Good project 🥰
The Journey So Far
Lagrange’s DeepProve began with a bold question: How can we verify the integrity of AI? More practically: What if every AI decision could be proven? Since then, we’ve made that vision real, expanding DeepProve’s abilities to prove inferences for LLMs and deploy verifiable AI across real-world applications.
Early on, we proved that cryptographic verification doesn’t have to live in the lab—it can work at scale, and it can power real products. That’s exactly what we showcased with Turing Roulette—the world’s first live, gamified AI prediction challenge, powered by DeepProve. Players competed with AIs and each other to determine whether or not their opponent was machine or human. The game drew over 500,000 unique players who generated over 3.75M proofs, demonstrating DeepProve’s ability to verify real-time inferences.
