Founder of Starknet in person: Reflections after the airdrop

CN
3 months ago

Original Author: Starknet Founder Eli Ben-Sasson

Translation: Odaily Planet Daily Azuma (@azuma.eth)

Starknet Founder's Reflections on Airdrop

The first phase of the Starknet airdrop activity (Provisions) has ended, and I would like to share some of my personal feelings. None of the content described below constitutes investment advice, nor does it necessarily reflect the position of StarkWare or the Starknet Foundation. DYOR.

What is Starknet? Why STRK?

Starknet is a ZK-Rollup network that we launched in alpha mode in November 2021. It aims to use STARKs' cryptographic protocol to scale Ethereum without compromising its core principles, such as decentralization, transparency, inclusivity, and security.

The STRK token empowers those who wish to contribute to the success of the ecosystem in governing, operating, and protecting the Starknet network. STRK has three main use cases — governance, paying gas fees on Starknet, and participating in Starknet's consensus mechanism.

The Starknet Foundation is distributing STRK to valuable community members who have already proven their willingness to advance, maintain, and protect Starknet through various activities, such as Devonomics, Catalyst, DeFi Spring, and the focus of this article, Provisions.

On February 14, 2024, the Starknet Foundation announced the first round plan for Provisions, expecting to distribute up to 700 million STRK tokens from the 9 billion STRK reserved for the Provisions activity. The claiming lasted for four months, from February 20, 2024, to June 20, 2024, resulting in approximately 5 billion tokens being claimed, with approximately 4 billion STRK remaining for future airdrop rounds.

What is the goal of Provisions?

The main goal of Provisions is to distribute STRK tokens to a wide range of individuals, i.e., real users, who will be active on Starknet and contribute to its security and governance, thereby helping to advance the decentralization of Starknet. Starknet is both a technology and a social tool, allowing individuals and societies to achieve any social function they need, such as currency, assets, and governance. Therefore, the security of Starknet is directly related to the size and resilience of the community caring about it.

One major challenge faced by Provisions is that "blockchains cannot represent real humans." By this, I mean that the basic unit on the chain is an account address, not a human or user, and there is no clear correspondence between the two; one person may control multiple addresses.

Based on the existing on-chain information, it is difficult to determine which accounts are more likely to contribute to the future operation, security, and governance of Starknet. In other words, our problem is, given the data on accounts and their activities on-chain and elsewhere, how to fairly distribute STRK to individuals who are aligned with Starknet's long-term mission?

First and foremost, existing data is fundamentally insufficient to precisely solve this problem. Everyone involved in the design of Provisions realized early on that all potential calculation methods can only produce a relative result, without accurately achieving all goals — some users completely aligned with Starknet's mission may receive a small amount of tokens, or even no tokens at all; while users less aligned with the mission may receive a large amount of tokens.

Given the public criticism faced by Starknet Provisions, as well as subsequent airdrops such as Eigenlayer, ZKsync, and LayerZero, I believe it is necessary to clarify this. To my knowledge, there are no existing solutions that can achieve the above goals more accurately or fairly than us. Adopting other metrics would also lead to inaccuracies in different scenarios.

How did the Starknet Foundation design the airdrop?

The Starknet Foundation included six groups in the airdrop, with the distribution within each group based on metrics/data relevant to that group.

  • Starknet Users: Mainly considering address activity metrics, and a third-party witch screening has been conducted;

  • STARK Early Adopters: Distribution based on the usage of StarkEx before Starknet;

  • Ethereum Contributors: Mainly including individuals who have contributed to Ethereum in various ways (staking, development, submitting Ethereum Improvement Proposals, etc.), with specific metrics for each subcategory;

  • Github Developers: Distribution to selected developers of open-source projects on Github based on Github activity metrics.

  • Early Community Member Program (ECMP): Individuals who have contributed to the Starknet ecosystem through activities and community development can apply for tokens in advance. A committee composed of ecosystem members will decide on the distribution based on the application review results.

  • Developer Partners (DP): Infrastructure developers who have previously reached agreements with the Starknet Foundation can also receive token distributions. This is an agreement reached in advance with developer teams by the Starknet Foundation.

In summary, the fundamental idea of Starknet is to attempt to distribute STRK to a diverse range of groups based on their past actions and contributions, believing that these groups are well-suited to operate, care for, and protect the future of Starknet.

Has Provisions achieved its goals?

As mentioned earlier, due to the inadequacy of metrics, we were clear from the start that the distribution of STRK could not be entirely accurate. This raises the following questions: Have we made the best effort based on the data at hand? How do we evaluate the distribution results — to what extent do the addresses included in the airdrop correspond to real humans/users?

