The following text is from the series Twitter Space #对话DEV, hosted by FC, founding partner of SevenX Ventures, Twitter @FC_0X 0
This episode’s guest: Marvin, founder of Phala Network, Twitter @marvin_tong
First time doing Dev: No top VC endorsement, but listed on Binance with the support of the community
After graduation, Marvin became an ordinary worker, working as a product manager in a large company. In 2018, the disruptive revolution brought about by smart contracts gave him and his brother @bgmshana, who worked on Machine Learning in the Gmail team, the idea of starting a business.
Which track to choose? Marvin and his team have a logic: the core value of Web3 is to solve the trust crisis, and in this process, new innovation models will be born, and the value of these innovations will be captured by the technology with the highest trust level.
TEE is exactly such a technology. So Phala was established, a decentralized cloud that uses TEE as hardware. After experiencing bumping into walls in the early stage, entering the Polkadot ecosystem became a turning point for the project. After a period of building, Phala became a very advanced project in the Polkadot ecosystem, and then seized the opportunity to successfully list on Huobi and Binance.
Regarding the listing of coins, Marvin believes that one of the things they did right at the time was the fairlaunch of the tokens. Most of Phala’s tokens are mined, and users do not need to run dedicated nodes. As long as there is TEE hardware, the coins can be mined.
This has allowed Phala to gain widespread support from the community and thus gain a lot of traffic. Although it does not have the endorsement of top VCs and a team with a first-class resume, the project has been listed on the blockchain thanks to its impressive data indicators in the community and user dimensions, aka the mass base, and the trend of Polkadot.
But the trade-off of Fairlaunch still exists. Because the team only has 5% of the tokens, many capital cooperations that came after the coin was listed have been scrapped.
Devs transformation path: Go to the United States and find PMF
The Defi Summer and NFT Summer from 2021 to 2022 was a very painful period for Marvin and the team. They kept trying to find some use cases, but unfortunately there were no results.
Marvin began to reflect. He had a vague feeling that in the past few years of crypto, due to some crypto rules such as prelisting and postlisting, he ignored the real focus of the project. He came to two conclusions:
1. If any Infra is to be used by someone, you need to find PMF
2. The entire team must serve only this matter before finding the PMF and not be distracted by doing anything else.
Of course, the premise of this conclusion is that the founders themselves have figured out what they want. We think that TEE still has 10,000 times of space that has not been fully explored, so we must do this, and we will admit it even if we die. This direction. If you focus on a direction, you must become the leader in it.
In order to find a breakthrough, Marvin and his team moved to the United States in 2023. One of their motivations was that 70%-80% of the projects that have real users and PMFs for infrastructure services for enterprises and developers are in the United States, so they wanted to be closer to this market, get to know these people, and understand how they do things. They talked to the projects, held hackathons, and talked to developers to find their own problems. Since last year, they have had some PMF systems.
What was it like to find PMF?
1. Someone is actually willing to pay for your stuff
2. Use it in strange ways, usually not the way you expect
Just like you sell bricks and ask people to build houses, but no one buys them. But one day, someone suddenly wants to buy your bricks to put out the fire on his head. This is probably what it feels like to find PMF. Once this scenario occurs, what you need to do is to find the person whose hair is on fire and change the product form of the brick so that it can put out the fire faster.
The key point of transformation: How are important partners onboarded?
Phala has been onboarding customers for almost 24 years. During this process, Marvin mentioned two key teams:
1. Flashbots
Flashbots was the key team that discovered that TEE must be used to solve the censorship problem under the Ethereum chain. It was also on the way to solving the TEE problem that they found Phala and had a series of technical exchanges and cooperation. Now Phala is also providing some services for Flashbots.
2. ELIZA
Marvin met Shawn in a hacker house and told him that if AI Agent runs in TEE, it can realize a fully autonomous Agent, which is the End Game of AI Agent in Crypto. Shawn bought in on this logic, and after that, some Phala developers began to make more in-depth contributions to the ELIZA ecosystem.
It was also because of this opportunity that Marvin created spore.fun in December last year.
The story behind the launch of Spore
After getting to know Shawn and having in-depth discussions, Marvin came up with the idea of creating an ELIZA+TEE case with two routes:
1. Create Unrugable AI. After running the Agent in TEE, it can eliminate artificial rugs. The product that Marvin and his team collaborated on is called AI POOL, and they provide technical support.
2. Do Autonomous AI and demonstrate complete autonomy to the extreme, and then you will have spore.
When is AI truly autonomous? When it does not need humans to support it. All AIs on the market still need people to pay the bills. Developers quietly pay Google and OpenAIs big model APIs. What Marvin has to do is to run the Agent into the TEE, so that it can pay itself and earn money to support itself. If it earns a lot, let it replicate its own genes, then evolve and compete, and finally the fittest will survive.
After Spore was launched, it was well received by everyone, and its ATH market value was probably 80 M, which made Marvin start to pay attention to this product and operate and develop it as a separate project.
“After GPT 3.5, we found that more and more AI projects were willing to try our products, or at least willing to spend time listening to our pitches. Later, as more and more AI customers onboarded, we also began to focus on serving the AI track.”
When Dev switches to a transaction perspective, how does he describe the potential of his project?
About Phala
First of all, if you believe that one day the penetration rate of Web3 can grow from todays 10% to 90%, then this process will definitely require extremely low-cost, trustless technology to be compatible with Web2 software, because it is impossible to rewrite Web2 software. As for the infrastructure track of Web3 Mass Adpotion, Phala is currently in a leading position, and if I am not being modest, it is the first.
