When banks failed us, Bitcoin emerged as a peer-to-peer money transfer solution where you could pay someone on the other side of the globe without the need to trust any intermediaries. Since then, the idea (and the technology) behind Bitcoin was generalized into a myriad of other solutions, all focusing on empowering the users and creating greater collaboration between the nodes. Ethereum particularly boosted this by introducing smart contracts.
Yet in a world spending billions on initial coin offerings (ICOs) to fund numerous blockchain promises, we still have no blockchain killer app, and Bitcoin, which remains the single most used blockchain project, is often associated with criminal activity.
Why is this the case? Are there any technical shortcomings that prevent blockchain from going mainstream?
The decentralized version of the problem
When Facebook’s Cambridge Analytica scandal erupted, blockchain offered a possible solution. However, having all the data of Facebook on the blockchain would require a massive infrastructure, as every node would need to store a copy of the entire system. According to futurist and Ted speaker Ian Khan, “If today Facebook data centers are the size of 10 Football fields, you would need a data center the size of the area of the United States to be able to provide the computing power” to run the system on current blockchain technology. That is not very practical.
The other problem is the processing capacity. Visa and Mastercard can process more than 5,000 transactions per second with capacity to process volumes multiple times that number, while Bitcoin takes 10 minutes to clear and settle a single transaction.
What is blockchain 3?
While Bitcoin was clearly the first generation blockchain (the founder) and Ethereum the second (with its smart contracts), there is still no clear definition on what the third generation should be doing, or what the technology should be. Different projects use different techniques to improve the existing blockchain framework by targeting issues such as scalability, interoperability, adaptability, sustainability, privacy and governance.
They also use different techniques to get there, one of them being Directed Acyclic Graph (or DAG for short). Contrary to traditional blockchains where each block is linked sequentially to the next block in the chain, in the DAG each “block” (or node, as they are called) can connect to many other nodes, as long as they don’t self-reference. This technique, which is best known from IOTA’s Tangle system, does not suffer from the sequential writing bottleneck regular blockchain projects face as it can process several chains in parallel. But the DAG also has its limitations.
“Having parallel strands requires less coordination between participants to insert a transaction in a DAG blockchain, possibly leading to a higher ‘write performance,’” Explains Shunsai Takahashi, founder of Takanium. “Unfortunately, for the “read operation,” instead of a single strand of a non-DAG blockchain, these thousands of the strands need to be ironed out.” This causes a problem, as a typical application performs 10x more read operations than write operations. The parallel nature of the DAG also has a security implication; while you need a 51% attack to break a blockchain network, the DAG breaks under a 33% attack.
Finally, IOTA is criticized because of its “Coordinator” system, rendering it like a centralized body. For these reasons, the blockchain 3 projects take different approaches.
The many faces of blockchain 3
Logos is a payments network, which uses a hybrid DAG/blockchain design. The “chain mesh,” as Logos calls it, creates independent transaction chains for each account. This way, they leverage the advantages of the DAG (parallel validation of transactions and substantially improved performance) and the blockchain (clear, easy-to-follow history) without being bogged down by the mathematical limits of the traditional, serial blockchain on scalability. As a result, the prototype was capable of performing over 17,000 transactions per second with a confirmation latency of 2 seconds.
COTI is another project working on extending the DAG. They have developed a multi-DAG blockchain network called TrustChain, which uses a new consensus protocol named TrustScore. The TrustScore protocol uses parallel validation: every new transaction must be validated by two previous ones in the same trust score range. The other interesting part about COTI is their Clusterstamp technique, which strips the DAG history except the most recent parts and seals the transactions after verifying them at that point.
While some projects focus on extending DAG, others focus on changing the blockchain algorithm. For instance, Bitcoin started with Proof of Work (PoW), a computationally intensive algorithm which rewards miners based on their effort in validating transactions. Ethereum, which currently uses Proof of Work for block creation, is using Proof of Stake for “finalization” (it aims to eventually fully shift to PoS). The PoS algorithm is much less energy consuming and easier to adopt, but at the same time there are concerns regarding its security. But, the innovations don’t stop there.
IOTW, a competitor to IOTA has come up with a new protocol called Proof of Assignment which is even less energy consuming, making it suitable for IoT devices. On another front, Shunsai Takahashi has presented Proof-of-Approval, which is similar to PoS but focuses on solving the security issues. Finally, CyberMiles has a Designated Proof of Stake protocol that’s not only faster than Ethereum, but also does not force miners to perform meaningless work to compete for the next block.
As you’ve seen there are many great ideas and bright minds behind solving the problems with decentralized apps, but which one will finally prevail as the de facto standard for blockchain 3 depends on many factors, some of which we might not have even accounted for. The thing is that we won’t recognize a blockchain 3 project before it has been deployed and passed the test of time in the wild.
What’s certain is for any technological solution to go mainstream, it must be easy to use and comprehend for the average people on the street – who won’t care much for the underlying technology as long as it works correctly.