After the introduction of ordinals, BRC-20 tokens, and more recently, Runes, the next significant wave of innovation on Bitcoin will revolve around Decentralized Autonomous Organizations (DAOs). Aisling Connolly, the Senior Research Scientist at the Dfinity Foundation, has asserted this. He stated that the technology, which is now well-suited to support DAOs and the more mature ecosystems, makes a trillion-dollar DAO a distinct possibility.
One of the Greatest Coordination Tools
In written responses provided to Bitcoin.com News, the Dfinity Foundation scientist also commended the Bitcoin network for achieving an important milestone: the removal of a central point of trust. This achievement led to the creation of “one of the greatest coordination tools” globally, according to the senior scientist. However, Connolly believes that true decentralization will only occur when smart contracts are “governed in a decentralized manner and deployed autonomously.”
Meanwhile, when asked why Bitcoin enthusiasts favour the Unspent Transaction Output (UTXO) model, it was explained that it allows UTXOs to be bundled together in a single transaction, making it more predictable. Therefore, while the programming model for UTXO-based systems may be perceived as slightly more complicated, Conolly said the benefits of efficiency and functionality “are deemed to outweigh the cost.”
Turning to decentralized finance (defi) on the Bitcoin network, the Dfinity scientist noted that the protocol’s recent advancements, such as ordinals and Runes, enable the network to “accept data similar to NFTs and fungible tokens found in other ecosystems.” Below are Connolly’s responses to all the questions posed.
Bitcoin.com News (BCN): Bitcoin has been a digital store of value and it has proven its value. Why is there a rush among developers to build decentralized finance (defi) use cases with Bitcoin, in essence making it compete with Ethereum? In your opinion, does bitcoin need defi or its defi which needs bitcoin?
Aisling Connolly (AC): Until recently, due to its simple design, the Bitcoin protocol could not support functionality beyond basic storage and transfer of value. However, there have been a number of advancements that allow Bitcoin DeFi to become more realistic. New meta-protocols (like Ordinals and Runes) were developed to encode information in clever ways and hence extend Bitcoin to accept data akin to NFTs and fungible tokens found in other ecosystems.
Infrastructure providers, like alternative L1s and other layers, have made advances in their technology that allow the use of Bitcoin on other networks in more secure ways. DeFi builders struggle with the high gas fees and low latency of Ethereum and question the security offered by more scalable networks. Leveraging Bitcoin for building DeFi is highly desirable and, due to the aforementioned developments, now started to become more realistic.
BCN: One of the key differences between Runes and BRC-20 tokens is that Runes uses an Unspent Transaction Output (UTXO) model just like Bitcoin, rather than the account model used by chains like Ethereum. Can you explain UTXO and account models for our readers and shed light on why Bitcoiners believe the UTXO model to be superior?
AC: In the UTXO model, the blockchain keeps track of individual coins. Like cash, transactions consume the UTXO (banknote), and produce resulting UTXOs (the payment and the change). In the account model, the blockchain keeps track of balances associated with accounts. This is more like online banking where transactions directly update the balance of other accounts.
Take the example of Alice who has 5 tokens and wants to send 2 to Bob.
In the UTXO Model
Initially: Alice has a 5-token UTXO
Transaction: Alice sends 2 tokens to Bob
Create a 2-token UTXO (payment) for Bob
Create a 3-token UTXO (change) for Alice
In the account model
Initially: Alice’s balance: 5 tokens, Bob’s balance: 0 tokens
Transaction: Alice sends 2 tokens to Bob
Alice’s new balance: 3 tokens
Bob’s new balance: 2 tokens
People argue that the UTXO model is better because, like with cash, many UTXOs can be bundled together in a single transaction. It’s also more predictable, as Alice could not make a transaction without an initial UTXO, but she could try to make a transaction from an account with 0 balance, which will fail and result in lost gas fees. Although the programming model for UTXO-based systems is a little more complicated, the benefits of efficiency and functionality are deemed to outweigh the cost.
BCN: The ICP is said to talk directly to the Bitcoin network and recently announced the Threshold-Schnorr integration that would supposedly increase functionality. Could you explain in simple terms what this integration means for users as well as developers interested in building Bitcoin-native Web3 use cases?
AC: ICP has multiple distributed signing services that allow smart contracts to make bitcoin transactions in a secure way. This, coupled with the fact that ICP runs bitcoin adapters on the network, means that smart contracts can read, write, own, and program Bitcoin. Users can then leverage the features of the ICP network to use Bitcoin with 1-2 second finality and negligible fees.
The new Threshold-Schnorr implementation will allow developers to build the full suite of DeFi protocols without needing to rely on any centralized components or inefficient and insecure workarounds.
BCN: If the Bitcoin network can support NFTs and tokens, in your view, what other innovative applications could also be built with the original blockchain?
AC: If you think about the original goal of blockchains, it was to remove the central point of trust. That has been achieved particularly well in Bitcoin, and, as a result we’ve built one of the greatest coordination tools in the world!
Decentralizing the infrastructure is a good first step. It allows us to hold tokens, show off our NFTs, and experiment with DeFi. However, to truly remove any central point of trust, smart contracts should also be governed in a decentralized way and deployed autonomously.
I believe the next big wave of innovation on Bitcoin will be around DAOs (decentralized autonomous organizations.) We’ve seen the emergence of some DAOs during 2021/2022 in the Ethereum ecosystem, and we’ve heard whispers of network states starting to emerge. Now the technology is much more advanced to support them, ecosystems are more mature and form more serious entities, and the principles are aligned. I can’t wait to see the first trillion-dollar DAO.
BCN: The Internet Computer blockchain went live on the mainnet about three years ago. And you recently released a new roadmap. Going forward, what’s going to be your key areas of focus?
AC: This year we are heavily focussing on Chain Fusion (our multichain tech stack), and Decentralized AI. The original goal of the Internet Computer was to decentralize computation on the web. We started by developing our own ecosystem of smart contracts and integrating with Web2. Given that the ultimate mission is to build Web3 as the standard future internet, our goal now is to make it easy and secure for developers to build applications across all of the blockchains (bitcoin, ethereum and others) and Web2 services (data feeds, email etc).
It’s still the case that large portions of decentralized apps run on centralized infrastructure, we aim to change that.
BCN: With the growing convergence of artificial intelligence (AI) and blockchain, do you envision AI-powered smart contracts changing defi? If yes, in what ways?
AC: There are many ways that AI will influence DeFi. If you look at the trad-fi industry, AI is used for things like fraud detection, AML services, fees or rates calculation, and market analysis. All of these use cases can immediately be transferred over to the DeFi realm.
As DeFi is technically more advanced, it’s likely that new features and a higher level of automation will be built too. Smart contracts could be programmed to have personalized experiences like having an AI portfolio advisor and algorithmic trading are all possible.
BCN: While certainly promising, artificial intelligence (AI) is not without flaws. Since there’s no real way to validate the sources of data, AI models could be trained on malicious or tampered data that could affect its accuracy. Many have called for running and training AI models on-chain for the sake of trust and transparency. Do you believe that the blockchain infrastructure today is capable of handling substantial computational resources and data sets that the AI models require?
AC: Verifiability of computation is one of the great things that blockchains have brought to the internet, and this applies massively to the AI space. It has been shown (on ICP at least) that smart contracts can support AI inference. AI training is more difficult as the computational requirements are larger. Clever techniques to reduce the size and complexity of training could be used, as well as other verifiability tools, like zero-knowledge proofs.
What are your thoughts on this interview? Let us know what you think in the comments section below.
