Okay, my bad. I thought he was talking about predicting the next puzzle, but he was actually referring to Puzzle 66 with some BS. However, I still have my doubts that retired_coder and satoshi_rising are the same person.

---

This thread spans 369 pages filled with debates, groundbreaking ideas, successes, and failures. It holds a wealth of knowledge—some theories dismissed, others leading to real breakthroughs. Many thanks to those who contributed to this field by developing BitCrack, VanitySearch, keyhunt CUDA, JLP Kangaroo, Sota Kangaroo, Rotor CUDA, wifcudasolver, and many more.

Now, with all this information, it’s time to build something new.

During my reading of more threads, I came across another person's idea that seems very close to the actual method used for creating these puzzles. Here is the post:

https://bitcointalk.org/index.php?topic=4453897.msg61848712#msg61848712

In Python, it's quite straightforward to implement multiple derivation methods for private keys that are masked with leading zeros. I'm curious—has anyone developed a CUDA version of these methods? If so, could you share a reference link?

Thanks in advance...