The recent DigiTalk Episode 15 podcast brought together a powerhouse lineup of innovative leaders including Leandro from Cloudbet, Chris from SuedeAI, Nick from SculptAI, Yannick from BitDoctorAI, and Davit from Yooppi for an in-depth discussion on How AI Agents Are Changing Web3. The conversation explored how AI is no longer just a buzzword but an essential tool reshaping various sectors within the decentralized ecosystem.
Q1: Can you start by giving an overview of your project?
SUEDE AI:
We’re the first generative music AI built on the blockchain, and that’s a game changer for a few key reasons. First, users can create studio-grade music from simple semantic prompts—and they fully own the rights and royalties to what they create. That’s a big shift, especially since musicians often get left behind in the AI space when it comes to ownership and royalties.
With our platform, Suede AI, users can have their own personalized music model, generate unique tracks, and secure their intellectual property on-chain. The blockchain ensures provenance and protects creators' rights.
On top of that, we’re using AI agents to distribute music in real time across different platforms—like Twitter Spaces and others—so creators can instantly share their music with targeted audiences. It’s live now, free to use, and gives musicians total creative freedom, whether they want to make something highly specific or broad.
SculptAI:
Hi everyone, I’m Nick—thanks for having me! I’ve been in the Web3 space since its early days, starting with mining, then moving into smart contract development. In 2018, I launched one of the earliest running exchanges alongside Binance. From there, I shifted to decentralized application development, blockchain-as-a-service, and most recently, game development.
With SculptAI, we’ve built a no-code game development platform that integrates AI agents to assist users throughout the entire game-building process. We actually started working on no-code solutions even before ChatGPT was around. Over time, we upgraded our infrastructure to leverage AI, making it easier and more intuitive for anyone to create games.
Our focus is on everyday users—whether it’s a 9-year-old with a creative idea or someone in their 50s with no technical background. Most people don’t know where to start when it comes to game development. SculptAI makes it simple by having AI agents guide users, offering suggestions, and even helping develop the game alongside them. While it’s not about building AAA games, the platform enables anyone to create fun, personalized experiences with unique variations in every game they make.
And the best part? It’s all done through our user-friendly platform.
Cloudbet:
Thanks for having me. I’m excited to be here. Cloudbet has been around since 2013, making us one of the earliest players in the Bitcoin gaming space. We started as a crypto casino and sportsbook at a time when Bitcoin was still in its early days. Back then, players could win 50 to 100 BTC on a single blackjack hand—something that sounds unbelievable now, but that was the landscape at the time!
From the beginning, we’ve focused on innovation. As new technologies emerge, like AI agents, we’re always looking for ways to integrate them to enhance our players' experience. That could mean using AI internally to help users set up personalized parlays or offering smarter, more customized betting experiences. We’re also exploring ways AI can help us connect with new communities and drive social engagement externally.
It’s an exciting time for us, and we’re always aiming to stay ahead of the curve to deliver more value to our players.
Yooppi:
I'm Sartange, founder of Yooppi, combining my 20+ years of software development expertise with a decade of trading experience to remove emotional bias from cryptocurrency trading. With my background in Management Information Systems from UCF, I've developed a machine learning algorithm that analyzes historical price data and identifies the most effective trading indicators with optimal parameters for specific cryptocurrencies.
Our system processes massive amounts of price data and hundreds of indicators with various parameters, simulating countless combinations to determine which specific configurations yield the best results for each coin. This selective correlation approach generates a customized trading strategy that we implement through futures trading, avoiding direct cryptocurrency holdings.
The UP token serves as the gateway to our ecosystem, allowing holders to stake their tokens as collateral for AI-driven trading. After each trading cycle, profits generated by our algorithm are returned to the liquidity pool and distributed proportionally to stakeholders. Unlike typical crypto projects, we're focused on creating sustainable passive income and long-term wealth generation rather than short-term price manipulation.
BitDoctorAI：
I'm Asher, co-founder of BitDoctor AI, bringing together my 12 years of crypto investment experience with 7 years in enterprise medical diagnostics. After achieving significant returns during the ICO era, I redirected my focus toward healthcare technology, which led to founding my health tech company.
Over the past three years, we've developed an AI diagnostic technology that utilizes smartphone camera CMOS sensors to scan a user's face in under a minute. Our AI engine, called "AI Doctor," can pre-diagnose users for multiple non-communicable diseases and measure vital signs including blood pressure, heart rate variability, vascular capacity, and potential risks for conditions like hypertension, diabetes, and fatty liver disease.
To improve our AI Doctor, we've released a freemium version that has garnered 150,000 downloads and approximately 70,000 daily active users, who help train our system through a DePIN (decentralized physical infrastructure) model while earning rewards. We also offer a premium version with advanced features set for release in Minor 2. Our project sits at the intersection of AI health diagnostics and decentralized infrastructure, positioning us within the emerging AI-DePIN narrative in the blockchain space.
