Ethereum Co-Founder Vitalik Buterin Says Cryptography’s Hardest Problem Is Still Unsolved

What Obfuscation Actually Does

Most people in crypto are familiar with encryption, which hides data in transit, and zero-knowledge proofs, which verify data without revealing it. Obfuscation does something different: it hides the code itself.

Think of it this way. Encryption locks a letter inside an envelope. Obfuscation scrambles the machine that writes the letter so that no one, not even someone watching it run, can figure out how it works.

The specific version Vitalik Buterin described in his latest blog post is called indistinguishability obfuscation, or iO. In plain terms, iO means that if two programs do the same thing but use different internal logic, you cannot tell them apart once obfuscated. The program runs, produces real outputs, and reveals nothing about how it got there.

Why Blockchain Needs This

In the analysis, Buterin connected iO directly to the blockchain use case. Obfuscated programs cannot prevent themselves from being copied, which means they cannot hold state like a balance or a transaction record. That is exactly what blockchains do.

Combine the two, Buterin wrote, and the result is something close to a “trustless trusted third party.” A system that enforces rules, processes inputs, and produces honest outputs without anyone needing to trust the people who built it.

One concrete example: a secure, private, manipulation-resistant voting system with no multisig committee, no trusted setup run by a small group, and no assumption that at least some participants stay honest.

The Progress So Far

Researchers have spent roughly 20 years trying to build a provably secure version of iO. Early approaches were broken repeatedly. Buterin referenced a key result from 2001 showing that the ideal version of obfuscation is mathematically impossible. Researchers then shifted to iO as the next-best target.

In recent years, the field reached a milestone. Cryptographers now have constructions that achieve iO under what Buterin called “reasonable security assumptions.” That is genuine progress.

The Problem: Runtimes Longer Than the Universe

The catch is runtime. Current iO schemes are technically polynomial, meaning they scale with input size rather than exploding exponentially. But Buterin described the actual overhead as “galactic.”

The schemes work by stacking several layers of advanced cryptographic tools inside each other. Each layer wraps the one beneath it. The result is a system where processing a single input can require overhead factors estimated above 10 to the 10th power, with security requirements that push ciphertext sizes thousands of times larger than standard constructions.

Buterin’s estimate: expected runtimes on current schemes exceed the age of the universe.

3 Ways Forward

Buterin outlined three paths researchers are pursuing:

• Optimize the existing tower of constructions using smarter engineering and AI assistance to shave orders of magnitude off each bottleneck, similar to how SNARK performance improved dramatically after 2010.
• Build iO using more aggressive but simpler cryptographic assumptions, trading some theoretical rigor for practical speed.
• Discover an entirely new approach to obfuscation, possibly outside the lattice-based math that underpins most of today’s work.

What It Means for Crypto

Buterin framed the stakes plainly. If any of those paths succeed, the result is a version of cryptography where any protocol that can be described with a trusted third party can be built without one. That covers private smart contracts, anonymous credential systems, encrypted computation on sensitive data, and governance systems that cannot be gamed by insiders.

“If we succeed in either path, the reward is high: there is a real sense in which we will have ‘solved cryptography’: any protocol that can be described using an idealized trusted third party, provided the adversary is allowed to rewind the clock, will be implementable securely. But getting there is still a formidable challenge,” Buterin concluded.

The technology is not ready. But the roadmap, Buterin wrote, is clearer than it has ever been.