Secondly, all problems are caused by insufficient growth. Problems such as insufficient dealer chips and insufficient capital are not worth mentioning in the face of violent growth, so we should start to pay attention to the growth space of the project.
Finally, quantitative data may not be of much reference value, including the so-called value capture and profit, etc. What matters is the application scenario, that is, whether more people are using Phalas TEE and who is using Phalas TEE.
Lets talk about spore
Now its value is brought by the team + themes, and this product is still being continuously built by a passionate team, so its growth space is reflected in:
Spore may become the most popular launchpad in the ai16z ecosystem, and this launchpad is autonomously generated by AI.
For excellent low-code agent platforms like Virtual, 95% of the AIs produced have a uniqueness or intelligence level of 60 points. The AIs generated by Spore themselves also have an intelligence level of 60-70 points. The difference is that on Virtual, if you pile up resources or have a good team, there is a 1% probability of turning an AI from a level of 60-70 points into a level of 90 points, such as AIXBT, but the remaining long-tail AIs have the same level as those on Spore.
Therefore, Spore has a great opportunity. Now many teams with ideas, or teams that have made some good products but cannot issue tokens, can achieve the following through Spore:
The AI Agent does not need any AI knowledge. It has unique skills, which can be added to the 60-70 points Agent automatically generated by AI, turning it into an 80-90 points Agent.
2. All Agents produced on Spore automatically run under a TEE ecosystem, using the ELIZA framework by default, and tokens are issued on SOLANA, which is interactive.
What dimensions should a Dev consider?
Marvin described the current Crypto market as extremely retail-oriented and extremely attention-driven.
He made an analogy with the development of short video live streaming sales. At the beginning, unknown internet celebrities and actors were broadcasting, and then more and more anchors gradually poured in. At this time, three types of people will appear: 1) Audiences, who find it interesting and indulge in it every day, giving rewards and buying things; 2) Despisers, who criticize from a moral perspective, thinking that this is eye-catching and moral decline; 3) Practitioners, who think that if the world is selling goods in the future, what do people need? What do I need to do to capture market growth?
The same is true for the current Crypto market. Even the president has issued a coin. As a practitioner, what we should think about is, if issuing coins becomes as common as air and everyone expresses their needs, then what should I become? Should I become a seller? A supplier to sellers? A tool in the process? Or a short video platform itself?
If there were no internet celebrity system, it would not be easy to become a seller.
If you want to become a supplier, it is best to find an active market where the products are not widely circulated/scarce, and then sell them through bound channels. In fact, many teams are suitable for doing this, and there are many people doing this.
Providing tools, becoming a platform or part of a platform is a process of finding PMF, and the success rate is relatively small. But like GMGN, it grasps both the demand and the trend. It may be a product that can be produced by an extremely practice-driven team and way of doing things.
In fact, for Dev, under the premise that it is impossible to predict the market in the long term, there is no need to think about what will happen in the market two years later. What you need to think clearly is what you want and what you can do.
What are the Stop Doing Lists for being a Dev?
“Don’t copy others. I think the thing I regret most is not thinking from a first-principle perspective to build my own uniqueness.
The core logic of business is that you can provide value that others cannot. All actions, products, and services actually come down to this point.
Once you get lazy about deep thinking and first principles, the price you pay is that although you have done a lot of things, they may have some effect, but it is a waste of time and resources because it may not be what you really want.
Regardless of whether he has experience in the crypto industry or not, he should quickly distinguish which are the rules of the game to master when entering the industry, and which are the ones that he thinks are valuable based on his first thoughts. Even if he is wrong, he will be faster than others in iterating through negative and positive feedback. If he does not realize this at the very beginning, he may need to take many detours.
Crypto AI Trading Opportunities
If Marvin brings in a trading perspective, his way of observing the market is not to look at mindshare, because the object of that is people. He should focus on the content itself and look at the most popular posts related to AI topics in Kaito. Read the top 20 posts every day, and you will find a sense of the market just like looking at K-lines. The top posts with the most likes and the most forwarded by bigwigs are talking about something about AI, and you will slowly get the trend.
Regarding the new directions of AI, Marvin himself focuses on two:
1. AiFi
AiFi is not simply Ai+DeFi. It is not AiFi just because an Agent can swap or a DeFi can adjust AAVE. Marvin believes that this thing has practical value, but lacks imagination. AiFi is most likely an Agent that can have some economic exchanges.
2. Swarm concept
That is, how can the interaction between multiple AI agents lead to interesting scenarios, rather than simply returning search results? So how do we find them? Just like what I just said, look at the posts that have the most likes and forwarded mindshares every day. Don’t look at the people, just look at the posts and you will gradually get the feeling.
Finally, there are three of Marvins bells:
1. @SH L0 MS and @ropirito from @NousResearch team
Recommended reason: Builder is strong, AI development capabilities and cultural accuracy are very good, and the mindshare of Shape is very high.
Recommended reason: True Builder is one of the most powerful AIs. It is not only powerful in technology, but also its product is very close to the market and can shape what everyone wants.
FC Last words
From the perspective of results, Marvin did several things right as a Dev:
1. Choose the right track with demand
2. Know where the leaders of the game are and stay with them
3. Because I chose the right track, I encountered a new opportunity and met the key person in this opportunity
4. Know the rules of crypto, how to play in the second level, how to play in the first level, where are the resources, and how to leverage them
5. Luck and persistence
I hope this can bring some inspiration to everyone.
Thanks again to Marvin for participating in the conversation with Dev. Space audio will be updated in the small universe, and we will open a separate podcast column. If you have any suggestions for the new series, you are also welcome to leave a message to interact.
