"Beyond the Hype: The Convergence of AI & Blockchain" BitMart Off-the-Chain AMA #5 Recap
AI and blockchain have the potential to revolutionize various industries. Microsoft, Google, and Meta are investing heavily in AI, while blockchain is already facilitating secure data sharing in finance, identity, digital ownership, and supply chain. The convergence of these technologies can create new use cases and applications. In our last AMA, five AI companies in the blockchain industry shared their insights on the topic.
The AMA has been made more concise and refined for clarity, and you can find a link to the audio recording at the end of this article.
Synesis One
With Joshua Hong (Co-Founder of Synesis One)
Kelechi Ibe: How does crowdsourcing of data for AI become a competitive edge for companies?
Joshua Hong:
Big tech companies rely on crowdsourcing to clean their data and ensure accurate results from their AI systems. Amazon’s Mechanical Turk, OpenAI’s Scale AI, and Microsoft’s clickworker.com are examples of such services. However, recruiting high-quality workers and training them to generate better results presents a challenge. Our system operates on a Web3 environment, enabling anyone worldwide with a Phantom wallet to start training AI and earning payments instantly. We offer a decentralized platform that employs gamification technologies for ease of use and provides a competitive solution for AI companies looking to crowdsource data, particularly for conversational AI.
Kelechi Ibe: Why is the lack of high quality data sets a problem for AI systems?
Joshua Hong:
AI companies spend millions of dollars to ensure content moderation for unacceptable content, such as child pornography. They outsource this work to companies in Silicon Valley that hire thousands of workers in Africa to tag images based on specific criteria. This situation, where human judgment is required, is where AI companies must spend money to comply with societal acceptability criteria. The Web3 environment can provide a unique, highly secure, decentralized, and gamified solution for crowd-sourcing data for conversation AI.
Kelechi Ibe: Can you tell us a little bit about what Symbolic AI is?
Joshua Hong:
Symbolic AI uses logical reasoning and linguistic patterns directly encoded into data sets with a programming language called LISP (List Processing). It’s deterministic and requires less computing power, making it faster and more precise than modern machine-learning techniques. Generative AI like ChatGPT uses pattern recognition to generate answers with varying degrees of certainty. The future of AI involves merging these two systems to create a hybrid approach that combines the conversational abilities of generative AI with the precise, authoritative answers of symbolic AI based on specific data sets. This approach can offer useful services by leveraging the strengths of both approaches to benefit humanity.
Kelechi Ibe: Can you give us what role gamification plays in Synesis One’s approach to data ownership and monetization?
Joshua Hong:
Synesis One’s purpose is to provide Mind AI with quality data sets. To achieve this, the company created a blockchain-based platform that crowdsources data sets from builders, validators, and architects. The platform runs on the Solana blockchain and uses SNS tokens to pay workers. The system has been functional for nine months and underwent six months of open beta testing. The Kanon NFT functions as a membership card to participate in data entry work and make money. Additionally, the GameFi system encourages competition and collaboration among the different roles, resulting in high-quality data sets for AI companies.
Fetch.ai
With Humayun Sheikh (Founder and CEO of Fetch.ai)
Kelechi Ibe: Can you give us an overview of what Fetch.ai is?
Humayun Sheikh:
Our approach is to use an agent-based system that breaks down complex systems into agents that interact with each other to perform tasks. These agents are like microservices and can be autonomous or deterministic, depending on the task. We provide tools that combine predictive machine learning with rule-based systems to enable agents to perform actions. This changes the search and discovery model, enabling agents to look for each other and perform tasks without relying on aggregators. The goal is to simplify processes by allowing users to ask their agents to perform tasks for them.
Kelechi Ibe: What are the specific AI elements that Fetch.ai leverages to perform its job?
Humayun Sheikh:
Our system comprises three main components. The first component consists of autonomous economic agents that use machine learning and AI to perform transactions and integrations. These agents act as the front end and connect with other agents through an agent framework using symbolic reasoning, similar to a microservices architecture in Web3 or Web 2.5. The second component involves collaborative machine learning models that train the agents and provide input into their decision-making by inviting IoT devices and multiple stakeholders to deliver predictions. The third component is the search and discovery engine, which connects agents and creates marketplaces for users to offer and find services. The Fetch chain orchestrates the entire system, recording all transactions and enabling economic value transfer using blockchain technology.
