Ethereum Co-Founder Vitalik Buterin published a detailed progress report on major cryptographic challenges for the last five years. While protocol problems were largely solved, more practical issues saw limited progress.
Buterin made a post in 2014 detailing a list of 16 “hard problems” in math, computer science and economics that he considered important for the cryptocurrency space. These include commonly known technical issues such as blockchain scaling and proof of stake implementations. Practical economic questions were also present, largely focused on providing fully decentralized alternatives to existing societal mechanisms. Progress has been lackluster for most entries in the list, though some of the problems were solved on a theoretical level. However, even the challenges that saw the most progress have yet to be fully battle-tested.
Core Issues Were Solved
Buterin considers blockchain scalability to be largely solved, at least from a theoretical point of view. Taking into account only sharding designs, he mentioned OmniLedger, LazyLedger and Zilliqa as alternative solutions to the upcoming Ethereum 2.0. Practice is a different matter, however, as he highlighted that “fully sharded blockchains have still not been seen in live operation.” Only Zilliqa made progress through a partially sharded implementation.
Proof of computation — the ability to cryptographically prove that a certain calculation was performed correctly — has progressed a lot in five years according to Buterin. Zero-knowledge proof technologies such as SNARKs and STARKs are now “increasingly well understood” with several proofs of concept such as STARKDEX being released.
Issues with efficiency remain, but they are generally “in the details rather than the fundamentals.”
ASIC resistance was solved shortly after the initial list was published, Buterin asserts. Ethereum’s Ethash algorithm uses a memory-heavy approach where ASICs are unable to significantly improve speed. Though the protection has been broken since then, Ethereum ASIC miners are at best five times more efficient than GPUs. Buterin considers this to be an acceptable level of centralization risk before the transition to Ethereum 2.0. “In the long term the better choice for blockchain consensus is proof of stake,” he concluded.
Finally, proof of stake developed significantly since 2014. Though its proponents had to come to terms with some compromises, many staking-based solutions are available today. More real-world evaluation is however necessary for these systems, according to Buterin.
More Work Necessary for the Rest
Only four issues are considered to be solved by Buterin, albeit with some caveats. The remaining 12 saw various levels of progress — sometimes almost none at all.
“Stable-valued crypto assets,” that is stablecoins, only made limited progress according to Ethereum’s founder. While there are dozens of stablecoin projects, only MakerDAO attempts to maintain stability through decentralized incentivization. Though the project survived Ethereum’s 93 percent price drop in 2018, he maintains that it hasn’t been tested nearly enough. Stronger and faster price declines, attacks against the underlying blockchain and exploits against the oracle system could all threaten the stability of the Maker ecosystem.
Other decentralized replicas of various economic mechanisms saw little development as well. Reputation systems, unique-human verification, contribution and success metrics, together with public goods incentivization remain mostly unsolved.
Some technical issues such as code obfuscation and physically-useful proof of work pose a significant theoretical challenge to this day.
New Challenges Emerge
Buterin concluded the report by tentatively making a new list of hard problems. Unsolved issues from 2014 are all present, though sometimes with a different focus. Notable new challenges include quantum-resistant cryptography, responses to 51 percent attacks, anti-collusion infrastructure and new Decentralized Autonomous Organizations.
Buterin concluded by noting that Base-layer issues are decreasing. Application-layer problems, however, “are only just getting started.”