A challenge on this platform has two components, and neither requires trusting the creator. The first is a Bitcoin address. The second is a prize that is deducted from the balance of the person who published the challenge and remains locked there until someone solves it.
Notice what this second component resolves. A puzzle published on a forum is worth as much as the word of the person who published it: who guarantees that there is a prize and that it will be paid? Here, the value ceases to be a promise the moment the challenge is created. It comes out of the creator's pocket, is locked by the system, and even that person cannot touch it while the challenge is active.
The verification of the answer is the part that concludes the course because it is exactly the process from Lesson 3 of Module 4, executed by a program instead of by hand. Whoever thinks they have solved it sends a private key. The platform multiplies this key by the secp256k1 generator, obtains the public key, applies SHA-256 and then RIPEMD-160, constructs the address, and compares it with the challenge address. If the two are equal, the prize is paid immediately. If not, a minimal fee is charged for the attempt, and nothing else happens.
It's worth emphasizing a point that many people take time to believe: the platform does not store the answer anywhere. There is no field with the correct key against which to compare, and no one here knows it. What exists is the process, repeated with each attempt. It's the difference between a system that knows the answer and one that only knows how to recognize it — and it's the same difference that Lesson 1 of Module 4 explained about hash functions.
The fee for a wrong attempt exists for an arithmetic reason. Without any cost, the optimal way to play would be to fire off millions of automatic guesses against the form — which would turn the server into someone else's search hardware. With a small cost per attempt, blind guessing is no longer worthwhile, and the search returns to where it belongs: the machine of the person searching.
Hints function like a market, and the design is deliberate. Any hint can be purchased, and each purchase increases the price of the next by ten percent — the more people consider the hint valuable, the more expensive the shortcut becomes. Of what is paid, a portion goes to the person who wrote the hint, and another portion increases the prize of the challenge itself. Those who help solve it increase the reward for solving it.
There are also alliances, which are this platform's response to the problem in Lesson 6 of the previous module: people searching together, with their own channel to coordinate who is covering what, instead of a thousand people digging the same hole.
And there is the published solution. After a challenge is solved, two people — and only they — can write about how it was done: the creator and the solver. Before that, no one can, because a published explanation with the challenge still open would give away the answer to those still searching. Afterwards, it becomes what this entire course is: study material with the method exposed.
You started by asking what money is and why what we use fails. You went through Satoshi's proposal, the ledger, mining, and the 21 million. You learned to use it safely, to store a seed, and to inherit it. You understood the hash, the curve, the signature, and the address byte by byte. You saw the network from the inside, from UTXO to difficulty adjustment, and what exists above it. And you finished with the calculations on the table: finite bodies, ECDSA, kangaroos, and the real cost of searching for a key.
All that remains is the part that cannot be learned by reading. The challenges are at /pt/puzzles, the prize for each is real and locked in waiting, and the verification is the same process you now know how to perform by hand.