Author: Blue Fox Notes
This long post by Vitalik can be considered an updated version of the future Ethereum roadmap. It builds upon the classic 2023 roadmap (Merge → Surge → Scourge → Verge → Purge → Splurge), overlaying it with the current Strawmap (the 'strawman roadmap' maintained by the Ethereum Foundation).
The new roadmap shows us the vast and immense panorama of Ethereum's future engineering.
Simply put, the overall direction remains largely the same, but with today's development, the priorities and implementation paths have changed significantly. Furthermore, several new elements that didn't exist in 2023 have emerged.
Parts that are still continuing:
• Continued improvement of the consensus layer (single-slot finality, support for more validators, etc.)
• Data availability scaling (continuing to increase blob size)
• Light client verification becoming easier (using SNARK/STARK proofs)
• Protocol simplification, cleaning up historical technical debt
These things will still be worked on, but they have been re-prioritized, and some have found new implementation methods.
Main adjustments:
Most importantly, quantum security has been significantly moved forward. Previously a low priority, it is now a high priority.
The reason is that the perceived real-world threat from quantum computers has become much stronger than in 2023.
Downgraded technical parts:
• VDFs (Verifiable Delay Functions) are essentially marginalized
• Many pure EVM improvements have also been downgraded
Replaced technical parts:
• Verkle Trees → Unified Binary Trees → Ultimately moving towards PBT (a more optimal structure)
• State Expiry → New State Types, a completely different state management paradigm to better support scaling.
Finally, the new parts (not present in the previous roadmap)
This is also the part Vitalik finds most noteworthy:
First, privacy becomes one of the highest priorities.
It is to be directly written into the core design. This includes keyed nonces, mechanisms related to FOCIL, lean privacy pools, wormholes, etc. The ultimate goal is to make privacy a native protocol capability, upon which the application layer can build and optimize.
Second, aggressive scaling (under post-quantum assumptions)
Using leanSPHINCS signatures + aggregation, and zkzk frames (zero-knowledge framework within zero-knowledge framework). The idea is to ensure scalability does not degrade even after a full transition to post-quantum cryptography.
Third, specification 'Leanification' + Formal Verification
Writing the protocol cleaner and simpler, facilitating complete formal verification. Vitalik explicitly stated: Comprehensive formal verification has now become feasible primarily because of modern AI tools.
This is an important underlying change; protocol complexity is now being actively kept low and cannot be allowed to grow increasingly complex.
Fourth, Blob and Gas Futures
Pre-committing to/marketing future capacity.
Fifth, Native Rollups
SNARK technology has now matured to the point where deep integration of Rollup capabilities into the protocol itself is no longer just an idea.
(This is very important for L1/L2 integration)
Sixth, The Future of the EVM is Reopened
The protocol may in the future expose a simpler, more modern instruction set (current candidates are leanISA or RISC-V).
The EVM could become an intermediate representation (IR) running on top of these new instruction sets, not forever hardcoded into the protocol core. This means future developers might directly use more efficient instruction sets, with the EVM becoming a compatibility layer.
Two core ideas behind the above technical roadmap:
First, no longer attempting to infinitely scale 'all Ethereum activities'.
It will design specialized mechanisms with certain restricted properties but more suitable for scaling, specifically handling today's heaviest loads (transfers, swaps) and tomorrow's heaviest loads (privacy protocols).
Trading specialization for higher scalability.
Second, treating recursive STARK and AI-assisted formal verification as truly first-class infrastructure.
A primitive called 'aggregate to union verified dependencies' is expected to be used simultaneously in three places: the execution layer (EL), consensus layer (CL), and data layer (DL).
This kind of cross-layer reuse necessitates formal verification for safety, which in turn relies on modern AI. This is a deep 'bet'.
In summary,
Ethereum is moving from the 'second generation after Merge' towards a third-generation overhaul (Lean Ethereum).
Almost every major component will be replaced or rewritten:
For example, verification becomes recursive STARK, cryptography becomes fully post-quantum, privacy becomes a native goal, the protocol itself is proactively simplified to support formal verification, while scalability is greatly enhanced through specialized mechanisms and Native Rollups.
The ultimate goals to achieve are:
Quantum security + Privacy first + Security + Decentralization + Highly scalable under the conditions of all the above + Maintaining leanness.
In short, it's not minor fixes, but rather a systematic refactoring, arguably the largest engineering push in the entire crypto space.
