Original｜Odaily Planet Daily (@OdailyChina)
Author: Wenser (@wenser 2010 )
Mass Adoption has always been a hot topic in the cryptocurrency industry. In this new wave of AI led by OpenAI, AI x Crypto has become an alternative solution for the mass adoption of cryptocurrency. As one of the projects in Binance Labs' Season 6 MVB Accelerator Program, Sparkle recently announced that it will shift from AI astrology applications to "Disney in the AI field", aiming to create a new AI interactive entertainment platform through the encryption paradigm.
Odaily Planet conducted an in-depth online interview with Harry, the founder of Sparkle, to understand the different strategies of front-line entrepreneurs in the AI x Crypto track for readers to learn and refer to.
Odaily Planet Daily: 1. Can you please give a brief self-introduction and talk about the origin of the Sparkle project?
Born in a traditional Internet company, he embarked on the "AI metaphysical psychology entrepreneurial journey" by chance, and believes in "creating value and making impact".
Harry, founder of Sparkle: I come from the Internet industry, and have been responsible for growth, creators, innovative business, and data science in Internet technology companies such as Meta (Facebook) and Tencent. In early 2023, I officially started my business and founded Sparkle. Later, the project was selected for the sixth phase of the Binance Labs accelerator program and received investment from Folius Ventures, a well-known venture capital institution in the Asia-Pacific region. I have a wide range of personal interests and hobbies, and I like snowboarding, ACG, and 3D models. I am also an independent developer. I have independently developed some small projects since my student days. This is both a personal interest and a way to keep myself sensitive to the latest technology and industry progress.
In the psychologically high-pressure environment of the COVID-19 pandemic that has not yet ended at the end of 2022, I noticed that many people around me sought help from fortune-telling consultations, and there were many jokes on the Internet like "psychological counseling that is more suitable for Chinese babies." This "metaphysical fever" phenomenon made me realize that it can not only help people with psychological counseling and emotional adjustment, but also has certain commercial value. I personally also hope to do something "interesting and valuable", so adhering to the core concept of "Create Value and Make Impact" (which is also related to my previous work experience at Tencent and Meta), we began to cut into the product from the perspective of "AI Astrology", and Sparkle was born naturally.
Odaily Planet Daily: 2. What is the user feedback on the AI astrology product? What is the reason that drives the project towards the vision of "AI Disney"?
There were some minor setbacks, but user feedback far exceeded expectations; based on the core concept, we are moving towards the bigger vision of "AI interactive entertainment platform".
Harry, founder of Sparkle: In fact, the initial product stage was much more difficult than we imagined, especially in terms of compliance with the application listing (such as restrictions on Apple's App Store). But after the full version of the APP was launched in May this year, user feedback data far exceeded expectations - our next-day retention, 3-day retention, 7-day retention and other indicators are already on par with the industry's top competitors. Commercial performance such as product payment conversion rate also gave us strong confidence, and product market fit (PMF) has been initially achieved. Subsequent development focuses mainly include user growth, brand awareness building and usage scenario expansion.
Sparkle’s vision upgrade is also a decision made based on our core concept of “Create Value and Make Impact”. AI astrology is only our first product. With the second AI product ready to go, it is only natural to upgrade Sparkle from an “AI astrology application” to an “AI interactive entertainment platform”. Analogous to the development process of ByteDance, Toutiao used to be its main product, but after the emergence of TikTok, ByteDance gradually developed into a platform brand, and the same is true for Sparkle.
Odaily Planet Daily: 3. How do you view the combination of “AI x Web3”? What is the most important driving factor in the development of the Sparkle project?
AI x Web3 empower each other and solve problems together. Sparkle has chosen to combine the two to solve more scenario needs. Mindset is the top priority of the project development.
Harry, founder of Sparkle: This topic is very broad. From a personal perspective, there are only three ways to combine "AI x Web3":
First, AI empowers the Web3 industry and solves its inherent problems. For example, 0x Scope, an AI on-chain data analysis platform;
Second, Web3 empowers AI to solve the challenges it faces, such as DePIN projects such as Render Network and io.net.
