Hey HN, this is jbenet -- an author of Filecoin.
We think it's great that people ask hard questions, and get involved. It's great to see others studying our work and we really appreciate the open discourse. There are a few things from this article I’d like to address. (Despite the length of this post...) these are quick comments, and not a proper in-depth response.
- (a) The article gets some things right and some things wrong -- there is good summarizing of several of our projects, and discussion of many difficult aspects in these projects. The article discusses many technological aspects in good depth, and highlights difficulties in building these systems, aligning incentives, and the trials of past projects. The article also has significant inaccuracies. For example, the sale figure -- which appears in the title and impacts the analysis -- is incorrect. We raised $205M -- officially here: https://protocol.ai/blog/filecoin-sale-completed/
- (b) The authors chose a provocative title. As some commenters have already pointed out, the conclusion is “[we] believe that it is not one.” Despite Betteridge’s law ( https://en.wikipedia.org/wiki/Betteridge's_law_of_headlines ), many people who only read the headline will come to the opposite conclusion, and now we (not they) will have the burden of correcting those misunderstandings. Provocative titles, though they may drive imagination and clicks, can do a huge disservice to everyone in the space, and contribute to misinformation. Most people will only read the title, maybe the abstract, and use that to form and drive opinions. We choose titles of our research with diligence and care, and hope others do the same.
- (c) The article has a great technical overview of the Filecoin stack, and how it fits with IPFS and libp2p. This is a large structure with many pieces, and it is rare to see articles grasping how all the pieces fit together so well, and then explaining it cogently. In particular, it’s great to see this article diving deep and discussing advantages and disadvantages of low level technical structures (multihash, ipld, libp2p, and more). We modularized everything in the hope to generally improve peer-to-peer systems, and improve reusability. We hope these components will be useful to the author’s Tribler project (a network similar in goals to Filecoin), and we hope that we can also learn from and leverage solutions they have made.
- (d) many of the objectionable things described in this article are common in ICOs in general. Put another way, consider those claims also in terms of other significant token sales, such as Ethereum, Tezos, Polkadot, Blockstack, Cosmos, Golem. People were saying similar things about Ethereum when they did their sale in 2014. Perhaps worth doing a survey / analysis over all of them, comparing and contrasting the different things groups have done, how the ecosystem has improved, and suggest new directions.
- (e) It’s worth mentioning that the analysis gives a definition of a ponzi scheme, but their discussion does not map to that definition. Instead, the discussion centers on claims about future investor sentiment or speculation as the driver of value in the token, which is not the only way to establish value in token networks, and ignores the value of the services provided. That kind of analysis does not work for projects like Ethereum and other live and functioning crypto tokens. If the network is useful, and there is a way to generate or introduce value, by providing new or better services, and if the network can capture that value in the token itself (important step), then the tokens can hold value, based on the utility of the network as a service and not just or primarily speculation. Networks like Ethereum, Bitcoin, Zcash, and Filecoin aim to provide useful services, and much of the value stored in their tokens will be thanks to the utility of the networks.
Perhaps it’s worth pointing out that most crypto token projects are compared to ponzi schemes at some point :(
- http://www.google.com/search?q=is+bitcoin+a+ponzi+scheme
- http://www.google.com/search?q=is+ethereum+a+ponzi+scheme
- http://www.google.com/search?q=is+zcash+a+ponzi+scheme
- http://www.google.com/search?q=is+tezos+a+ponzi+scheme
- (f) The article discusses the SAFT and assurances to investors, but does not discuss them in contexts of other token sales and ICOs. Most ICOs are structured as
donations
(not investments) to a project, with little to no legal recourse -- even though many people refer to these “donations” as being “investments”. In our case, we raised investment through an instrument (the SAFT) that is a direct liability to us, and gives investors greater guarantees on the completion of the project, or consequences otherwise. If we fail to deliver the network, we must return the proceeds of the token sale. Few token sales ever have such a clause. Our structure gives investors greater accountability, not less. The article discusses this in sec IV, but does not take into account that startups are similarly risky (i.e. that startup dissolution events return only remaining capital from the efforts), and does not mention how our structure improves on the ICO landscape in general.
- (g) We do share the legitimate concern that ICOs need stronger accountability, and some are structured in a way that leads to abuse. The community as a whole needs to raise the bar on accountability and ethical behavior. We have taken significant steps in this direction, not just in our sale but to improve the ecosystem -- the SAFT project, which was a gargantuan undertaking that many other networks are now using, is one example. Many other networks are introducing and improving structures. We believe token networks present a very important new way to form capital, with promising advantages to users, investors, and creators, but the space is still in its infancy, and significant changes are still ahead. Token sales have improved dramatically in the last three years, and we hope they continue to improve to find the right balance and protection of the interests of all parties involved with the network.
Thanks,
Juan