Try. Tinker.
“Oh, now I get it.”
Four small experiments for the ideas that matter. Make a choice and see what changes. No timers, no grades, no real funds.
Buttons need JavaScript. Worked examples below remain readable without it. Nothing is sent to a server, stored or connected to a wallet.
The satoshi challenge
One BTC = 100,000,000 sats. Multiply the BTC quantity by 100 million. This is a unit conversion, not a market price.
Choose an answer to see the explanation.
See a worked example
0.001 BTC × 100,000,000 = 100,000 sats. Changing the unit does not change the amount you have.
Where does the change go?
Make a fictional 7,000-sat payment with a fixed 500-sat classroom fee. Select whole inputs below. Any remainder becomes a change output in this simplified model.
Choose enough fictional inputs to cover 7,500 sats.
See the balanced example
6,000 + 2,000 = 8,000 sats in. Then 7,000 to the recipient + 500 change + 500 fee = 8,000. Change is a new output, not an unused piece of an old input.
Illustration only. Real fees depend on transaction weight and fee rate; input choice affects privacy and fees. This model does not build transactions, apply dust rules or recommend coin selection.
The 2-of-3 signing room
Imagine a policy that requires two distinct valid signatures from three keys. Select fictional signers and watch the threshold. Then remove one and ask what recovery would require.
0 of 3 selected. Two distinct valid signatures are required.
See the three combinations
A + B, A + C, or B + C can meet the threshold, assuming valid signatures under the actual policy. A single key cannot. Two copies of the same key do not count as two different signers.
No keys are generated. A satisfied signing threshold does not establish permission from people, safe backups or independent devices. Recovery design needs more than this counting model.
Would you stop here?
Read the fictional situation. Decide whether it calls for stopping or continuing your checks. “Keep checking” never means “send money.”
Choose a response to explore why.
Read a sample situation
Someone calling themselves support asks for recovery words. Stop: those words can provide spending control. Never share them with a helper, a website or an AI.
Take the “aha” back to a lesson.
These tiny models leave out real-world complexity on purpose. Follow the linked lesson and primary sources to understand the limits. New lab references checked October 5, 2026.
Teach it to someone else.
Explain where change comes from, why key independence matters, or what made a request suspicious. A clear explanation is worth more than a lucky guess.
Visit the community open table →
AI-assisted classroom exercises, not financial advice, a wallet, certification or a guarantee of safety.