Assumption surfacing

Have the unstated premises in your own reasoning named, the things you're relying on without having noticed you decided them.

When to use this: When you want to go deeper
Learning impact: Reliable gain
Learning evidence: Credible with caveats
Where you are with the topic: Building understanding
How often to use it: Worth repeating often

When to use it

Your argument feels watertight until someone asks what you assumed.
A seminar or tutorial is coming and you need the hidden premises named.
Two smart people disagree and you suspect they are not arguing about the same thing.

What to ask AI

Type something like: “Here's my reasoning: [paste]. Don't tell me if I'm right. List the premises I'm relying on but haven't stated.”

Ask it to mark which premises are load-bearing versus ornamental.

Keep deciding which assumptions you actually accept.

How it helps

Most reasoning rests on premises you never wrote down.

You notice the conclusion. You miss the silent ifs.

Assumption surfacing names those silent premises so you can accept, reject, or test them.

Often the disagreement was never about the conclusion, it was about an unstated if.

Once named, a shaky assumption becomes something you can check instead of something you defend by habit.

How to run it

Paste a short piece of your reasoning, not a whole essay.
Ask for unstated premises only, no verdict on whether you are right yet.
For each premise, say: accept, reject, or need evidence.
Rewrite the argument with the load-bearing premises stated out loud.
Attack or test the weakest accepted premise next.

Example exchange

You

Here’s my reasoning: remote exams lower cheating because students fear getting caught on camera.

One silent premise in there: that cameras deter more than they open new cheating channels. Is that what you’re assuming?

You

Yes. I hadn’t named it. I don’t actually have evidence for that.

Then your conclusion is conditional: it holds only if deterrence beats new channels. What would count as evidence either way?

Copyable prompt

Here's my reasoning: [paste]. Don't tell me if I'm right. List the premises
I'm relying on but haven't stated. For each: is it load-bearing or optional?
Wait for me to accept, reject, or flag each one before you critique the conclusion.

The Tell

If every assumption surfaced is one you already would have named, the pass was too shallow.

If the AI jumps to “you're wrong” before listing premises, it skipped the method.

You are done when you can point to one silent if that actually carries the conclusion.

Principle evidence

Strength of the underlying learning idea, not a claim about AI products.

The underlying learning idea is rated moderate. Surfacing unstated premises is core argumentation pedagogy and critical reasoning. Making hidden ifs explicit improves evaluation of arguments.

AI delivery evidence

Whether an AI tutor delivers this method well is a separate question.

AI may invent premises you never relied on, or miss ones you did. You still judge the list. Delivery quality is speculative.