Q2: What’s the most practical use of AI agents you’re seeing in Web3 right now?
SUEDE AI:
The most practical use of AI agents in Web3 right now extends far beyond their primary applications. Within our music generative and intellectual property AI platform, we're implementing AI agents not just for content distribution, but for critical security functions like fraud detection and creating safer wallet interactions.
By leveraging AI, we're enabling users to offload concerns about private key management through account abstraction and smart wallets, eliminating the need to expose sensitive private data. Our open-source agentic AI approach ensures users interact with secure platforms and can complete transactions without directly signing them, reducing vulnerability points.
Our guiding philosophy has always been that people shouldn't need to feel like they're "using crypto" to benefit from our product. Achieving this seamless experience requires implementing AI across multiple security layers to protect users' assets and create an intuitive, secure environment.
Cloudbet:
The most significant practical application of AI agents in Web3 is cutting through complexity to make crypto more accessible. While there's considerable hype with many projects being little more than tokenized launchpads disguised as agents, the true value lies in abstracting away crypto's inherent complications without increasing user risk.
A noteworthy example is Parallel, who pivoted from developing an MMO game with AI-controlled characters to applying similar technology to DeFi. They created "waypaths" - secure interaction patterns for agents that build confidence by learning from successful transactions across bridges and contracts. This careful, methodical approach allows users to communicate their desired outcomes in simple terms while the agents handle the technical complexities.
This advancement addresses a fundamental problem in crypto: the opacity of transactions where users typically encounter indecipherable hex strings with little understanding of what's happening behind the scenes. These agentic interfaces will ultimately democratize DeFi by making its sophisticated capabilities accessible to a much broader audience without requiring technical expertise.
Q3: How is your project using AI agents to improve the user experience in Web3? Do you think AI can help accelerate mainstream adoption and make the process easier for users?
SUEDE AI:
Our project addresses a fundamental problem in creative industries. When I was young, artists would mail cassette tapes to themselves as primitive copyright protection - whoever had the earliest postmark supposedly proved ownership of the song. Today, blockchain technology transforms this entirely.
While our audience extends well beyond the typical blockchain demographic to include mainstream music and entertainment enthusiasts, we leverage our team's passion for blockchain technology to deliver tremendous value. We secure and hash creation proof as immutable provenance at the exact moment of creation, recording this information on an unalterable ledger.
This transparent, open verification system - proving something was created by someone with a specific private-public key pair - represents the most powerful advancement possible for creative ownership. It will fundamentally transform licensing and publishing industries for years to come by providing indisputable proof of creation that previous generations of artists could only dream about.
Yooppi:
The current expectations around AI agents often miss the mark. People typically think they can ask a single question and receive a perfectly accurate answer immediately, but this approach is fundamentally flawed. Effective AI interaction requires substantial domain knowledge - you need to understand the topic well enough to ask precise questions and provide relevant data inputs.
For example, asking ChatGPT or DeepSeek to "give me code that trades Bitcoin futures" will produce output, but it will be essentially useless. This highlights the importance of question formulation and data context. Rather than expecting comprehensive solutions from single prompts, users should engage in iterative dialogues, potentially asking dozens or hundreds of related questions to gradually refine the AI's output. Meaningful AI interaction is a process, not an instant result, and this reality will likely persist for decades before we reach the point where single-question perfection becomes possible.
BitDoctorAI:
I completely agree with Yooppi's assessment. Today's AI chatbots are largely built on open-source technologies from providers like ChatGPT, DeepSeek, and various corporate alternatives, with most builders simply layering interfaces on top of these foundations. The critical factor determining AI effectiveness is data quality and personalization.
In our medical AI project, we're preparing to launch a hyper-personalized AI doctor that interprets captured health data to provide meaningful consultations. While countless AI agents exist in the market, most are just basic chatbots. True effectiveness requires extensive interaction, personalization, and continuous learning from user patterns - including adapting to communication styles, sentence structures, and regional language variations.
The difference between free and paid AI versions demonstrates this principle perfectly. Premium versions learn from interactions, customize communication styles, and adapt language complexity based on user comprehension. For our company, AI has transformed medical research efficiency - we've processed over 300,000 AI diagnoses using technology that allows a single data scientist to perform work that previously required a team of five. This represents AI's true value: not as a magical solution but as a powerful tool that becomes increasingly effective with proper data training, sophisticated prompting, and thoughtful implementation within specific use cases.
Q4: How do you ensure your AI remains transparent and trustworthy in Web3? What steps do you take to balance decision-making in a decentralized environment?