  • Among the six groups mentioned above, the last three groups can be confirmed to correspond one-to-one with real humans, and we can even further speculate that these individuals are likely to continue caring about the future of Starknet.

  • For the third group (Ethereum Contributors), except for the staking sub-group, most of the sub-groups included in the airdrop likely meet the standard of "one address corresponds to one user," and their past actions have indicated their willingness to care about the decentralization process, so we can hope that they will also care about and help Starknet.

  • The second group (StarkEx Users) as early adopters of STARK technology, is the group with the lowest claiming rate and airdrop scale (only 2.4 million STRK claimed, less than 1% of the total allocation), and can therefore be disregarded.

  • The most difficult to evaluate is the distribution results for Starknet Users, this group received over 87% of the airdrop share (over 430 million STRK). Public dissatisfaction after Provisions was mainly focused on the distribution to this group.

There has been a lot of discussion on social media about this matter, most of which is very negative. Many people have mentioned the issue of the balance threshold — Starknet requires holding at least 0.005 ETH on a specific date. There have also been other controversial events, such as strong language from a senior StarkWare executive that sparked community outrage, followed by a quick apology; and criticism of the unlocking plan for StarkWare shareholders (including investors, founders, and employees), which was later revised.

The criticism of the "0.005 ETH threshold" and the "electronic beggars" has persisted for a long time. Although the criticism of these two issues has greatly diminished recently with the emergence of new airdrop controversies, it has not completely disappeared to this day.

How should we view this community anger? To what extent does it come from professional farming teams trying to rationally influence the current and future (not limited to Starknet) airdrop standards? To what extent does it represent a certain group (farmer or non-farmer)? If different distribution methods were adopted, would they contribute to the long-term success of Starknet? These are research questions for which I hope to see answers. If you have a way to address this issue, please post your suggestions on the Starknet community forum and @ me.

So far, I have discussed the social media sentiment related to the distribution of Starknet users, and now it's time to consider a bigger issue. Was the Starknet airdrop done well? The answer is, I don't know, because we lack the metrics needed to answer this question, which is the same problem as our inability to accurately complete the token distribution initially. On-chain metrics available, such as TPS, TVL, STRK price, and the number of addresses, do not directly answer the following question — "Are the holders of STRK a diverse and varied group? Will they stay and continue to improve, operate, and protect Starknet?"

I also want to get the answer to this question. If you have ideas on how to address this issue, please post on the Starknet community forum and @ me.

How do I feel personally?

This question sounds a bit strange, but I believe many people want to hear the answer. The entire team has been under tremendous mental pressure in the Provisions work, especially Abdel and I, who have been the targets of personal attacks.

To deal with the Twitter feed filled with false information (and even worse), we have relied not only on mutual support within the Starknet Foundation or the StarkWare team, but also on the unwavering support of the amazing ecosystem of Starknet. Although this period has been difficult, it has ultimately proven to be valuable, highlighting areas that need improvement and testing the resilience of our team.

We have learned the importance of resolute decision-making, while also realizing the need to be open to constructive feedback, even if it is harsh. This experience has strengthened our belief that handling public pressure in the crypto space is as important as technical decision-making. It has been very encouraging to see people from other ecosystems (sometimes even our competitors) reaching out and offering support, and I will never forget this help. Most importantly, we have drawn strength from the amazing Starknet ecosystem.

How to do better in the future?

There are still approximately 4 billion STRK tokens to be used for future airdrop rounds. How can we do better?

Clearly, "identity verification" on the blockchain is a very difficult problem to solve, and we cannot be sure if it can be solved, but this is exactly what we are passionate about in our scientific and technological research.

The motivation of professional airdrop farming teams to influence future airdrop rounds remains strong, which means that whatever we do will cause public protests on social media. For me, this is an unavoidable and unpleasant part of the cryptocurrency industry.

I hope that the Starknet Foundation and its Provisions-related teams can find new solutions to distribute tokens to a diverse group that cares about the long-term vision and mission of Starknet and is willing to stay and help ensure its secure operation. I know this is their desire, and they are researching and discussing ways to achieve this goal.

In conclusion, Provisions aims to place STRK in the "right hands." Honestly, I don't know if the design of the Starknet Foundation (especially the distribution scheme for Starknet users) is precise enough, and I hope future community research can answer this question. I will definitely continue to ponder this question and plan to share my suggestions at some point in the future.

I would love to hear more thoughts on the token distribution mechanism from people inside and outside the ecosystem, if any, please speak freely on the Starknet community forum.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
Download

X

Telegram

Facebook

Reddit

CopyLink