Update aib.md

@yimikaerinle @aik31 please review each science note with care before publishing. Use Grammarly to eliminate spelling, grammatical and punctuational mistakes.

blogs/AIB/aib.md

@@ -12,7 +12,7 @@
 
 ## Introduction
 
-Artificial intelligence and blockchain are two progressive technologies that hold great potential in stimulating the intelligent evolution of various industries. Each possesses inherent strengths but also faces its own unique challenges. AI is gradually transforming industries by introducing advanced capabilities but struggles with issues such as interpretability and effectiveness. It operates based on three key elements: algorithms, computational power, and data {cite}`zhang2021recent`. On the other hand, blockchain, despite being an advantageous foundation for trustworthy transactions, grapples with hurdles concerning energy consumption, scalability, security, privacy, and efficiency {cite}`zhang2021recent`. Despite these separate research directions and associated challenges, AI and blockchain exhibit a natural synergy due to shared requirements for data analysis, security, and trust. This intersection between the two technologies can significantly amplify their respective strengths. According to the estimations of Spherical Insights, the Blockchain AI Market, valued at USD 230.10 million in 2021, is projected to grow to USD 980.70 million by 2030 {cite}`sphericalinsights2022blockchain`. The merger of these two technologies is an area that still calls for in-depth exploration. Moving forward, we'll examine AI in the context of blockchain, investigating in detail the possible intersections and inherent value these two technologies may possess when coalesced.
+Artificial intelligence and blockchain are two progressive technologies with great potential to stimulate the intelligent evolution of various industries. Each possesses inherent strengths but also faces its own unique challenges. AI is gradually transforming industries by introducing advanced capabilities but struggles with issues such as interpretability and effectiveness. It operates based on three key elements: algorithms, computational power, and data {cite}`zhang2021recent`. On the other hand, blockchain, despite being an advantageous foundation for trustworthy transactions, grapples with hurdles concerning energy consumption, scalability, security, privacy, and efficiency {cite}`zhang2021recent`. Despite these separate research directions and associated challenges, AI and blockchain exhibit a natural synergy due to shared requirements for data analysis, security, and trust. This intersection between the two technologies can significantly amplify their respective strengths. According to the estimations of Spherical Insights, the Blockchain AI Market, valued at USD 230.10 million in 2021, is projected to grow to USD 980.70 million by 2030 {cite}`sphericalinsights2022blockchain`. The merger of these two technologies is an area that still calls for in-depth exploration. Moving forward, we'll examine AI in the context of blockchain, investigating in detail the possible intersections and inherent value these two technologies may possess when coalesced.
 
 ## The mutual empowerment between blockchain and AI technology
 
@@ -26,7 +26,7 @@ Conversely, AI can reciprocate by improving the architecture of blockchain syste
 
 ## Enhancing DeFi Operations with AI
 
-Most commonly, research and applications have explored the potential of using blockchain to augment AI {cite}`inproceedings`. However, our focus is on exploring how AI can enhencing Decentralized Finance. As a core application of blockchain, Decentralised Finance (DeFi) has consistently been a subject of keen interest in both academic and commercial research.
+Most commonly, research and applications have explored the potential of using blockchain to augment AI {cite}`inproceedings`. However, our focus is on exploring how AI can enhance Decentralized Finance. As a core application of blockchain, Decentralised Finance (DeFi) has consistently been a subject of keen interest in both academic and commercial research.
 
 Raheman et al. {cite}`9686345` designed an infrastructure of AI agents or "Oracles" for portfolio management, liquidity provision, and price prediction in various decentralized financial markets. As shown in {numref}`defi&ai`, these Oracles will increase investment value and returns by offering liquidity. They serve end business applications, smart contracts, and other agents. A key aspect is the distinction between an "inventory/portfolio" that comprises multiple assets and a "DEX swap pool" or "DEX balancing pool," which is one of several portfolio maintenance strategies. Therefore, a single "inventory/portfolio" may contain multiple "DEX swap pools" or "DEX balancing pools" with various strategies.
 
@@ -57,11 +57,11 @@ Key components include:
 - A Signal Generator Oracle for real-time trading and liquidity advice
 - A Sentiment Watcher Oracle for monitoring social and online media buzz about specific tokens
 
-These components are informed by the Price Predictor, which predicts price trends and volatility using AI and machine learning. Data is sourced from various channels, including centralized exchanges and live Ethereum Nodes. The architecture is currently under construction, with primary components being the Strategy Evaluator, Price Predictor, Portfolio Planner, and Pool Weighter. The research {cite}`9686345` also indicate that there are opportunities to improve the predictor model's efficiency, such as sharing models across similar markets or using pre-trained models that match expected market conditions. For more detail of the architecture and test result, please see [here](https://docs.singularitydao.ai/research-papers/artificial-intelligence/architecture-of-automated-crypto-finance-agent).
+These components are informed by the Price Predictor, which predicts price trends and volatility using AI and machine learning. Data is sourced from various channels, including centralized exchanges and live Ethereum Nodes. The architecture is currently under construction, with primary components being the Strategy Evaluator, Price Predictor, Portfolio Planner, and Pool Weighter. The research {cite}`9686345` also indicates that there are opportunities to improve the predictor model's efficiency, such as sharing models across similar markets or using pre-trained models that match expected market conditions. For more detail of the architecture and test result, please see [here](https://docs.singularitydao.ai/research-papers/artificial-intelligence/architecture-of-automated-crypto-finance-agent).
 
