better.codes launches to raise hash-based SNARK security
The Ethereum Foundation’s Formal Verification team has launched better.codes, an open autoresearch challenge aimed at strengthening hash-based SNARK security benchmarks through agentic collaboration. Participants run their own AI agents to improve the machine-checked soundness lower bound of koalaIRS12, a Reed–Solomon proximity problem.
Each submission is validated by the Lean kernel against a pinned theorem statement, parameters, and verification harness. The score is measured in “bits,” and every promoted proof increases the proven bound toward a fixed 128-bit target. New lemmas, techniques, and impossibility results are upstreamed to a public leaderboard so other solvers can build on verifiable progress and avoid dead ends.
The article ties the work to Ethereum’s post-quantum roadmap and to production hash-based SNARKs used across zkrollups and zkVMs, where security depends on proximity gaps and correlated agreement conjectures. better.codes is designed to narrow the gap between conjectured benchmarks and formally proven benchmarks.
This launch focuses on the soundness challenge for koalaIRS12; the program says additional challenges may be added over time. Key figures include the initiative authors Gal Arnon, Dan Boneh, and Giacomo Fenzi (via the earlier Proximity Prize/Open Problems work).
Neutral
This is a cryptography research and tooling initiative (better.codes) focused on formally verified security benchmarks for hash-based SNARKs. It does not directly change Ethereum’s protocol parameters, tokenomics, or near-term network usage. As a result, market impact is likely limited and indirect.
In the short term, traders may react mainly to “Ethereum post-quantum readiness” narratives, but there is no immediate catalyst tied to ETH supply/demand or on-chain adoption. Historically, research announcements around ZK/cryptography (e.g., verified circuits, proving performance benchmarks) often create modest hype, then fade unless paired with engineering milestones like audited implementations or shipped client changes.
In the long term, improved, machine-checked soundness benchmarks can strengthen confidence in hash-based SNARK assumptions used in zkrollups/zkVMs and Ethereum’s post-quantum roadmap. That can be strategically bullish for ecosystem credibility, but it won’t translate into price momentum unless and until the research results are integrated into production systems with measurable performance or security upgrades.
Therefore, the expected impact on market stability is neutral: positive for sentiment/credibility, but not a direct trading trigger for ETH.