Ethereum 2023 roadmap overhauled: quantum safety and privacy now core

Ethereum’s blueprint for its own future looks almost nothing like it did three years ago. In a comparison posted this week, Ethereum co-founder Vitalik Buterin laid the network’s 2023 upgrade plan directly against the Ethereum Foundation’s current draft, revealing that the Ethereum 2023 roadmap once anchored by different priorities has been substantially reshuffled, with quantum safety and privacy now sitting at the center of Ethereum’s long-term strategy.

Key takeaways

• Buterin overlaid Ethereum’s 2023 roadmap onto the Foundation’s current “strawmap” and found that privacy, quantum defenses, and non-EVM execution engines weren’t part of the original diagram at all.
• VDFs and many planned EVM improvements were deprioritized, while Verkle trees and state expiry were dropped in favor of newer constructions.
• The strawmap targets seven network upgrades by 2029, roughly one every six months, built around five long-term “north star” goals.
• Recursive STARKs proofs and AI-assisted formal verification are now treated as core dependencies for the protocol’s future correctness and speed.
• Ethereum’s fee base has shrunk 79% in eleven months, from $39.6 million in August 2025 to $8.3 million in July, cutting into the ETH burned through transaction fees.

Vitalik Buterin Reprioritizes Ethereum’s Roadmap for Quantum Safety and Privacy

Buterin’s own framing of the shift is blunt: some items on the old plan got reordered, some were shelved, and some were swapped for better designs entirely. What stands out most, he said, is the list of things that simply didn’t exist in the 2023 version — privacy protections, defenses against quantum computers, and the idea of an execution engine that goes beyond the Ethereum Virtual Machine.

Changes from the 2023 Roadmap

Quantum safety jumped up the priority list, while verifiable delay functions, cryptographic timers meant to secure Ethereum’s randomness, and a broad swath of EVM improvement work got pushed down. Two projects that had anchored the earlier plan disappeared outright: Verkle trees, a redesign meant to shrink the data needed to verify account and contract information, and state expiry, a scheme for letting the network forget data no one had touched in years. Both were replaced. Verkle trees gave way to a unified binary tree structure specified in EIP-7864, which Buterin said will eventually evolve into a further construction he abbreviates as PBT. State expiry, meanwhile, was replaced by what Buterin called new state types — not simply a substitute, but what he described as a fundamentally different approach to how Ethereum scales.

New Focus Areas: Privacy, Post-Quantum Scaling, and Execution Engines Beyond EVM

Seven items on the current strawmap have no equivalent on the old diagram. Privacy now gets what Buterin called first-class attention, covering tools like keyed nonces, recent roots, a lean privacy pool, and “wormholes” — mechanisms designed to obscure who is transacting with whom — alongside parts of FOCIL, the fork-choice enforced inclusion list design filed as EIP-7805 that forces block producers to include transactions the network has already seen.

Post-quantum scaling is another addition, built around leanSPHINCS, a signature scheme resting on hash functions rather than the elliptic-curve math a quantum computer could eventually crack, and a proposal Buterin opened himself in June, EIP-8288, which lets transactions bundle their signatures and proofs to keep gas costs manageable despite the larger size of post-quantum signatures.

Rounding out the new list: simplifying the protocol’s specification so it can be formally verified, futures contracts on gas and blob space, native rollups, and a wider design space beyond the EVM. Buterin was explicit that some of these simply weren’t feasible before — the cryptographic proofs called SNARKs weren’t mature enough in 2023 to even consider native rollups, and blob and gas futures didn’t exist as concepts at all.

That last point — moving past the EVM — carries the most structural weight. The strawmap already assumes Ethereum will eventually expose an instruction set that isn’t the EVM, with two leading candidates: leanISA, developed in-house by Ethereum researchers, and RISC-V, an open standard already used in conventional computer chips. Buterin went further, suggesting the EVM itself could eventually become an intermediate representation sitting on top of that new instruction set rather than a hard-coded feature of the base protocol — though he acknowledged that deeper exploration here is still too early even for the strawmap.

Ethereum’s Vision and Technical Dependencies Through 2029

Buterin closed his post with six declarations that now function as Ethereum’s north stars: the network will be quantum-safe, privacy-first, secure, censorship-resistant, and highly performant and scalable, all while staying, in his words, “Lean.” Those goals are meant to be reached by 2029.

Strawmap Upgrade Schedule and North Star Targets

The strawmap — the Foundation’s living, quarterly-revised draft for base-layer upgrades — is maintained by the Architecture team of Justin Drake, Thomas Thiery, Toni Wahrstätter, and Buterin himself. It lays out seven network upgrades through 2029, arriving at a pace of roughly one every six months. Beyond quantum resistance and privacy-first design, the document’s long-range targets include transaction finality within seconds, 10,000 transactions per second on Ethereum’s base layer using zero-knowledge proofs of the EVM, and layer-2 networks reaching as high as 10 million transactions per second.

