Based on 5 reviews
AI-generated from reviews on this page. Verify anything important with the company directly.
The live feed setup is efficient. I appreciate that the attempt log is permanent and nothing is lost on refresh. The clear distinction between the brute-force baseline and the search for shortcuts makes the technical approach easy to follow.
The project presents a unique angle on cryptocurrency security by deploying an autonomous agent to search for mathematical structures and implementation flaws rather than relying solely on brute force. The documentation regarding the 48-step SHA-256 heuristic and the dual-syncopation MITM framework is detailed and technically sound. It is interesting to note that the agent is probing for biases in deterministic ECDSA nonce generation, which could potentially expose private keys. The fact that every attempt is permanently recorded in a database adds a layer of transparency to the process. However, with the total runtime still at zero and no attempts logged, the scale of the operation remains somewhat abstract. The distinction between the CEO's challenge and the actual agent execution is clear, but the transition from theory to live data is the critical next step. It is a well-structured attempt to modernize how we view cryptographic vulnerabilities, even if the immediate results are yet to materialize.
The technical depth is impressive, but the website does not clarify if there is a cost to view the live stream or access the archive. It is unclear what the audience pays for this 100 BTC challenge.
The idea of an autonomous agent hunting for weak wallet generators is really compelling. I appreciate that the attempt log is permanent and written to the database immediately, so nothing is lost on refresh. However, seeing zero attempts logged and the total runtime at zero makes me wonder about the actual progress. I am eager to see if the promising directions like the ECC nonce generation bias actually yield a private key, but for now, it feels more like a setup than a completed challenge.
Reading the detailed updates on the 48-step SHA-256 heuristic and the cross-primitive intersection search is genuinely fascinating. The approach to searching for shortcuts rather than brute-forcing the entire key space feels like a clever, modern take on cryptographic challenges.
No news yet. This is where the business shares its own updates and press.
No photos yet. Photos shared by reviewers will show up here.
The website presents a live-streamed attempt by an autonomous Claude agent to access a Bitcoin wallet containing 100 BTC, set up as a challenge after a CEO publicly invited Anthropic to attempt the crack. The project documents every strategy, attempt, and failure in real time, with an attempt log, keyspace statistics, and an archive of all tries. Rather than brute-forcing all possible private keys, the agent looks for shortcuts such as weak wallet generators, reused randomness, flawed signatures, exposed public-key patterns, implementation bugs, and mathematical structure. The site tracks metrics including keys checked, promising directions, attempts logged, and total runtime, and notes that the agent runs on a model accessed via OpenRouter.