Another week gone by, another data debacle from Facebook.
This time, it was about its sharing of user’s private messaging data with tech giants such as Netflix, Spotify, etc. A couple of months back, it was an attack on the network that exposed personal information of at least 50 million users.
To compound matters, we then discovered that hackers could have accessed accounts on all apps of affected users that required their Facebook logins.
As consumers, it seems that centralized databases present major inadequacy in safeguarding user’s data. The guardians who we entrust with protecting us and our data are failing repeatedly.
And often, spectacularly.
Whether it’s harvesting of our Facebook user data that enabled Cambridge Analytica to create psychographical profiles for special political purposes, or our crypto holdings being stolen from centralized exchanges, the centralized model seems vulnerable against increasingly sophisticated hacking operations.
dApps to the rescue?
Speculation grows over whether their decentralized equivalents – dApps – will take their place. Can they deliver?
We’re still in the very early stages of the dApp’s development life cycle, which means that a cemented definition of a decentralized application remains undetermined thus far. But invariably a dApp will involve using smart contract code run by a peer-to-peer network (P2P) of users – rather than a single controlling entity – to connect it to an existing blockchain platform such as Ethereum.
It should be mentioned that decentralized applications have been around since before the dawn of blockchain. P2P networks for file sharing, for instance, have existed for over 20 years, with apps such as BitTorrent proving highly successful to this day.
But in this current early era of blockchain, it’s dApps such as CryptoKitties that have thus far generated much interest – and much debate – over their potential scope to achieve mass adoption. Using smart contracts, CryptoKitties allows users to collect these digital cats on the Ethereum blockchain, which is used to track ownership of these collectibles. In fact, CryptoKitties has proven so popular that soon after launching, a digital cat was purchased for a whopping $140,000. The kitties can be collected, bought, sold and traded with cryptocurrency, and users can also pay other users a ‘siring fee’ that allows their kitty to breed with theirs and produce new baby kitties.
But it’s not just novelty games that have proven popular in the dApp world. Decentralized exchanges such as Idex could draw an end of this painful era of exchange hacking. And gambling dApps such as MÖBIUS 2D could transform the traditional structure of lotteries. Indeed, dApps are growing strongly in number, with 1,947 in total being recorded; as well as their types, including decentralized file storage, governance, and financial applications.
From the outset, the core characteristics of dApps help us understand the limitations of centralized applications. Centralized architecture, for instance, involves a front-end and back-end that communicate with each other via HTTP, and often utilize an API to pull in user data for display on a website. The back-end server is usually responsible for holding all the data on behalf of the app-owner, such as Facebook.
So, while many users may exist on one side, only one entity ultimately controls the back-end. This means that centralized applications have one single point of failure, and as such, a successful attack on this point could cause the whole system to fail. It could also compromise substantial amounts of our data.
But dApps have their back-end code running on a decentralized network, which removes the need to rely on a single server to maintain the integrity of user data. Using smart contracts and cryptographic proofs, dApps can eliminate that single point of failure and secure customer data much more effectively. The decentralized network also ensures a trustless ecosystem, one that prevents a central authority from exerting singular control over the app’s business objectives. So, no entity – not even the app designer – could compromise your data, as Facebook has done on various occasions.
This prevention of control may also extend to such issues as censorship. A dApp version of Twitter, for example, could be developed whereby users can tweet messages that can’t be moderated, censored or deleted, not even by the app creators. Once posted, the messages are permanently recorded on the blockchain, meaning that they can be retained forever.
And on the monetization front, moreover, centralized apps will normally generate revenues through advertising revenues. As we have seen with the recent Facebook controversies, our user data can be used to target ourselves with highly personalized ads.
In contrast, dApps involve a native token that can be tracked across the blockchain. By having the potential to appreciate in value in the future, this token incentivizes user participation and engagement with the application, with the opportunity of being rewarded with tokens for greater usage of the dApp network. And in the absence of a central authority that takes the lion’s share of the revenues, more value can be distributed among the participants of a dApp ecosystem.
So why haven’t DApps taken over the world?
Despite their hypothetical advantages in architecture, dApps do face certain challenges, not least of which is their low user adoption rates. Even though dApps experienced substantial investment and user activity during these early stages, interest in many applications has fallen away rapidly.
A sobering report in August from research site Diar shows that the largest seven dApp platforms by ICO funds raised shed 74% of their active users, on average, from their all-time highs.
Most astonishingly, the number of daily visitors to CryptoKitties in August was recorded at just 510, representing a 96% drop from its peak just a few months prior. Decentralized exchange Bancor, meanwhile, lost a hefty 74% of its active users; while Augur, which launched in July, counted less than 50 users just a month later:
Although a tokenized ecosystem may encourage user adoptions, the incentives for both users and token issuers will have to be aligned to ensure the platform’s sustained success in the long run. As we recently discussed, this is easier said than done. Indeed, the decline in dApp usage this year is almost certainly due to falling crypto prices, which makes designing incentives for users to hold onto their tokens and remain active participants of the ecosystem during such times a distinct challenge. I will discuss these issues in depth in a future piece.
That same pesky issue of scalability – or lack thereof – that so many blockchain projects have endured also limits the usability of dApps. Although we can be critical of the architecture of centralized services such as Uber, Facebook, and Visa, they can nonetheless handle tens, if not hundreds of thousands of transactions every minute.
They completely dwarf the capabilities of Ethereum at present. Just a week after CryptoKitties launched in late-November 2017, Ethereum was recording a mighty sixfold increase in pending transactions. The popularity of the new dApp and sheer volume of transactions caused a massive slowdown for users of the platform, thus revealing the limitations of blockchain architecture in its current form.
This hasn’t gone unnoticed by blockchain leaders. Ethereum founder Vitalik Buterin remarked back in April, “If you want to build a decentralized Uber and Lyft on top of an unscalable Ethereum, you are screwed. Full stop.”
It should be emphasized that just because the underlying blockchain (eg Ethereum) may be considered decentralized, it does not mean the same applies to the DApp. In CryptoKitties, for instance, the ownership contract for a kitty is owned by a single wallet – not exactly the most decentralized solution. As we have pondered the extent to which cryptocurrencies are decentralized, so too must we acknowledge similar limitations being present in the world of dApps.
Can decentralized applications provide the solution to centralized data breaches and exchange hacks?
Eventually, yes. But there is a very long way to go before we get there. Until successful scaling solutions to blockchain are implemented, it’s unlikely we will see a decentralized Facebook any time soon.
And while there are around 2,000 dApps in existence today, it’s worth observing that there are over 2 million iOS apps available in the App Store. This shows how early we are at the development cycle. Much work remains to convince the world to shift toward a decentralized architecture. It will be a steady process to get there, but as the dApp Fund’s David Johnston has assured us, ““Everything that can be decentralized, will be decentralized.”