The following text is compiled from the series Twitter Space #DialogueTraders, hosted by FC, founding partner of SevenX Ventures, Twitter @FC_0X 0
This episode’s guest: Timo, AI investor, Twitter @timotimo 007
If you want Alpha, play something you can understand.
In Web2, Timo is a VC that specializes in investing in the primary market AI track and has invested in many domestic AI companies working in different directions.
As for Web3, his trading journey started with buying Bitcoin in 2017. In 2019, he played with contract rollovers, with the highest floating profit of tens of millions, but lost most of it in several liquidations. He started playing with the chain at the beginning of last year, and tried inscriptions, currency conversion, and pre-sales, and then the zoo and AI.
“Looking at AI targets is completely different from looking at the coins in the zoo before. You still have to play with things that you can understand. If you just follow others and rush in blindly, then you can only become someone else’s liquidity in this market.”
Timos strategy for trading AI targets on the chain can be summarized as playing the second stage, that is, looking for Alpha with a potential return of more than ten times and a relatively high upper limit, while doing a good job of position management, and investing no more than 15% in a single coin. So far, the winning rate of this strategy is over 90%.
In addition to looking for Alpha on the chain, Timos large Beta positions are all placed on exchanges, including BTC, ETH, SOL, DOGE, PEPE and AAVE.
What to buy: Which AI targets are worth paying attention to?
In general, Web3 AI is actually following Web2 AI, from the application of technology, to narrative, to the level of talent, so the trend of Web3 AI can basically refer to the development of Web2 AI.
Specifically, from the perspective of the industrial chain, there are problems that need to be solved in every link of Web3 AI, and where there are problems there are new opportunities.
1. Protocol layer
When the intelligence of the agent itself reaches a certain level, agents should be able to interact with each other intelligently and across ecosystems/platforms, and this interaction requires an underlying protocol. Currently, no one is actually working on this link.
2. Model layer
Currently, many models are general-purpose large models that know a little bit about everything, but not too much. So if you want the model to be better and more professional in a field, you need to train the model with proprietary data. There have been some models specifically for solana on the market before, and there is a high probability that more proprietary models will appear in the future. There are many opportunities here. In addition, the inference cost of the model is also a link that can be improved.
3. Data Layer
Just like people need to find high-quality learning materials, models also need high-quality data. However, most of the data is invalid and needs to be screened, cleaned, and labeled. Scale AI, a Web2 company specializing in data cleaning, has a valuation of more than $1 billion.
4. Tool layer
If we compare the model to the human brain, it needs tools to complete some tasks. Currently, many projects that want to be agents face two problems in terms of tools: first, whether they have the ability to intervene; second, whether the tools are willing to let you intervene. Therefore, Timo believes that if an open platform can be built, everyone can use this platform to call various tools, which will be very potential.
5. Application layer
In Timos definition, the application layer is actually a variety of AI Agents. In Web3, what we need to focus on is: applications that can attract the attention of Web3 users and make them use them frequently, which is why many projects are telling the story of AI+DeFi. But now it is relatively basic, staying at the fundamental and technical analysis of tokens, and cannot automatically execute the desired trading strategy, so there are many opportunities here, but we need high-quality on-chain data to train the model.
How to buy: How to play the second stage?
The gameplay of the second stage can be summarized into two steps: first analyze the fundamentals, then analyze the market. Fundamental analysis is to determine whether a target can be bought, while market analysis is to determine at which position to buy.
Fundamental analysis can be carried out from four aspects:
The first one, narrative, is essentially the positioning of the product.
Narratives determine the upper limit of a project. A good narrative can attract enough funds in the market to FOMO. Timo suggests that you can focus on three types of narratives: 1) those that users may use more 2) those that can cut liquidity 3) those related to infrastructure
The second one is chip control, which is whether there is a strong banker.
Why should we play projects with high degree of fund control (strong banker’s market)? In the previous conversation with Mr. Mai, he explained in detail. In summary, strong banker’s market has a relatively longer life cycle and greater imagination space, and it is a win-win game for bankers and retail investors. It is also very simple to judge whether it is a strong banker’s market. Go to GMGN to look at the position wallets. If the wallets in the front row are basically small fish, bought in early, and then made a lot of profits, they are basically all banker’s market.
Third, products and technology
As the market matures, pure stories are no longer bought in, and attention needs to be paid to products and technology. For the analysis of these two aspects, Timo suggests that if you don’t understand them, you can form a team with your friends, with some people looking at the code and others responsible for “feeling” the chain. It is extremely difficult for one person to do all the analysis.
Fourth, the background of the team.
Timo believes that if Open AIs DEV can score 100 points, the average level of AI DEV in Web3 may only be 30 points. However, this industry is gradually attracting more outstanding talents, so we should try to find regular troops and fully verify their resumes and scientific research results.
In fact, direct dialogue with the DEV team and obtaining first-hand information will be of great help in judging the fundamentals of the project. Many DEVs are willing to have this kind of communication.
As for the position of opening a position, according to Timo’s experience, there is no need to rush for the second stage of many projects, because there are at least two bottoms, usually when the market maker takes over the funds, and it is generally more reasonable to open a position at this time. The operation of the market maker can be known by monitoring and tracking the market maker’s wallet.
How can you not make mistakes when playing the second stage?
There are two aspects to note.
1. Research must be done sufficiently
There is a lot of time to analyze the project in the second stage, so dont be lazy. In addition, Timo suggested to collect first-hand information, such as talking directly with the Dev, as such information is more valuable.
