Interpretation of Ambient: Maintaining high-speed and efficient characteristics, introducing PoL Consensus fork of Solana.

Author: Fairy, ChainCatcher

In today’s world where blockchain and artificial intelligence technologies are continuously evolving, effectively combining the two has become one of the goals of many innovative projects. Ambient was born in this context, dedicated to integrating decentralized blockchain architecture with large-scale AI reasoning, exploring a new model of intelligent economy.

As a complete fork of Solana, Ambient retains Solana’s speed and efficiency, and by introducing the Logits proof (PoL) mechanism, it creates a brand new blockchain ecosystem.

What is ### Ambient?

Ambient is a Layer-1 blockchain that combines Solana SVM compatibility with a novel proof-of-work mechanism, providing ultra-scalable validation inference. The core idea of the Ambient project is to deeply integrate AI inference with blockchain, creating a decentralized AI economy.

Unlike traditional Proof of Stake (PoS) systems, Ambient employs an incentive mechanism similar to Bitcoin’s, providing predictable profits to each node participating in network inference, fine-tuning, or training. This approach avoids reliance on enterprise-grade GPUs and ensures sustainable profitability for miners through compensation based on transactions and inflation. Both miners and users can receive rewards that match their contributions, while the value of the platform continues to increase with the growth of the network.

Characteristics of Ambient:

• Efficient reasoning and security: Provides fully verified reasoning with overhead of less than 1%, while ensuring high security on large intelligent models (600B+ parameters) and their fine-tuned versions.
• Excellent training performance: Training performance improved by 10 times compared to existing methods, enhancing the training efficiency of AI models.
• High miner utilization: By optimizing on a single model, the utilization of miners has been improved, enhancing the efficiency of the inference and verification processes.
• Non-blocking Proof of Work Consensus: Adopts a non-blocking Proof of Work mechanism to ensure economic competition in core network activities (inference, fine-tuning, training), while maintaining high TPS and avoiding the performance bottlenecks of traditional blockchains.

Ambient Team Background and Development Status

In addition to the background of the founder, Ambient has not disclosed information about other members. Travis Good, the CEO and founder of Ambient, has a diverse academic background that spans four fields: government, economics, computer science, and machine learning. Travis’s leadership style emphasizes execution and practicality; he consistently adheres to hands-on approaches and focuses on actionable solutions while promoting technological innovation. Furthermore, Travis is very active on Twitter, often sharing his unique insights on technology, innovation, and industry trends.

On April 1st, Ambient completed a $7.2 million seed round funding, led by a16z CSX, Delphi Digital, and Amber Group. Big Brain Holdings, Superscrypt, Proof Group, Rubik Ventures, Aethir Foundation, and Edessa Capital also participated. Ambient plans to launch a testnet in the second/third quarter.

Logits proof of consensus mechanism

The “Logits Proof” algorithm leverages a key fact: logits (which can be understood as logical units) are both unique fingerprints and effectively capture the model’s “thinking” state at a specific moment (i.e., when the model is producing “streaming” outputs) through the hash values generated during the model’s generation process. Under this mechanism, the hash value of the logits proof is a hash list of the hash values of each group of logits before each output token. In short, for each token n, up to the final token t, the hash value of the logits proof is:

Hash(Hash(n) … Hash(t))

The hash value of the logits progress mark proof is the logits hash after generating x tokens, where x is between n and t (inclusive of n and t), that is:

Hash(n) … Hash(x) … Hash(t)

Based on this principle, a verification mechanism can be constructed: first, the miner generates text; then, the verifier randomly selects a word from the text and requests the miner to provide the “thinking state” at that point (i.e., the corresponding logits progress marking proof hash). The verifier then performs an inference on that word in the same model and context to generate its own “thinking state.” If the two “thinking states” (represented by hash values) are consistent, the verification is successful.

This proof-of-work mechanism is consistent with the design principles of Bitcoin: mining (in this case, repeatedly executing the model through the inference of 4000 tokens) is costly, but the verification process is very cheap (requiring only the inference of 1 token). This mechanism not only improves efficiency but also ensures the security and reliability of the verification.