Kelechi Ibe: What’s the main problem that Fetch.ai is trying to solve and what technology do you leverage to address it?
Humayun Sheikh:
Our focus is on agent-based systems, but we recognize that there are multiple components to this approach. Currently, the search and discovery model is broken, particularly for the gig economy and service-based economy, as advertising costs have skyrocketed. We aim to make search more efficient and cost-effective by integrating small agent-based microservices that can be easily discovered and connected with just a few lines of code. We also aim to solve the automation problem by providing tools for automation, allowing individuals and small businesses to cut out unimportant tasks and automate processes easily. Our goal is to bring automation to everyone, not just corporations.
Kelechi Ibe: Can you give us some highlights of a few items from the 2023 roadmap for the audience to look forward to?
Humayun Sheikh:
Our main focus is to integrate all the components we have been building and showcase the technology’s versatility. Our standalone system, Co-Learn, enables machine learning model training and provides insights into the agent-based system. Our aim is to demonstrate how our search and discovery tools, agent deployment, and automation can be used in various verticals, including DeFi, gig economy, and travel. This technology is not limited to any specific vertical and is a game changer for all. Our roadmap for this year is to onboard developers to build on our technology and deploy solutions to showcase its potential. Our key objective in 2023 is to demonstrate how to use and build on our technology.
Aspecta
With Jack He (Founder of Aspecta)
Kelechi Ibe: Can you tell us what specific role AI plays in Aspecta?
Jack He:
Aspecta’s AI technology analyzes code contributions on linked accounts such as GitHub to determine skills and proficiency through text and network analysis. It identifies collaborators and their seniority through network analysis as well. The dynamic AI layer helps Aspecta ID holders demonstrate their achievements and proficiency, providing proof for potential employers. Additionally, the technology can be applied to Twitter, StackOverflow, and on-chain activities.
Kelechi Ibe: How can Aspecta help layer 1 and layer 2 to protocols and DAOs to streamline their operation?
Jack He:
In collaboration with various blockchain ecosystems, Aspecta has been using its ID system to register developers for hackathons, providing a better understanding of their background and skill set. Achievements from these hackathons are documented on their Aspecta ID, providing verifiable credentials to demonstrate their proficiency. Aspecta is also collaborating with different ecosystems and applications to provide educational content that can be documented on the Aspecta ID as verifiable credentials. The collaboration aims to onboard more developers and build a dynamic and exclusive developer ecosystem.
Kelechi Ibe: How are Soulbound tokens relevant to Aspecta ID?
Jack He:
Aspecta ID Pack provides an AI-generated personal web page and a Soulbound token that can be put on-chain. The token can include data and content generated by Aspecta. Additionally, Aspecta is exploring the creation of metaverse twins for Aspecta holders.
Kelechi Ibe: How do you help DAOs?
Jack He:
Aspecta has a close relationship with Builder DAO, which encourages builders to use Aspecta ID to better match developers with similar skills and interests so they can build together. It also helps verify members of the DAO and serves as a management tool to create a more efficient calibration system. Aspecta ID is not just for developers, but for everyone in Web3 and Web2. Even if you’re not a developer, you can follow Aspecta ID and engage with different courses and content related to different protocols, layer 1, and layer 2 solutions from a developer’s perspective. Stay tuned for more updates.
DeepBrain Chain
Feng He (Founder of DeepBrain Chain)
Kelechi Ibe: What DeepBrain Chain is to our audience?
Feng He:
DeepBrain Chain is a computing network built on blockchain technology that provides high-performance computing infrastructure for AI and the metaverse. It consists of a high-performance computing network, a blockchain mainnet, and a GPU computing mainnet. The mainnet is developed based on Polkadot’s Substrate and has achieved large-scale implementation, being widely used in various scenarios such as AI intelligence, cloud gaming, and cloud intellect interface. More than 50 manufacturers worldwide have deployed high-performance GPU cloud platforms based on the DeepBrain Chain network, serving hundreds of enterprises and tens of thousands of AI developer groups. Products based on DeepBrain Chain include high-power IO for AI training and reasoning, deep link for cloud gaming, and towering in the cloud for mining.