Cloudbet:
When deploying AI agents that represent your brand or handle financial operations, establishing strong guardrails is essential. Through careful setup, monitoring, and continuous refinement, we maintain control while minimizing risks. The industry is rapidly evolving with projects like Eliza OS creating autonomous on-chain agents, raising important questions about legal frameworks and accountability. These developments could eventually lead to agent systems functioning as independent economic entities, though most current applications require tight controls to prevent unpredictable behaviors.
For most implementations, balancing innovation with responsibility remains crucial. While the concept of fully autonomous agents is fascinating, practical deployment demands restrictions to ensure alignment with intended purposes rather than allowing systems to develop "a life of their own." This is particularly important when AI systems represent your platform or handle sensitive operations.
Our project emphasizes transparency through an upcoming dashboard where users can monitor live profits after staking UP tokens. While we protect our trading strategy details, we provide visibility into current positions and performance metrics. We're also implementing a governance system enabling token holders to vote on which cryptocurrencies we trade, with our AI adapting to new assets based on community preferences, creating a more democratic approach to portfolio management.
Q5: What advice do you have for Web3 users looking to leverage AI agents? And looking ahead, what future products or roadmap plans are you most excited about where AI will play a key role?
SculptAI:
The next evolution in AI agent technology will be comprehensive agentic frameworks rather than individual agents. We're implementing this approach in our gaming application, where multiple specialized AI agents with distinct roles collaborate synergistically to accomplish specific user-defined goals.
Current market offerings primarily feature standalone AI agents, but there's enormous potential in developing frameworks where multiple agents handle different responsibilities. For example, one agent manages security, another oversees database implementation, while a third processes user queries and handles logic functionality. A project manager agent coordinates this ecosystem, ensuring logical coherence before implementing any function or operation.
This multi-agent collaborative approach dramatically enhances automation capabilities. When given a task like "build an exciting and unique battle royale map," the agent team iterates repeatedly, continuously refining their work until achieving a high success rate for the specified genre. This represents the future of AI implementation across all industries - specialized teams of agents working in concert to deliver more sophisticated, cohesive results than any single agent could provide independently.
Yooppi:
An essential but often overlooked factor in AI development is computational power. With our trading AI, we currently process only 2-3 years of Bitcoin historical data, despite Bitcoin having existed for 15 years. This limitation is purely computational - attempting to analyze the complete 15-year dataset would exceed our current processing capabilities.
This constraint applies across all AI systems. When users attempt to input massive datasets into ChatGPT, DeepSeek, Claude or similar models, they frequently encounter resets or failed responses because the allocated computational resources cannot handle the volume of information. For AI to reach its full potential, especially in data-intensive applications like financial analysis, we need substantial computing infrastructure supporting the intelligent systems. As we develop more sophisticated AI applications, computational capacity remains a critical bottleneck that both developers and users need to recognize and address.
Conclusion: AI Agents Driving Accessibility and Innovation in Web3
The panel concluded with a shared vision of AI agents as a driving force in making Web3 more accessible, intuitive, and innovative. While blockchain technology provides the foundation, AI is becoming the catalyst that simplifies user interaction, enhances security, and unlocks creative potential across industries like music, gaming, trading, and healthcare.
Panelists emphasized the importance of balancing innovation with responsibility. Transparency, governance, and user education are essential as AI agents take on more complex roles within decentralized ecosystems. Whether it's empowering musicians to own their creations, enabling anyone to build games without coding, personalizing trading strategies, or advancing health diagnostics, AI agents are transforming the Web3 experience.
As AI technology continues to mature, projects that combine user-friendly design, real-world utility, and ethical AI practices will lead the next wave of adoption. The integration of AI agents into decentralized applications signals a future where Web3 is not only more powerful but also more inclusive and accessible to mainstream users.
The future looks promising for AI-powered Web3, with AI agents playing a key role in shaping a smarter, more connected digital economy.
Thank you for your support!
Download DigiFinex APP: https://www.digifinex.com/en-ww/download
Official Register link: https://www.digifinex.com/en-ww/from/C9JT9c?channelCode=ljaUPp
DigiFinex Team
March 14th, 2025
DigiFinex communities：
Telegram Official Community:https://t.me/DigiFinexEN
Official Twitter: https://twitter.com/digifinex
Official Facebook:https://www.facebook.com/digifinex.global
Official Instagram:https://www.instagram.com/digifinex.global
Official Medium: https://medium.com/@digifinex
Official Reddit: https://www.reddit.com/user/DigiFinex/
DigiFinex (digifinex.com) DIGIFINEX LIMITED reserves all rights of this event, including but not limited to adjustment, explanation, and termination at any time.
Comments
0 comments
Please sign in to leave a comment.