Over the last few months, people have been asking me what inspired me to start Mind Club.

Alex Kinsella did a wonderful job at capturing the origin story in this piece on the BitBakery Software blog.

Check it out!

Mind Gym: How Aggie Branczyk Built an Accidental Community in the Age of AI.

It's popular among physicists to think you can't understand quantum mechanics without the math. I actually agree with that.

But I also think you can teach a curious person enough of the math to get them pretty far toward that understanding. That's what I want to try out in this next series of Mind Club sessions.

Sign up here

What it is

Go Quantum is a new series of Mind Club sessions. Each session is one self-contained module, and each module focuses on a single quantum phenomenon: superposition, entanglement, measurement, and more.

The sessions are drop-in. The modules run in order, but they are not tied to specific dates, so if you miss one you simply pick up the next module the next time you come in.

You don't need a physics background

It's not for academics. It's for curious knowledge workers who don't mind putting their brains to use.

Don't let the idea of doing math scare you off. The sessions are designed to be accessible. It will feel more like solving puzzles than doing math.

By the end of each session, you will have:

• an intuitive feel for the phenomenon
• an idea of how real researchers model it
• a grasp of how the popular media gets it wrong

Who's running it?

Dr. Aggie Branczyk has a PhD in quantum physics and spent seven years at the Perimeter Institute for Theoretical Physics teaching in the PSI Masters program, an intensive one-year theoretical physics bootcamp. She was heavily involved in curriculum design and developing problem sets for topics at the frontier of the field.

How it differs from a regular session

If you've been to a regular Go Mental session, this will be a bit different. The format is the same, in person, pen and paper, calculators provided, no phones and no AI, but the problem set is much more guided and pedagogical than a usual Go Mental sheet.

The first module is superposition.

Sign up for first session

If you're in KW, I hope you can make it.

Can't make it in person?

The Go Quantum problem sets will eventually be available as PDFs in the shop, alongside the sets from past sessions. Join the waitlist and I'll let you know when they're ready.

When I started writing Mind Club problems, I had a strong gut feeling about what I was trying to achieve. I wanted problems that were less about testing knowledge/specific skills and more about practicing problem solving and critical thinking more broadly.

I expected these to be different to problems you encounter at school or university, broader than logic problems, and different to those found on IQ tests and interviews for consulting firms.

I decided to not overthink it and write the first two problem sets based on my intuition. I think it was a good start, but I'd like to refine my approach and inform it by more than just my gut.

So I started digging into problem design and discovered some very interesting work by an education researcher, David H. Jonassen. In his 2000 paper Toward a design theory of problem solving, he identifies 11 problem types.

Problem Types

There is a lot of nuance, but at a high level:

🔗 Logical problems are abstract puzzles where the method must be discovered; the challenge is finding the most efficient sequence of moves or manipulations

🪜 Algorithmic problems are where a known procedure or formula is applied to produce a correct answer; the challenge is selecting and executing it correctly

📖 Story problems are narratives that embed a procedure; the challenge is comprehending the context, identifying the relevant variables, and selecting the right algorithm to apply

📏 Rule-using problems are ones where multiple rules and solution paths are possible; the challenge is selecting and applying the right rules to produce a system-constrained answer

⚖️ Decision-making problems require selecting one option from a set of alternatives by identifying benefits and limitations, weighing options, and justifying the choice

🔧 Troubleshooting problems involve a closed system with one or more faults; the challenge is diagnosing the fault state by generating and testing hypotheses against known fault signatures

🩺 Diagnosis-solution problems involve a more open system with faults and numerous optional treatments; the challenge is selecting, evaluating, and justifying the best remedy

♟️ Strategic performance problems are real-time, complex situations where tactics must be applied to meet an ill-structured strategy while maintaining situational awareness as conditions shift

🏛️ Case analysis problems are ones where an existing situation must be interpreted with incomplete information and no consensual answer; the challenge is building a defensible argument from the evidence available

🏗️ Design problems are ones where something new must be created from a vague goal with few constraints; the challenge is first imposing structure on the problem, then producing a solution that can only be judged as better or worse

🌀 Dilemmas are situations where perspectives are irreconcilable and every solution involves compromise; there is no satisfying answer, only an articulated preference with justification

Application to Mind Club

It seemed like many of these problem types could be applicable to Mind Club, with the exception of Troubleshooting, Diagnosis, and Strategic Performance. This is because Troubleshooting, Diagnosis, and Strategic Performance are more about interacting with the physical world, rather than thinking with pen and paper. They also often require deeper domain expertise, which doesn't suit Mind Club, as I want to make it accessible to people from different career backgrounds.

Before trying to apply this taxonomy as a design guide, I wanted to see how existing Mind Club sessions scored against those 11 types. So I used Claude Code to evaluate each problem according to the criteria in the paper and present the scores on a heat map.

Session 1

It was interesting to see that in the first session, the problems leaned heavily towards Algorithmic and Story.

Perhaps not surprising given my background designing problems in the area of theoretical physics.

One piece of useful feedback I got in the first session was that some of the problems felt like helping their kids with HW!

Session 2

Based on this feedback, I adapted the problem types in the second session. I was happy to see that reflected in the heat map, where there is much less concentration in Algorithmic and Story columns.

I was surprised to see such a large representation of logic type problems. This was not deliberate.

Overall trends

As expected, there was not much representation from Troubleshooting, Diagnosis, and Strategic Performance in either session.

But it was interesting to see that in both sessions, Dilemmas were not well-represented either. This seems like a very interesting problem type to explore.

Session 3

I’m now using this to guide the problems I’m writing for the third session. It’s helping me decide which problems to keep and how to arrange them. These are the ones I have so far.

I think this will lead to a better mix of problems that target the desired areas and help me adapt better to feedback from folks who attend the sessions.

If you're interested in attending a session, upcoming events are on Luma. Full problem sets from past sessions are available in the shop.

It referenced Dr Evan Risko, a psychology prof at the University of Waterloo.

The worry is that, as Dr Risko puts it, generative AI allows one to "offload a much more complex set of processes". Offloading some mental arithmetic, which has only a narrow set of applications, is not the same as offloading a thought process like writing or problem-solving. And once the brain has developed a taste for offloading, it can be a hard habit to kick. The tendency to seek the least effortful way to solve a problem, known as "cognitive miserliness", could create what Dr Gerlich [Risko's colleague] describes as a feedback loop. As AI-reliant individuals find it harder to think critically, their brains may become more miserly, which will lead to further offloading. One participant in Dr Gerlich's study, a heavy user of generative AI, lamented "I rely so much on AI that I don't think I'd know how to solve certain problems without it."

It’s given me some food for thought.

The article is about the undesirable effects of outsourcing our memory. This was written about 10 months ago. Now we can outsource our reasoning as well as our memory.

So I think it’s even more important for us to do this mindfully, to figure out how to use technology to achieve our goals, but without losing our own abilities in the process.