Kelechi Ibe: What’s the final goal for the DeepBrain Chain?
Feng He:
DeepBrain Chain aims to build a scalable and stable high-performance computing network for AI and metaverse, becoming the largest GPU computing power network with more than one million GPUs. However, the feasibility of this goal is uncertain.
Kelechi Ibe: What should our audience look forward to in your 2023 roadmap?
Feng He:
In 2023, the company has several highlights, which include supporting the deployment of ChatGPT, NIC, and nudge models, supporting short-term PADs event models of GPU servers, launching an AIGC development contest for global developers to create an AI computing power ecosystem, and developing and announcing the cloud-based system.
Gny.io
Zachary Barnett (Head of Operations at gny.io)
Kelechi Ibe: How is GNY innovating with the blockchain and AI industry technology?
Zachary Barnett:
The mainstream attention given to AI and blockchain by companies like OpenAI has driven innovation and investment. GNY, an established blockchain product founded by experts in AI and blockchain, is exploring the unique opportunities created by combining these two technologies. GNY’s focus has been on democratizing AI tools and making them available to everyday traders to enhance their cryptocurrency trading. The GNY range report, which launched last year, provides state-of-the-art AI and ML forecasting for Bitcoin and Ethereum and will expand to cover the top 10 tokens by spring. GNY aims to continue expanding collaborative AI tools and developing functionalities that can be applied across various industries.
Kelechi Ibe: How does GNY make AI accessible to everyone?
Zachary Barnett:
Our pragmatic strategy is to start with trading information and build infrastructure for sharing datasets and algorithms. We aim to expand our focus to other data-intensive areas once we perfect our core technologies. Our goal is to develop scalable platform technologies using trading information as a pathway. We have aspirations to broaden our focus as we continue to develop these technologies.
Kelechi Ibe: Can you share more about projects that you currently have in the pipeline for GNY?
Zachary Barnett:
We are expanding our forecasting and indicators to cover the top 11 tokens by spring. Additionally, we are introducing a gamified wisdom of the crowd tool and virtual trading games. Wisdom of the crowd is a scientific area that we are exploring, where groups of relatively uninformed individuals can often outperform experts. We are also looking to expand the availability of games within our system and use our main net to power specific algorithms and training tools.
Kelechi Ibe: Can you share more about the wisdom of the crowd?
Zachary Barnett:
Our research has shown that the wisdom of the crowd data is a more valuable source than current sentiment analysis technologies that scrape Twitter conversations around BTC. We’re interested in harnessing this data for our algorithms and forecasting and making it available for companies to use as a data point within their overall ecosystem for building powerful algorithms and ML networks. We’ve spent the last few years building our core technologies such as the mainnet, data warehouse, and mobile app. We’re excited to start connecting these pieces and growing tools and experiences for users to use live.
Kelechi Ibe: Where do you see the most exciting technological innovation emerging in the AI and blockchain space?
Zachary Barnett:
GNY is excited about the potential of smart contracts and powerful ML working together to create trustless systems where data can be shared for specific purposes and individuals can be remunerated for its use. GNY hopes to enable companies to collaborate more and share data in an unprecedented way by building tools that allow for trustless ML collaboration.
>> Listen to the full AMA here <<
Don’t miss out on the next Off-The-Chain Twitter Space AMA by BitMart! Follow us on Twitter, Instagram, and LinkedIn to be notified of future episodes.
Disclaimer: The content shared in this article reflects the views of the respective speakers and not the view of BitMart. Nothing shared in this article should be considered as financial or legal advice. The use of BitMart services is entirely at your own risk. The value of digital currencies can go up or down and there can be substantial risk in buying, selling, holding, or trading digital currencies. You should carefully consider whether trading or holding digital currencies is suitable for you based on your personal investment objectives, financial circumstances, and risk tolerance.
