Mental Models in the Age of AI
- 3 days ago
- 7 min read
Thinking with AI
So, the anthem stands by now:
“AI won’t replace you…but someone using AI will.”
Including someone that thinks with AI.
One counter-question would be:
"Why do I need to think when I can just ChatGPT it?"
And...that’s kinda the problem:
You don’t just naturally delegate your thinking to AI.
Mythbuster: ChatGPT is no Sage
See, the thing is...
ChatGPT is not—and will not—be the answer to all our problems.
It’s no guru...but a tool, like any other Generative AI.
Meaning, “Just ChatGPT it” is as vague as the thought process behind it.
And ChatGPT—or any AI for that matter—is only as good as the brains behind it. And not just data brains...but human brains, too.
Enter Mental Models
Mental Models—A subpremise of this article.
Combined with AI, Mental Models can make a user formidable in the next decade or so.
What are Mental Models?
“Thinking Shortcuts” in short.
If you’ve heard of concepts like:
↳ First-Principles Thinking (My Favorite)
And so on,
Those are some Mental Models.
Thinking frameworks that help you problem-solve scenarios.
Now that definition’s out of the way, let’s add AI to the mix.
MENTAL MODELS FOR THE AGE OF AI
So, as initially stated, ChatGPT alone isn’t enough—
We need some thinking to back it up and some mental models up our sleeves, considering how fast the world is evolving in tech.
Some mental models for starters:
First Principles Thinking
Bayesian Updating
Second-Order Thinking
Critical Thinking
Systems Thinking
Inversion Thinking
Bias to Action
Root-Cause Analysis
Latticework of Models
Mental Model Plasticity
Cynefin Framework
Probabilistic Thinking
Critical Thinking
Note
An overarching theme in this article is that AI is an assistant.
That’s already obvious off the bat, but the more we dive in, the more we’ll uncover that AI further brainstorms beyond our natural limits.
And at the end of the day, we bear the consequences of our actions, not ChatGPT.
1. First Principles Thinking
First-Principles Thinking is breaking down problems into their fundamental, undeniable components and solving them from scratch.
The art of problem deconstruction and building up solutions from notwithstanding pillars of facts.
Why is this important? It helps to have these irrefutable pillars of truth to lean on when all else fails, especially during ambiguity.
So in the case of AI, sure—You can use it to solve a problem...or use it upon these pillars for amplified results.
Or if you’re not sure where to start, prompt it.
2. Bayesian Updating
Bayesian Updating is revising one’s beliefs using new data.
Why is this important? Data is changing everyday, and so is AI—Both evolving at rates faster than the human mind can fathom.
And our thinking prowess is only as good as the data we hold.
So with Bayesian Thinking (or Updating, as you may), the more unlikely something is, the more compelling the evidence must be to update your belief.
This factually proves one conclusion: The ignorant and the close-minded shall be left behind if they don’t adjust their thinking.
3. Second-Order Thinking
2-o-Thinking (as I call it) considers the effects of our actions to the second order (or more.)
Meaning, “If this happens, that happens. And if that happens, what happens?”
Why is this important? AI can perform scenario analysis on different events. Meaning, if you can’t hypothesize an outcome in second-order, AI can map out potential outcomes, further boosting your analysis.
One caveat, though, is that AI generates outcomes from historical data, so while data provides context, it reduces (but doesn’t entirely eliminate) second-order ambiguity.
Meaning...the only way out is through. You must take action regardless. (See #7 - Bias to Action - below.)
Swipe left for more.
4. Critical Thinking
As the title suggests, this implies the art of thinking critically using whatever information is at our disposal to solve a problem logically.
And not everything that glitters is gold—Don’t take everything at face value.
Case example
When writing this article, I asked ChatGPT for “Sample mental models in the age of AI,” and it also recommended “R&D as Search,” whatever that is.
Inexperienced-me would’ve considered it, but critical-me, already aware of mental models, had to look it up to confirm if indeed it was a thing.
It wasn’t.
Why is this important? AI is a tool and an assistant for that matter...but don’t take its output as gospel. Critically assess if:
The answer makes sense, and
The answer is helpful.
5. Systems Thinking
Systems Thinking involves analyzing an entire system, its constituent parts, and their cause-and-effect relationships.
