Hoskinson Explains Hash vs Lattice-Based Cryptography

Charles Hoskinson Highlights Key Differences Between Hash Functions and Lattice-Based Cryptography

Cardano founder and IOHK CEO Charles Hoskinson drew attention this week to a technical but increasingly relevant topic in cryptography: the distinction between cryptographic hash functions and lattice-based cryptography.

Hoskinson’s remarks, shared under the headline phrase “Most Crypto People Have No Clue What’s About To Happen”, pointed to a broader gap between how cryptographic primitives are discussed in parts of the crypto industry and how they are formally defined and evaluated in modern cryptography.

The discussion referenced standard cryptographic requirements that separate “hash-like” outputs from actual cryptographic hash functions. In particular, it highlighted that some functions can be computationally difficult to reverse yet still fail the additional properties required for secure hashing. As noted in the source material, linear functions may be computationally difficult in certain mathematical settings (including ideal lattice contexts), but linearity typically prevents them from satisfying the full set of properties expected of cryptographic hashes.

The material also contrasted cryptographic hashes with checksum algorithms such as CRC-32 and other cyclic redundancy checks. Checksums are designed to detect accidental errors in transmission or storage and are built to meet significantly weaker requirements. As a result, they are generally considered unsuitable as cryptographic hash functions, especially in adversarial settings where an attacker may deliberately craft collisions or manipulate inputs.

The broader context is that lattice-based cryptography is widely studied as a foundation for “post-quantum” security, while hash functions remain a core primitive used across blockchains for integrity, linking data structures, and commitment schemes. Hoskinson’s comments underscored that these tools serve different purposes and must be evaluated against different security properties—even when the terminology used in casual crypto discussion blurs the lines.

• Hash functions are expected to satisfy strong properties such as collision resistance and preimage resistance under adversarial conditions.
• Checksums like CRC-32 prioritize error detection, not adversarial security, and are typically not appropriate substitutes for cryptographic hashing.
• Lattice-based constructions can provide hard mathematical problems for cryptography, but “hard to invert” alone does not automatically make a function an appropriate cryptographic hash.