The third is to leverage the dual advantages of AI and Web3 to solve diverse problems in various industries and life scenarios, improve productivity at work, or create a more interactive entertainment experience.
These three directions have their own advantages and disadvantages, so the main consideration becomes:
Market size: How big is the market behind the problem?
Market maturity: Is there sufficient market acceptance?
Difficulty of problem solving: How to seek consensus and recognition on a larger scale?
Different from the first two directions, Sparkle chose the third direction. This decision is based on our comprehensive assessment of market size, market maturity, community consensus and recognition. We believe that real needs and feasible application scenarios are the most important. Therefore, we focus on promoting projects under a mature market demand, rather than choosing directions that are difficult to implement or have only concepts. Moreover, the fields of pan-entertainment and pan-life have huge potential and opportunities for continuous innovation.
As for the key driving factors of the project, I think the core is the team's mindset. Take the user feedback function of our product as an example. From a traditional perspective, this requires a lot of operational resources to maintain. But from the perspective of AI practitioners, this only requires a simple "Chatbot + knowledge base" function. If we add the perspective of Web3, the knowledge base reserve and user feedback classification can be solved through community incentives. So, when we look at the same problem from a new perspective, we have a more efficient solution than before.
In addition, compared with projects in other fields, AI x Web3 projects need to have monetization capabilities on the first day of their launch. Because the large models required for AI projects to achieve reasoning capabilities (whether Claude, ChatGPT or privately deployed Llama) all involve Token (Odaily Note: a unit of measurement for API quota consumption in the AI field) consumption (or graphics card computing power consumption). Therefore, every Token consumption needs to be recovered through commercialization, which is the biggest difference between AI projects and traditional non-AI projects. Just like in the Web3 field, the essence of cryptocurrency mining is to use computing power to obtain Tokens, and there is also a corresponding shutdown price, which is the standard line for revenue to cover costs.
Wenser from Odaily Planet Daily: This is a very good point. Simply put, the tokens in the Web3 field are mined through mining (strictly speaking, it is the process of verifying and adding transaction records to the blockchain by solving complex mathematical problems); in the AI field, the token is actually a unit of measurement and a consumption cost. This difference is very intuitive.
Harry, founder of Sparkle: Yes, so when the token of the AI x Web3 project is issued, it actually implies a value anchor, which is the computing power consumption. The token consumption cost price during the project development process is the value anchor point. Both user cost and project value fluctuate around the value anchor point.
Odaily Planet Daily: 4. What is Sparkle’s current development pace? How to balance “product development vs market expectations”?
The project is progressing smoothly, and the second new AI product is about to be released; Sparkle chose the latter two among time, quality and cost in order to gain long-term development advantages.
Harry, founder of Sparkle: At present, we have basically achieved the goals set in the Roadmap: launch products, find product-market fit (PMF), and commercialize to achieve organic revenue. The focus of the second half of the year is on user growth, that is, to expand the user base through brand marketing and market promotion, while reducing customer acquisition costs and increasing gross profit margins. In addition, Sparkle’s second gamified AI DApp Prompt.Inc is also about to be released, and will soon be officially launched in public beta to meet community members. The successful launch and profitability of Starry Book also proves that we can create far-reaching meaning and value by applying AI to life scenarios. Therefore, we think it is time to attract more people to join through Prompt.Inc, scale up and create valuable AI applications. Under such an overall plan, we are getting closer and closer to the time point of TGE.
Prompt.Inc Sneakpeek
Regarding the debate on "product development vs market expectations", I think trade-offs at different stages are inevitable. Take us as an example. In the early days, in order to avoid the distorted data interference caused by the "fake party", we took cold treatment measures on some "product locusts" among Starry Book users. Later, with the iteration and update of the product, it has also been highly recognized by many real users. Therefore, it is a better way to develop to meet market expectations through practical actions, rather than over-raising expectations and creating a false hype situation. This also has a positive impact on the long-term development of the project and the industry reputation. After the project TGE, the best way to meet market expectations is "sustainable business development". At this time, the health of the project itself, market value management and a more transparent governance mechanism will become particularly important. Closer communication and a more transparent governance mechanism can not only enhance community cohesion, but also better manage community expectations, thereby promoting product and project development.