-In general, DeFi allows user to access financial services without intermediaries. Comparing to traditional finance, DeFi allows users to have greater control over their assets and avoid the fees associated with centralized exchanges.
+In general, DeFi allows users to access financial services without intermediaries. Compared to traditional finance, DeFi allows users to have greater control over their assets and avoid the fees associated with centralized exchanges.
 
-- With help of AI, DeFi platforms can analyze vast amounts of data to provide personalized investment and minimizes risks. Machine Learning Algorithms (MLA) can help identify investment opportunities and optimize smart contracts. These can help user to make more informed decisions to increase profits.
+- With the help of AI, DeFi platforms can analyze vast amounts of data to provide personalized investment and minimize risks. Machine Learning Algorithms (MLA) can help identify investment opportunities and optimize smart contracts. These can help users make more informed decisions to increase profits.
 
 Yield aggregator and dencentralised exchange (DEX) are specific applications under the DeFi ecosystem.
 
@@ -82,9 +82,9 @@ SingularityDAO is not solely a DEX, a liquidity pool, or a yield aggregator in t
 
 With AI technology becoming increasingly sophisticated, the line between AI-generated content and human-created content is blurring. As such, there is a growing need to ascertain the origin of content—whether it was generated by applying a specific AI model to a particular input. Zero-knowledge cryptography could provide a solution, offering a method to validate the outputs from these models without revealing any sensitive information about the input or the model itself—a process often referred to as ZKML (Zero-Knowledge Machine Learning). This capability could be extremely useful in sensitive fields like healthcare, where the confidentiality of patient data is paramount. However, it's important to note that ZKML currently focuses on creating zero-knowledge proofs for the inference (or output) of machine learning models, not the training phase, which is a highly computationally demanding task {cite}`dcbuilder2023zkml`.
 
-Battah et al. {cite}`9924181` proposed a solution using blockchain and NFTs addresses the limitations of existing methods in managing AI model ownership, trading, and access by providing traceability, transparency, auditability, security, and trustful features . It leverages blockchain technology to create a decentralized and transparent system where ownership rights and exchanges of AI models can be managed in a trustworthy manner . By using NFTs linked to AI models, smart contracts are employed to enforce ownership, ease of access, and exchange policies . This ensures that the ownership of AI models is clearly defined and can be easily transferred between parties . Additionally, the system tracks information regarding all assets and provides provenance of data, overcoming trust concerns . Overall, the solution aims to provide a secure and reliable framework for managing AI model ownership, trading, and access. This system keeps track of all assets and provides provenance of data, could potentially addressing the trust concerns that can arise with AIGC NFTs.
+Battah et al. {cite}`9924181` proposed a solution using blockchain and NFTs to address the limitations of existing methods in managing AI model ownership, trading, and access by providing traceability, transparency, auditability, security, and trustful features. It leverages blockchain technology to create a decentralized and transparent system where ownership rights and exchanges of AI models can be managed in a trustworthy manner. By using NFTs linked to AI models, smart contracts are employed to enforce ownership, ease of access, and exchange policies. This ensures that the ownership of AI models is clearly defined and can be easily transferred between parties. Additionally, the system tracks information regarding all assets and provides provenance of data, overcoming trust concerns. Overall, the solution aims to provide a secure and reliable framework for managing AI model ownership, trading, and access. This system keeps track of all assets and provides provenance of data, which could potentially address the trust concerns that can arise with AIGC NFTs.
 
-While blockchain technology does not natively recognize or comprehend real-world events, it could be advantageous if it were cognizant of such incidents. This understanding could enable the transfer of value in accordance with real-world situations. Oracles provide a solution to this by serving as intermediaries that fetch and verify real-world data for blockchains. However, they may not suffice in all cases because some real-world data requires computation before it's sent to the blockchain. For instance, a yield aggregator aiming to transfer deposits between different pools to maximize yield in a trust-minimized way would need to compute the current yields and risks of all available pools. This forms an optimization problem that machine learning is well-equipped to tackle. Nevertheless, executing machine learning computations on the blockchain is costly. This presents an opportunity for ZKML, this would allows machine learning computations to be conducted off-chain but verified on-chain in a zero-knowledge manner, which could potentially reduce costs and increase efficiency {cite}`samani2023convergence`.
+While blockchain technology does not natively recognize or comprehend real-world events, it could be advantageous if it were cognizant of such incidents. This understanding could enable the transfer of value in accordance with real-world situations. Oracles provide a solution to this by serving as intermediaries that fetch and verify real-world data for blockchains. However, they may not suffice in all cases because some real-world data requires computation before it's sent to the blockchain. For instance, a yield aggregator aiming to transfer deposits between different pools to maximize yield in a trust-minimized way would need to compute the current yields and risks of all available pools. This forms an optimization problem that machine learning is well-equipped to tackle. Nevertheless, executing machine learning computations on the blockchain is costly. This presents an opportunity for ZKML, which would allow machine learning computations to be conducted off-chain but verified on-chain in a zero-knowledge manner, which could potentially reduce costs and increase efficiency {cite}`samani2023convergence`.
 
 ## Conclusion
 
@@ -97,4 +97,4 @@ The symbiotic relationship between blockchain and artificial intelligence can pa
 
 ```{bibliography}
 :filter: docname in docnames
-```
+```