In the rapidly changing AI market, investors need sharp insights and decisive actions to seize hundredfold and thousandfold opportunities. This article will review the investment dynamics and successful experiences of the past few months. This article is derived from a piece written by Tomb and organized and translated by 深潮 TechFlow. (Background: The Economist shouts: 2025 is the era of 'AI Agents', but three difficulties must be noted) (Background Supplement: DeFi merges with AI: How the three core scenarios of DeFAI promote the evolution of decentralized finance) The main points of this article: Review the dynamics of smart agents and AI development over the past few months; Review some obvious investment opportunities in the market; During the New Year period, I took a slight break, observed many things from an outsider's perspective, and conducted an in-depth reflection on past successes and mistakes; Next, I will try to clearly share all my thoughts; Reviewing the dynamics of smart agents and AI development over the past few months AI Smart Agents Review I am very satisfied with the performance at the end of last year. From September to early October, we entered many narrative areas about AI, smart agents, and infrastructure very early. Without overcomplicating the analysis of these trends, it was clear to see that these are the next stage of the AI supercycle (there will be more important catalysts in the future). This mature judgment comes from long-term experience in capturing narratives in the market. When you can identify these trends, you can quickly invest funds without too much hesitation. Even now, I still see some people arguing about irrelevant issues such as the packaging applications of ChatGPT, the specific functions of a certain smart agent, or a certain protocol's failure to respond in time. But are these really important? Our purpose here is to make money, not to prove ourselves right. The key is to be clear about the phase of the market cycle you are in. Every cycle has its unique rules and strategies. This is also why many people hold onto old projects for too long, thus missing out on some hot trends of 'simple patterns'. What I mean by 'old' is not in the literal sense, but has become 'relatively old' compared to the rapid pace of industry development. Remember, new hotspots emerge every day, and the speed of capital flow will become faster and faster—if a team cannot clearly communicate their ideas or present a clear roadmap, funds will flow to those agreements that communicate better and can quickly execute projects. When I first came across Goat, I realized its potential and paid attention to it very early—I was surprised that others took so long to migrate to Base via cross-chain bridges or to start exploring smart agents on Sol. This phenomenon is largely due to the 'Mid Curve Effect' (referring to overly complicated thinking) and market panic. If you only realize these issues now, you need to seriously reflect on the reasons. I am very grateful for the past few months because I did not experience too many serious losses. I seized the opportunity with ai16z at a market cap of 30 million, and also entered at a market cap of 30 million for aixbt, in addition to some gaming and conversational projects at market caps below 10 million. Of course, there were also some missed opportunities, such as Zerebro, Fartcoin, Snai, and Swarm Node. But that’s okay; it’s impossible to seize every opportunity. The important thing is to reflect on why you missed them and why they succeeded. Quality As always, quality is always the key to determining the outcome: high-level developers; excellent branding; innovative ideas or unique technologies with competitive advantages; cohesive high-quality communities; clear and valuable information and communication; these are the core elements we have always focused on. If two or more of the above are missing, there is risk. Why? Because there will definitely be other teams that can do better. For example, in the current environment of Virtuals, as long as there is a high-quality team launching AI smart agents or related infrastructure, I will quickly invest early (i.e., 'go all in'). The identity of the developers is public and transparent (doxxed), they have been deeply involved in this field for over 5 years, and have participated in multiple protocols—as long as I discover such teams early, I will decisively enter. In fact, you don't even need to invest too much capital to achieve considerable returns. For example, I discovered the GEKKO project just as it was starting, when its market cap was only 4 million dollars. Because it was developed by the Axal team and had received investment from a16z—this was enough for me, so I decisively invested. As a result, it rose from a market cap of 4 million to 40 million, yielding very considerable returns. I believe it will have further growth. Turning 4000 dollars into 40000 dollars is equivalent to a year’s salary for some people. Sometimes, you don’t need to invest a large amount; investing 2000 or 4000 dollars can still bring considerable returns. Don’t underestimate the potential of these funds. Another project I have been laying out is Acolyt; this team is incredibly strong, with narratives similar to GEKKO. I have mentioned before that if there are 100 different AI smart agents that can provide high-quality market intelligence (alpha), then the only things that can distinguish them are identity, user interface (UI), branding, developer level, etc.—this is no different from other areas of life. This is what I have been looking for: those teams that have been focused on building for a long time—those passionate top developers who continuously tweet and constantly launch new products. They are the only destination for my funds; I will not consider other places. Things to Avoid Avoid single-developer teams with undisclosed identities, especially when building speculative smart agent projects; this carries a high risk. Another thing to be wary of is smart agent projects initiated by individual 'internet celebrities'. These projects are often abandoned shortly after their announcement because the developers realize the responsibilities exceed expectations. Struggling Narratives I have to admit that there are some narratives that I personally have high hopes for, but their development has indeed encountered some difficulties. This does not mean these projects are bad, but their progress is relatively slow. Investing funds into a project that seems promising, only to find its performance is below expectations while surrounding projects achieve 2-3 times growth in a short period of time, can indeed be frustrating. These projects often belong to the 'slow-burning' category, or eventually become 'auxiliary experimental products' for other projects. For example, $GRIFT+$REALIS+$OMEGA. I still hold some of these, but in practical operations, I should have cut losses earlier and shifted funds to stronger AI infrastructure projects, such as REI and ARC. The potential of these projects is more obvious, and their performance is more stable. At that time, I invested about 100,000 to 120,000 dollars in these 'experimental' projects, expecting them to capture a larger market share. But looking back, projects like Griffain, supported by Solana, are clearly stronger, while projects like Realis need more time to mature, possibly because their ideas are too advanced, or the team...

#AI