Odaily Planet Daily: 5. Compared with competitors in the same field, what are the advantages of Sparkle?
The three elements of time, quality and cost are key.
Sparkle founder Harry: In terms of comparing product advantages, I think we first need to review the current status of the AI track. Since the release of ChatGPT 3.5 in late 2022, the AI track has changed significantly. In the first half of 2023, the release of Agent and subsequent GPTs opened up the possibility of more scenarios, and the championship of large models is also changing hands in turn. At the same time, the rapid development of multi-modal tools such as Midjourney, Runway, Stable Diffusion, and Sora has caused the AI industry to undergo a large number of changes and be constantly reconstructed in the past year or so. It was not until the first half of this year that the underlying AI products gradually entered a state of convergence, providing opportunities for the development of the middleware layer. Frameworks such as Langchain and RAG are still in their early stages, and everyone is exploring the optimal solution.
In this context, startups may face huge silent costs for every choice they make and need to consider three key factors: time, quality and cost.
Take Sentient, which recently raised $85 million, for example. By entering the market at this point in time, this project may have made certain sacrifices in terms of cost and achieved rapid and high-quality development through a large amount of resource investment. When AI fever just emerged, many shell applications gained considerable income by quickly developing short-term dividends. However, these low-value-added applications have gradually disappeared because their model determines that their life cycle is limited.
Back to Sparkle, we chose to spend time and patience to polish the product matrix and underlying capabilities at a relatively low cost. From our perspective, the AI field has been overheated and unstable in the past year. Chasing hot spots and traffic may lead to huge waste of resources, which will eventually be reflected in overvaluation and damage the interests of market investors. As "high FDV and low circulation" projects are increasingly unrecognized by the market, assets with self-profit, reasonable valuation and rising potential will become a trend. This can also be seen from the recent actions of the industry's leading exchanges.
Based on these understandings, Sparkle has a greater competitive advantage in market opportunities, product quality, business health and strategic planning. After all, we are committed to long-term value creation and sustainable business growth.
Odaily Planet Daily: 6. As an investor of Sparkle, what support does Folius Ventures provide to the project? And in the latest market cycle, as an entrepreneur, how do you view "high-quality assets and asset pricing"?
Received great support in terms of resources and cognition; high-quality assets and pricing depend on different market roles, and Sparkle is ready.
Harry, founder of Sparkle: Foius Ventures is a distinctive investment institution that focuses on the Asia-Pacific market and has a clear investment theme. The layout in the Web3 application layer track has formed a huge ecosystem. At the same time, Foius is also an investment institution with a long-term vision, which makes me believe that the next killer application of Web3 will come from our invested projects. They also have layouts in the direction of AI x Web3, including leading projects such as MyShell. As a well-known investor in the circle, founder Jason also provides all-round support for Portfolio. Many projects that have switched from Web2 to Web3 have relatively limited understanding of crypto asset construction and community building. Jason has provided a lot of help around the different issues of these projects to help founders improve their cognition and strategic heights. He also contributed a lot to overseas resources.
Regarding the topic of high-quality assets and pricing, just like doing business, the most important thing is to clarify who the target audience of the asset is - users, institutions, exchanges or market makers? Ideally, a high-quality asset should bring value to all parties involved, although achieving this is challenging. Based on the target audience, we need to further clarify two decisive factors: token distribution and expected holding period.
In the past, the interests of retail investors and VC investment institutions were largely consistent, and both parties hoped to achieve higher investment returns by increasing FDV (Fully Diluted Valuation). However, from the perspective of the investment return cycle, retail investors usually tend to make profits as soon as possible; while institutions face a longer unlocking period, so they pay more attention to obtaining better chip prices in the primary market and realizing the cashing out of chips within the appropriate time period.
In the current macro environment of tight liquidity, only a small number of retail investors can make profits; and VCs also find it difficult to find sufficient liquidity to take over after FDV rises. Therefore, the market gradually no longer recognizes this model, and Meme coin has gained more attention and liquidity.