I think the biggest problem with the second part is that Dev doesnt do it anymore. You thought he would do it, but he didnt. Or maybe he didnt do it anyway, but you didnt contact him at all and didnt know the situation of the project. You still bought the second part. Thats stupid.
2. Position opening and position management
“Always remember not to go all in. As Charlie Munger said, if I knew where I was going to die, I would never go there.”
In addition, building a position requires patience. Don’t rush to buy the position you want. At the same time, don’t do the so-called technical analysis, because the pool capacity on the chain is limited. If a big investor suddenly smashes or deceives, it is easy to break the support level of the K-line, but this does not mean that you should enter the market at this time. It is still necessary to grasp the two points: first, have a sufficient understanding of the fundamentals; second, look at the behavior of the market makers wallet. Because the volatility on the chain is very large, only by doing these two points well can you ensure that the action is not deformed in the volatility and get a high winning rate.
In Timos own experience, most of the targets he bought had floating losses at first, for example, he bought 1% and suffered a floating loss of 30%, but his analysis was thorough enough, so he could accept it. Because I know it will definitely go back up, so I probably wont care too much.
How to value AI targets?
The reason for asking this question is that valuation determines how to stop profit. Timo believes that there is no absolute answer or standard for valuation, so he tried to classify it:
The first one is to use the benchmark valuation method for a new target in an existing track.
Basically, for an existing narrative/track, the market value of the first dragon in this track is the upper limit of all the targets in this track, and the second dragon is about 20% of the valuation of the first dragon. But in the field of AI, a new situation may arise, that is: in the same track, the level of the later projects exceeds all the previous targets, then the principle of the second dragon is 20% of the valuation of the first dragon may no longer apply.
For example, Timo believes that the AI targets on the market now only have some leading advantages but lack moats. If the core Dev of Open AI comes out to do a framework project, it will be necessary to re-examine it. It is likely that the valuation of this project will exceed that of ai16z.
The second type is that for a completely new project with no reference, valuation can only be done through cognition.
“For example, when AI16Z first came out, based on my understanding of this thing, I thought it was worth 1 B, but some people thought it would be higher. The same is true for TRUMP. Some people think it is worth 1 B, some think it is 10 B, and some think it can reach 100 B, so they dared to take a large position when it was 10 B. This thing can only be said to be a matter of opinion.
For SpeedPass, the most useful thing to refer to is emotions, and through group chats and content on social media, we can determine where the chain of communication has gone.
Back to how to stop profit, Timos method is:
Give the project three valuations: the valuation of the first stage, the valuation of the second stage, and the final valuation. If the first stage reaches the target position, 30% will be paid, and if it reaches the second target position, another 30% will be paid.
Of course, this is a changing thing and cannot be fixed into a standard. When it comes to taking profits, everyone should do it based on their own acceptance of profits and drawdowns. For most people, selling at high speed should be better than taking drawdowns.
Are the fundamentals themselves important, or is it what the market thinks they are that matters more?
If it is an ultra-short-term target, then what the market thinks it is is more important than what it is in itself, because when speculating on the first wave of fomo, what is essentially needed to push up the price of the currency is to attract everyones attention and liquidity.
But if you want to hold a target for the long term or medium to long term, you should be more concerned about what it is. Because the iteration of this market is very fast, if you are not good at it, it will be falsified soon.
If you are doing the second stage, you still need to spend more time to care about the basic capabilities of a project and judge its fundamentals, because these will screen out some targets. You have to accept missed opportunities and follow your own trading system.
How to continue to grow?
Timo raised four points:
The first is to shrink your ego. This view has been expressed in different words by many people, including Zhang Yiming. The essence is to put aside your emotions and external noise and see the essence of things.
Second, stay curious and open-minded. In the Crypto market, different things are played at different times. You should always stay curious and open-minded to accept them. If the market hotspots change or the mainstream trends change, you should be able to follow up quickly.
“Although I still firmly believe that the main trend will be AI, if it is something else, I will quickly change my own playing style and strategy.”
Third, learn to review and reflect. Don’t make the same mistake again. It seems simple but it is actually very difficult to do.
Fourth, accept opinions and doubts, don’t get hung up on whether you agree or not, it is important to absorb more good things from the outside world.
The three people and three books that have had the greatest influence on Timo
The first one is Mises, the founder of Austrian economics. He helped Timo understand the relationship between monetary cycles, governments and people’s economic behavior when he was in college, and helped him better understand the world.
The second one is Duan Yongping. A persons ability to win continuously in different fields is definitely not luck but ability. One of his famous sayings is to do the right thing and stop immediately if you make a mistake. I think this is far more important than being smart and working hard.
The third one is Taleb, the author of Fooled by Randomness. Many events are actually random, and there is essentially no so-called causal relationship that humans give them. I think the key is to respect objective laws, respect probability, and also respect black swans.
The three recommended books are: Fooled by Randomness, The Crowd and Built to Last. The third book was very helpful for Timo, who was investing in Web2 at the time, to understand what an evergreen company is.
Final Thoughts
Timo said that he started playing on-chain last year and actually spent a long time learning what kind of game on-chain is, from dev to narrative to market making, etc. It is very important to understand the market rules before participating.
I totally agree.
When I was playing games at a VC friends house before, he also said something similar to me: first determine the purpose of the game designer, and then start playing the game. You have to know the rules of the game, how the different links are thinking, how they operate, so that you can play well. This is actually a very important reason why I became a dialogue trader.
Thanks again to Timo for participating in the dialogue with traders. Past Space audios will be updated in Xiaoyuzhou one after another. Search for dialogue with traders.