Reliance on STARKs and AI-Accelerated Formal Verification

None of that is achievable, in Buterin’s telling, without leaning heavily on two tools that barely registered in the 2023 plan: STARKs, a family of cryptographic proofs that let one computer verify another’s work without redoing it, and artificial intelligence used to mathematically verify the protocol’s own code. Recursive STARKs — proofs about other proofs — are expected to show up across three separate layers of the protocol at once: execution, consensus, and data storage.

Because so much of the system will depend on those proofs being correct, Buterin argues the only realistic path to safety is Ethereum formal verification — mathematically proving that code matches its specification — accelerated by modern AI tools that weren’t practical for this kind of work even a few years ago. “This can only be safe with formal verification, which is itself only feasible with modern AI tools,” he wrote, describing STARKs and AI-driven verification as the two things Ethereum is effectively betting its technical future on.

That framing has already drawn pushback on timing. Dankrad Feist, a researcher at the Stripe-backed blockchain Tempo and a former longtime Ethereum Foundation researcher, argued the strawmap’s own three-to-four-year horizon is too conservative, suggesting large language models could realistically compress the work into roughly a year. The strawmap’s FAQ itself acknowledges the draft assumes human-paced development and that AI-accelerated research could significantly shorten the timeline — an admission that underscores just how much of Ethereum’s next phase hinges on tools still being tested in practice.

Tokenomics and Market Dynamics: Uncertainty Surrounding Issuance and Fees

None of Ethereum’s five north stars directly addresses how much new ETH the network issues to validators each year, and Buterin’s post didn’t touch the question either — a gap the top reply on his thread went straight after. Vanessa Grellet, managing partner at Arche Capital and a board member of the Enterprise Ethereum Alliance, pressed on whether token economics had been revised, calling EIP-8363 “a critical first step.”

That proposal, known as the tapered issuance burn, would destroy a rising share of validator rewards specifically to discourage staking beyond half of all existing ETH. Core developers put it forward for consideration on Aug. 6 as part of Hegóta, the upgrade following Glamsterdam — the earliest and least committed stage of Ethereum’s process, meaning nothing about its adoption is guaranteed. The move followed two days of pushback from DeFi figures including Stani Kulechov and ether.fi chief executive Mike Silagadze. On the strawmap diagram itself, last revised Aug. 4, the tapered issuance burn box sits under a fork pencilled in for 2027 to 2028 and carries the symbol the diagram’s own legend defines as uncertainty about the upgrade ever shipping. Justin Drake, who co-maintains the strawmap, is listed among the proposal’s six authors — but Buterin has not weighed in on any of the 181 posts debating it in the Ethereum Magicians forum.

Why this matters: issuance policy directly shapes how much ETH validators earn for securing the network, and any change to that balance affects staking incentives across the entire ecosystem — from solo validators to institutional stakers.

The debate over issuance lands at an awkward moment for Ethereum’s tokenomics. According to DefiLlama data, the network’s base layer collected just $8.3 million paid as transaction fees during July represented a 79% decrease compared to the $39.6 million recorded in August 2025 — eleven months. Fees over the first ten days of August totaled only $2.1 million. Because Ethereum destroys the base portion of every transaction fee rather than distributing it, under the burn mechanism adopted in 2021, a shrinking fee base directly means less ETH removed from circulation over time.

That combination of soft fee revenue and unresolved issuance policy sits alongside a battered market price. According to CoinGecko data, ETH trades at $1,873, down 55.7% over the past year, with a market capitalization of $226 billion. Its all-time high of $4,946 was set on Aug. 24, 2025. The gap between that peak and current levels illustrates how far sentiment has shifted even as the Foundation pushes forward with one of its most ambitious technical overhauls in years.

FAQ

What major changes did Vitalik Buterin make to Ethereum’s roadmap in 2023?

He reprioritized quantum safety and privacy, deprioritized VDFs and many EVM improvements, replaced Verkle trees and state expiry with new constructions, and introduced new focus areas such as post-quantum scaling, native rollups, and alternatives beyond the EVM.

What are Ethereum’s long-term goals through 2029 according to the updated roadmap?

Ethereum aims to be quantum-safe, privacy-focused, secure, censorship-resistant, and performant and scalable by 2029, delivering roughly seven network upgrades at a pace of about one every six months.

What technical dependencies are critical for Ethereum’s future upgrades?

Ethereum’s future upgrades depend heavily on STARKs cryptographic proofs and AI-accelerated formal verification to ensure the protocol’s correctness and scalability as it grows more complex.

What uncertainties exist around Ethereum’s issuance and fee-burning mechanisms?

The tapered issuance burn proposal, EIP-8363, aimed at reducing validator rewards remains uncertain and debated, while Ethereum’s transaction fees have declined sharply, reducing the amount of ETH burned and reshaping the network’s tokenomics.

Article produced with the assistance of artificial intelligence and reviewed by the editorial team.