From the perspective of the "strongest players in the ecological niche" - exchanges, they always follow the logic of screening high-quality projects and high-quality assets, that is, "If Meme coins are strong, then list Meme coins; if VC coins are strong, then list VC coins." Based on their judgment and understanding of the market and the entire industry, the top exchanges currently prefer projects that create positive external income and bring fresh blood to the industry. The consideration of listing coins also gives greater weight to the project parties and the community, which has led to the slight weakness of VC coins in this cycle.
According to Yijie (Binance co-founder He Yi), Binance is currently more inclined to look for projects with small and medium market capitalizations, allocate more chips to the community, and be able to run a sustainable growth PMF model to achieve self-sufficiency. We believe that the high-quality assets emerging in the new cycle will be a business target like Discord from 2017 to 2018. By solving the voice call needs of gamers, Discord has created impressive business value while growing its revenue scale, and has not overly pursued short-term hot narratives such as the metaverse. This case also illustrates the importance of entrepreneurs focusing on solving specific needs and creating real value. High-quality value targets are inseparable from the solution of real needs, which is also the concept upheld by the Sparkle team.
Odaily Planet Daily: 7. In the AI wave, which one is more optimistic about UGC or AIGC? What role does Sparkle play in it? From the perspective of entrepreneurs, which AI x Web3 projects can "survive" and even occupy their own ecological niche?
Ideas are priceless, and AI is a tool to improve efficiency in the concept development process; hematopoietic capabilities and a multi-product matrix are more conducive to occupying an ecological niche.
Harry, founder of Sparkle: Regarding the comparison between UGC and AIGC, the situation is unclear in the short term, but in the long run, I think UGC will still be the mainstream. The essence of all content creation is to express creativity and opinions, and AI is just a tool to improve efficiency. Because for high-quality content, the most important thing is your Idea and Concept. All content is essentially a process of Concept Development, and AI is just a tool to improve efficiency. From the perspective of the entire industry, AI applications are still in their early stages. For people who have not been exposed to AI tools, the threshold for use is relatively high, whether it is effective questioning or prompts. Sparkle's goal is to bridge the reality gap, which is also the value mission of Prompt.Inc.
As for what kind of projects can "survive", from the perspective of entrepreneurs, the key to whether a project can survive lies in its commercial hematopoietic ability. From the perspective of the industry's end game and ecological niche, I personally believe that it is difficult for the AI wave to produce giant companies in the mobile Internet era. AI consumer business still needs to rely on user volume and scaled network effects to build barriers. Existing giants can quickly replicate new functions or products through existing traffic and network effects. Therefore, in the future, there may be more companies with annual revenues of tens of millions, hundreds of millions, or billions, and it will be more difficult for giants to emerge. For AI entrepreneurs, building competitive barriers through a multi-product matrix is a more practical development path.
This is why Sparkle has been positioned as a multi-product matrix project from the beginning. In the future, Sparkle will become a holding company rather than a single project business entity. Through a diversified and practical business model, it can maintain long-term competitiveness in the ever-changing market environment.
Odaily Planet Daily: 8. What influence do you think the personal style and interests of entrepreneurs will have on the entrepreneurial process? Please share your views with us.
Having a wide range of interests and hobbies, and an ISTJ personality, is very helpful for starting a business.
Harry, founder of Sparkle: Of course. As mentioned at the beginning, I have many hobbies, especially skiing and model making, which require great patience and concentration, so I have a strong focus on achieving long-term goals, and I enjoy the sense of accomplishment brought by delayed gratification. As a developer and creator, I often need to invest a lot of energy in details, and these hobbies also help me become a detail-oriented person. From the perspective of MBTI, I am a very typical ISTJ personality.
Odaily Planet Daily Wenser: I understand that hobbies and personality types are indeed a plus factor for entrepreneurship. Among the existing business leaders, Amazon founder Jeff Bezos and Apple CEO Tim Cook are also ISTJ personalities. Thanks again to Harry for the wonderful sharing, and look forward to more updates from Sparkle.