A prompt hack for learning new things quickly

Give LLMs a few compressed mental models as examples, then ask them to teach you about whatever you want to learn, starting with a few compressed mental models and expanding from there.

Once you understand the basics, ask LLMs to give you multiple examples from different perspectives so your brain can connect the dots and form a map.

A few examples of compressed mental models:

To gain weight, calories in > calories out. To lose weight, calories in < calories out.
The short-term stock market is a loser's game and a zero-sum game. For most people, the way to win is to take a long-term approach and not do much.
Software infrastructure scaling is largely about letting stateless services handle entropy and variable load while keeping stateful services behind a consistent, predictable load.

Use these examples in the prompt (if you don't have any of your own), and watch how quickly you can learn new stuff!

Elaboration

If you have a deep understanding of one domain, you can use it to quickly learn about another domain, thanks to LLMs.

Why/How?

LLMs are pattern/token predictors. The more detailed/in-depth your prompt, the more detailed/in-depth the result. An added advantage is that, because LLMs have knowledge of so many domains, your expertise in a single domain can quickly translate across the domains they know about.

Every domain has levels of depth. When you can discuss a topic deeply and ask LLMs to teach you about a new domain at that level, they will match your depth in the new topic.

The more deeply a person understands a domain, the more simply they can explain it. Deep understanding often leads to simplification (the bell-curve meme is applicable here).

This is why compressed mental models help beginners grasp things quickly. Only after gaining substantial experience do people start to realize the power and depth of these simplified/compressed mental models.

Elaborating on the examples above

Let's unpack the compressed mental model of weight gain and loss:

The mental model assumes that you are taking care of the basics, such as:

Getting enough sleep
Eating moderately-healthy food
(etc)

This isn't black and white i.e it doesn't apply to every person in exactly the same way. For example:

You have a health condition, such as diabetes, a thyroid disorder, or high cortisol levels.
In some cases, your genetics may make it difficult for you to gain or lose weight.
(etc)

To summarize, human biology is such that this model of calories in V/S calories out applies to the vast majority of people.

Concluding

Once you prompt LLMs with a few compressed mental models, they will respond similarly in whatever domain you ask it to teach you.
Then, as LLMs expand on those topics and you get the hang of the basics, you can ask them to give you multiple examples, each from a different perspective. While going through those examples, you start to notice a pattern / framework and understand the domain much better.

Now that you have the knowledge, it's time to apply it (i.e., turn it into wisdom). Soon, you'll get good in this new domain!