What is DFinity and why should you care
In today’s apocalyptic cryptoland of crazed alpha-seeking zombies blindly dragging behind every DeFi spinoff with barely a logo to show, DFinity stands out with its global research team of notable CS scholars, it’s 5 years in the making and its almost $200M in funds raised, but most of all, it’s bold mission statement of becoming the true infinitely scalable internet computer. DFinity claims to be the only blockchain to fully deliver on the decentralized revolution wet-dream: taking down the web2.0 giants the likes of Amazon, Facebook and Twitter, creating a world of fair, user-owned ecosystems that can not be controlled or stopped by centralized authorities. It’s recent genesis launch followed by its ICP token peaking at $465 ($68B marketcap), call for a critical review of the strengths and weaknesses of DFinity and the feasibility of it fulfilling its vision.
DFinity for Dummies (A highly simplified overview of DFinity’s tech)
In a nutshell, DFinity is a strange hybrid between a blockchain and a cloud-compute system. Rather than the common monolithic single blockchain structure, DFinity is designed as a collection of small blockchains hosted on a network of participating datacenters. These blockchains (Subnets in DFinity lingo) can inter-communicate, split and scale as needed. Developers can upload Canisters (software modules equivalent to smart contracts) to these subnets that can handle web front-end, backend and data management tasks at internet speed and with transparent infinite scalability.
Here are the main properties that make DFinity’s infra unique:
- Coding simplicity: canisters can be written in any language that compiles to web assembly (including a proprietary language named Motoko) rather than the much-hated and programmatically limited Solidity or its equivalents.
- Data persistence is immediate and transparent without the need for complex external database management.
- In addition to blazing-fast transaction time, scalability can be further enhanced by using multiple-canister architecture and by splitting subnets/blockchains.
- All this while maintaining tamperproof data and true decentralization/censorship resistance.
In other words, developers will be able to create autonomous applications on DFinity that automatically scale to infinity and are governed, maintained (and paid for) by users without ever having to deal with rent-seeking, bill-bloating, cost-hiding, potentially censoring third parties (looking at you AWS).
All this goodness is enabled by a host of algorithmic/tech innovations including a fast and energy efficient consensus mechanism based on Random Beacon and other techniques (as promised — this is for dummies).
Criticism, strength, weaknesses, thoughts
The following is a list of points to consider pertaining to the strengths and weaknesses of DFinity and it’s likelihood of delivering on its promise and becoming the next crypto giant.
DFinity’s capacity to support high-load applications such as Facebook and Twitter is, at least at this stage, a hypothesis at best. Without taking anything away from DFinity’s elite dev team, no infrastructure of this magnitude has ever been built by a single team of <100, in five years and without the experience gained from servicing real world high-load apps. There is a real and non-negligible probability that the Internet Computer simply will fail to meet the needs of such applications. This doubt will only be removed when a number of high throughput applications are successfully running on DFinity.
At present, a few demo projects have been built and presented (i.e. CanCan — a TikTok look alike, and LinkedUp — a decentralized LinkedIn, none of which are live at the moment.) However these are no more than simplified coding examples, light years away from real-life TikTok/LinkedIn alternatives.
Let’s take for example a one of the many facets of high load application delivery: CDN (Content Delivery Network). High performance apps rely on a complex stack of best of bread CDN Providers, with decades of experience and dedicated development. (The real web giants such as Facebook use homegrown solutions as no commercial provider can meet their specific needs).
On Dfinity (according to its documentation), CDN is referenced briefly as one of the built-in properties of its networking protocol stack. In fairness, I did not get to look under-the-hood of DFinity’s CDN implementation, however the chances that this miniscule feature developed as part of DFinity delivers anything near what the CDN industry has to offer are slim.
DFnity node participation is permissioned, limiting it’s censorship resistance and security in general. At Genesis DFinity was hosted on 48 participating datacenters, a small enough number to enable a central authority to take down the entire network. While this may change over time, the fact that the network does not enable anyone with a laptop to spin up a node in no-time to make up for shutdown nodes (as is the case with bitcoin) is a major limitation. Nevertheless, this can be alleviated in the future if hardware/infrastructure improvements enable nodes to be set-up faster and with a smaller investment.
Adoption and Use Cases
The DeFi Resilience Use Case
Acknowledging (at least in subtext) that its blockchain is not likely to be deemed secure enough for DeFi core protocols, DFinity offers DeFi projects better censorship resistance by hosting their front-end layers instead (currently hosted on centralized services such as AWS). The value proposition being that these Dapps are not truly decentralized as the hosting service can decide to cut them off.
This claim is somewhat weakened by the fact that DeFi protocols can continue to function even when their front-end hosting is cut off, since the critical information (user balances and transaction history) persists on the blockchain and can be revived with alternative front-ends hosted elsewhere.
The Decentralized Social Network Use Case
One of DFinity’s favorite examples for possible use cases are decentralized social networks, pointing out recent privacy scandals, censorship incidents and user-data exploits as reasons for the inevitable adoption of decentralized alternatives. Past attempts at blockchain social media (Steam, Hive) have all failed to gain traction, though performance issues were never the reason. The more likely reason is that privacy and censorship resistance are not strong enough value propositions to generate the network effects required to disrupt the social giants.
The Fair Economy Use Case
Censorship resistance aside, it may turn out that DFinity’s most appealing feature is its simplicity for creating and hosting community managed Dapps. Things like data persistence/management, scalability, availability, data security, user management and load balancing are transparent and effortless on DFinity (at least on paper), eliminating the need for complex/centralized IT and support teams. DAO governance is also natively supported, as are token systems that help create advanced economic mechanisms for value creation and distribution.
With that in mind, allow me to speculate the first Dapps to achieve significant traction on DFinity would be marketplace/gig economy alternatives that keep most of the value in the hands of participants and allow them control over major decisions. Recent phenomena that support this thesis are the meteoric rise of NFTs and relative success (in crypto terms) of Audius (a SoundCloud alternative on a blockchain), both demonstrating the hunger for fair value creation/distribution platforms and its capacity to draw normies into cryproland. A DAO managed Uber/Upwork/Fivver/AirBNB/Wolt alternative on DFinity could be the first killer app.
Team, Community, Investors
This is by far one of the strongest bullish signals for DFinity. A quick and partial review of people involved shows a fair amount of PhDs (Mathematics, Computer Science), top MBA grads and senior alumni of prestigious tech companies. Major Investors such as Andreessen Horowitz and others complete the picture, and last but not least, the sheer fact that the team pushed through five years of development of this impressive engineering enterprise, traversing multiple bull/bear crypto cycles without faltering, serves as prove of long term vision and tenacity not to be taken for granted in cryptoland.
DFinity is built on a bold and unique vision that, should it become a reality, will make it one of the heavy-weights of crypto. None of the existing alt-blockchains is set up to support DFinity’s use cases. While some (EOS, Cardano, Solana) also boast blazing fast transactions they lack the native architecture support for full-stack decentralized Dapps. This means if things go DFinity’s way, it will be hard for a contender to catch up to its advanced development stage. Two big question marks remain, forming DFinity’s risk profile: Tech feasibility and traction. While DFinity’s team and cred can help alleviate the former, the latter is more a question of chance and will be resolved with time. It took Ethereum a long while to find real use cases, and DFinity may find itself in the same multi-year limbo Ethereum was in post cryptoKitties. But if the vision is eventually realized, even the current 40B marketcap seems small for a potential AWS/GCP/Azure killer, hosting tomorrow’s Facebooks, Ubers and more.