Why is this important? AI is a second brain in cultivating your system. It can help analyze a system, create new ones, or identify blind spots in existing ones.
Swipe left for more.
6. Inversion Thinking
Inversion Thinking involves thinking the opposite of the intended goal, then doing its inverse.
In practice, it involves seeing what guarantees failure, then doing everything in your power to avoid it.
Charlie Munger was a huge proponent of Inversion Thinking, using it in much of his investment and life theses.
Why is this important? Sometimes the best thinking comes before AI.
Sometimes the best accelerator is not running towards something...but from something, and AI can’t do that for you.
Quick Plug - The Weekly Chronicle
Hopefully you’re gaining some value from this article.
If you are, kindly consider supporting me by joining my newsletter—The Weekly Chronicle— where I share Motivational and Personal Development Insights every Sunday in under one minute. Specifically for people wanting to improve their lives and better themselves every day.
More eyeballs on my work helps me be more candid, clear, and insightful. Plus, I want to learn more and share better, too.
It’s free.
Thank you.
We resume reading.
7. Bias to Action
Bias to Action is a natural tendency to act, favoring action over inaction. It’s more of a behavior than a model...but exemplary for this age of accelerated information access.
Why is this important? Bias to action is a trait independent of AI—A non-causal relationship. In other words, the presence of AI shouldn’t stop someone from taking action regardless.
You don't need the full picture to get started, but you need to start if you want to get the full picture. (Or at least achieve what you needed the picture for.)
One could brainstorm their way to planet Pluto, but action will always be the catalyst. People feel safe researching...but it’s the outliers that win by taking action.
AI is not needed here—Common sense and Execution are.
8. Latticework of Models & Mental Models Plasticity
2-in-1.
The Latticework of Models is of a network of interconnected models from various disciplines. (Disciplines including Technology, Engineering, and more.)
On the other hand, Mental Models Plasticity (MMP) is the ability to switch mental models interchangeably depending on the scenario.
Plasticity refers to the quality of something that can be easily shaped or molded.
Just like Neuroplasticity—The brain’s ability to change through the reorganization of its neural networks.
So when you combine MMP & the Latticework of Models, you get:
So, by definition:
Why is this important? No one model can solve an entire problem on its own.
Another thing: (Generative) AI can generate a latticework any time, depending on the problem to be solved. It can also drill down to the First Principles of each model and map interrelationships that are not readily identifiable to the human eye.
Additionally, both models provide a "playground" to test your mental models arsenal, just like you would try new tricks on a pitch. You get a good ground to test your models, hypotheses, and combinations across different scenarios, honing your decision-making muscle over time.
9. Cynefin Framework
Cynefin categorizes problems into five domains: Clear, Complicated, Complex, Chaotic, and Confusion.
Why is this important? Some problems are not so straightforward for AI to solve, plus the skill of identifying a problem’s complexity helps even before introducing AI to the mix.
Furthermore, AI isn’t needed to solve all problems. (Ref. Bias to Action above.) Cynefin classifies these problems—You judge whether or not AI is needed.
10. Probabilistic Thinking
Probabilistic Thinking estimates the likelihood of different outcomes.
Why is this important? PT is a mental muscle to be built that comes with experience and arithmetic, specifically Probability.
You don’t have to be a Math expert; Rather, you’re better off developing the ability of foresight.
AI can also map out possible scenarios...and to some extent, predict the chances of each happening (at least from the data it’s trained on.) And again, the human bears the brunt of the consequences.
11. Critical Thinking
Pretty straightforward.
Critical Thinking is analyzing the available information to form sound judgments and make informed, logical decisions.
Why is this important? Even AI itself admits so.
🌟 THE BOTTOM LINE
AI has infinite memory—We love that already.
But humans still possess the edge in judgment, wisdom, and discernment. (Or at least...so long as we realize that.)
Mental models can help us become great thinkers and, to some extent, prolific problem solvers.
Legends of the past have shaped cities, moved economies, and changed the course of history through their use of mental models.
And so can we.
And with the onset of AI, it will be no different.
Take nothing else from this essay but this: AI is but a tool—A means to an end.
And those who harness it and know when to use it will rule the world.
The best is yet to come.
Peace,
.
.
.
~T.K.K
Subscribe to The Weekly Chronicle Newsletter for motivational insights and high-impact ideas in under one minute. Delivered every Sunday.
Comments