Vitalik Buterin opened a new technical series on June 29 arguing that obfuscation is cryptography's hardest unsolved problem. The specific target: indistinguishability obfuscation (iO), which encrypts a program so thoroughly that two functionally identical programs become computationally indistinguishable, even to someone who runs them.
The appeal is architectural. If you can hide a program's logic while preserving its behavior, you collapse the need for a trusted intermediary. A voting system, a collateral custodian, a censorship-resistant relay, any protocol normally requiring an honest gatekeeper could run on chain with zero trust assumptions. Buterin framed this as achieving a "trustless trusted third party." Pair that with a blockchain to anchor state and timestamp, and the primitives for building without intermediaries start to click into place.
But obfuscation alone cannot prevent code from being copied. That's where blockchains step in, Buterin wrote: they can anchor ownership and prevent replication. The two together solve a problem neither can handle alone.
The runtime barrier is brutal. Buterin traced roughly two decades of research that proved the ideal form of obfuscation impossible in 2001, forcing cryptographers toward iO. Recent work has shown iO is theoretically achievable "under reasonable security assumptions," but the practical cost is staggering. He described the expected runtimes as "literally galactic," with conservative parameter choices producing schemes whose execution would exceed the lifetime of the universe.
but the practical cost is staggering. He described the expected runtimes as
That's not hyperbole aimed at dismissal. Buterin drew a parallel to zero-knowledge SNARKs around 2010, when they too were mathematically proven but computationally absurd. Researchers and AI tools then spent a decade chipping away orders of magnitude. He suggested obfuscation could follow the same arc, and promised a second post examining two leading candidates: "diamond iO" and "local mixing obfuscation."
This moves sits inside Buterin's broader privacy-first research agenda. He introduced the Kohaku toolkit earlier and has mapped out an Ethereum privacy roadmap. Cracking either of the two iO approaches, Buterin wrote, would mean "we will have 'solved cryptography," letting any protocol describable through a trusted third party be built securely on chain. "But getting there," he added, "is still a formidable challenge."