto a page via words. GenAI does - which is the way the relationship is completely different from traditional technology. So I won’t be interacting with GenAI the same way I interact with my oven - or any other traditional technology appliance. Which is why new frameworks are needed that place GenAI interaction “between” traditional technology and human partnership. Unfortunately, we don’t have the language for that relationship yet, which means we have to borrow from other fields until we do. Hence this post: “All The Word We Cannot Use.”
The hilarious thing is that I work with some who are still in their anti-tech reaction. I therefore get quite a lot of sparring practice from humans. That said, I totally used Claude (or one of them) to act in character to push back against my presentation of the use of AI in schools to some higher ups who had misgivings; I asked it to give me strategies and scripts, all of which I then condensed to three phrases which I carried in my head into the said presentation.
It worked.
I stayed in control and had pivots as needed.
Result = everyone on the same page, food for thought fed in a calm way, no explosions and no defensiveness.
Deep irony!
That is a great example. I find that, for my own use, it makes sense to use it as a sparring partner when I am most unclear on the subject, my goals and objectives, or the intended output.
We could argue that students - having less domain expertise and experience producing the objectives we seek — benefit the most from the sparring partner approach.
However, sparring is tiring! So, for adults I advocate that we use it in specific moments where it is most necessary. And for students, the same - we can’t ask them to spar every time - it just wouldn’t be fair. But we can generate opportunities for the occasional “sparring round” with GenAI as a means of pushing their thinking.
Thank you for sharing this!
Right? It did go all go a bit meta. Asking AI to spar with me to help me to spar with humans about using AI with younger humans🤣🤯.
Great point about the exhaustion. Was talking today with a colleague about the need, at a certain point, for the human in the loop to step in and just help if the pupil / AI chat gets too much for the kid.(This being hypothetical though: we haven’t gone all out with AI tutors in class, yet. All tbc next academic year…)
Mike, you land this perfectly.
We don’t grow from nods. We grow from sharp, unexpected hits that force us to anchor our logic and refine our stance.
Your framing reminds us that inviting challenge is an act of trust, not threat. Too many treat AI and even human collaborators as mirrors to reinforce what they already believe.
It’s a clear reminder that using AI, or working with any thinking partner, takes courage. Courage to face blind spots, rewire bias, and embrace stronger outcomes.
Real progress lives in that tension.
Thank you for sparring in public. That’s where the real proof of thinking happens
“Inviting challenge is an act of trust, not threat.” Beautiful!
It takes a special person to do so - I can’t say I do it all the time (or even enough). But giving young people the option, opportunity, or choice to do so (with and without AI) feels like an imperative for the future.
Thank you for your thoughtful reflection. Wonderfully said.
Hitting the nail on the head once more Mike! That's what I meant in my last piece on that disastrous MIT study: "Imagine what we could have learned if that impressive EEG setup had been applied to a well-designed task"
https://wesstrabelsi.substack.com/p/when-mit-is-also-beating-dead-ai
Absolutely! I would love to know what happens in someone’s brain when they “spar” with AI. Now - we just need to buy an EEG machine haha!
Mike,
My email is kevin.hallinan@udayton.edu.
I free (all times EST)
Monday - all day
TuTh - after 1
Wed - after 9:30
Fri - all times but 10:30-11
Well said.
I've asked my students to leverage AI at the end of a project or assignment to spar with the AI about the appropriateness of their solution.
Do you mind if I incorporate your 'sparring' into activities I am creating for my platform "Girls AI-ing and Vibing" as 'AI Learning' activities?
Here's the platform: https://girls-aiing-and-vibing.vercel.app/
Of course! Let’s talk offline. I’d love to hear more.
It is kinda cool I am using AI this way without realizing it was a better way. Instinctively, I knew that AI as a creator was not only terrible at the job, but terrible for me. So, I have used it test me and spar with me. For example, I created a prompt which turned the LLM into a Socratic guide (or rather, sparring partner) and it made me think about an idea in a different way.
Recently, I had to learn about a new service at work, and I asked it to test my knowledge by asking me questions. As a result, I actually remember prime elements of the service much better.
AI is not a “partner.” It does not understand anything, including what it means to be a partner. It simply makes predictions. Stop anthropomorphizing AI.
I suppose I should treat it like I treat my toaster oven?
Actually, yes, unless you regard your toaster oven as your partner in preparing your morning English muffin.
No, I don’t view it as a partner - which is precisely the point. Toaster ovens don’t produce language, fetch data, or pattern out ideas onto a page via words. GenAI does - which is the way the relationship is completely different from traditional technology. So I won’t be interacting with GenAI the same way I interact with my oven - or any other traditional technology appliance. Which is why new frameworks are needed that place GenAI interaction “between” traditional technology and human partnership. Unfortunately, we don’t have the language for that relationship yet, which means we have to borrow from other fields until we do. Hence this post: “All The Word We Cannot Use.”
https://mikekentz.substack.com/p/all-the-words-we-cannot-use-why-talking