https://www.gate.io/learn/articles/do-all-roads-lead-to-mpc-exploring-the-end-game-for-privacy-infrastructure/3934

The article argues that if the ultimate goal is to create programmable privacy infrastructure for blockchains that can handle shared private states without a single point of failure, then Multi-Party Computation (MPC) may be the most viable solution.

Key Points:

1. Current Privacy Infrastructure Limitations: Existing privacy solutions in blockchains, such as private payments or voting, are limited in scope. To expand the design space beyond speculation and trading, more advanced privacy technologies are needed.
2. Three Hypotheses:
   • Abstraction of Cryptography: Developers shouldn’t have to deal with the complexities of cryptography. Privacy solutions should be abstracted away, enabling easier development of private applications.
   • Shared Private State: Many applications require a shared private state, and to manage this securely without centralized trust, MPC is necessary.
   • Larger Shielded Sets: Enhancing privacy often depends on minimizing information leakage when entering or exiting shielded sets, which could be improved by increasing the size and number of private applications on the same blockchain.
3. End-Game for Privacy:
   • MPC vs. ZKPs and FHE: While Zero-Knowledge Proofs (ZKPs) and Fully Homomorphic Encryption (FHE) offer strong privacy guarantees, they struggle with shared private states. MPC, however, provides a way to securely manage shared private states, making it a critical component for future privacy infrastructure.
4. Challenges with MPC:
   • Trust Assumptions: The main risk of MPC is collusion among parties. The strength of privacy guarantees depends on the specific MPC protocol and how it manages this risk.
   • Maturity of Technology: MPC suffers from significant communication overhead, particularly in complex computations and large networks. Most existing protocols are restricted to small, permissioned operator sets due to these limitations.
5. Alternative Approaches:
   • TEEs (Trusted Execution Environments): Offer hardware-based privacy solutions but come with their own trust issues and vulnerabilities.
   • Intermediated Privacy: Relies on trusted third parties, which may be suitable for lower-value or high-performance applications.
   • Stealth Addresses: Provide privacy for transactions but lack the flexibility needed for general computation.
6. Risks to the Thesis:
   • MPC may not be necessary if shared private state proves less important, or if the trade-offs in performance outweigh the benefits of privacy guarantees.
   • Regulatory Hurdles: Privacy solutions face significant regulatory challenges that could hinder their adoption.
7. Conclusion: While MPC is not the perfect solution, it offers a significant improvement over current centralized privacy solutions. However, the complexity and overhead involved mean that alternative approaches should also be considered depending on the specific needs and trade-offs of the application.