Haotian

趋势研究员| Advisor @ambergroup_io｜产业洞察 | Crypto・AI・具身智能・美股代币化・预测市场等｜Previously：@peckshield | DMs for Collab

0Following

0follower

Feed

After in-depth discussions with several experienced friends in on-chain activities recently, we reached a strong consensus on the survival rules for the current cycle. The market has completely shifted from "listening to stories and speculating on expectations" to "looking at cash flow and verifying implementation." Below are some trading principles for crypto (for reference only): 1) Prioritize assets with real value capture ability. In a bull market, the market is willing to pay for stories and expectations; in a bear market, only real cash flow and buyback/burn records count. The true "immunity card" this cycle is protocols that can continuously generate fees and directly return these fees to token holders through buybacks, burns, or dividends. For example, recently well-performing launchpad concept tokens like $UNI, $PUMP, $PONS, and the buyback king of this cycle, $HYPE; 2) Only choose projects with PMF (Product-Market Fit) realized and a complete closed loop formed. Because barring surprises, the next cycle will only have two major narratives related to "asset tokenization" and "Agentic Economy" (Perps, prediction markets, stablecoins, Payment). The market will shift from favoring technical narrative extensions to practical implementation verification. Projects without real users, real transaction loops, or real revenue will be quickly filtered out. Concept tokens following this logic include $ONDO, $VVV, $VIRTUAL, etc., where actual AUM, trading volume, and fee generation ability will be key data indicators; 3) Choose assets with strong "consensus." It must be admitted that after several cycles in the crypto industry, the only thing that withstands the test is "consensus." Note that this consensus is naturally formed by the market and has cross-cycle capability. Do not mistake a reply to a tweet or industrialized hype for so-called "consensus." The truly promising assets are those that newcomers don’t understand at all but still maintain good liquidity and survive well as old assets. For example, old cult MEME tokens like $DOGE, $PEPE, $PEOPLE, or leading assets in various niche sectors like $ZEC, $TAO, which have gone through multiple bull and bear cycles, have strong organic community vitality, and are often targeted by major funds for repeated turnover and manipulation; 4) Try to avoid pure VC tokens. If I say altcoins are dead, you might argue with the cyclical nature of finance, but if I say VC tokens are dead, basically no one disagrees. Because VC tokens with high FDV, low circulation, and continuous large unlocks can only rely on airdrop expectations around TGE to generate hype. If the project lacks value capture ability, it inevitably faces insufficient development momentum and the awkward situation of price dumps upon unlock. This is the fundamental reason why "bulls don’t go crazy, bears go deep" this cycle. Many hungry VC tokens are waiting to unlock and dump—how can retail investors dare to touch such tokens? Note: The above is only a summary of personal and friends’ discussions, and the tokens mentioned are examples only, not investment advice.

Let's talk about the much-debated latest Ethereum proposal EIP-8363: 1) The starting point of this proposal is certainly good. At a future critical point, specifically when the total network staking rate exceeds 50%, the consensus layer's issuance rewards will be completely "burned," relying solely on the execution layer's real income such as MEV and tips to incentivize stakers. This approach will be a monumental benefit for Ethereum holders, as it reduces the continuous issuance of ETH and indirectly forces the market to grow economic vitality to generate value, rather than purely relying on extracting value from holders' shared interests to maintain network security budgets; 2) Short-term resistance to the proposal is inevitable because it infringes on vested interests and may disrupt the current DeFi market balance. For example, LST protocols like Lido and etherFi will have their earning capabilities affected since most of the rewards they give to staking holders currently come from the consensus layer issuance rewards of about 2.6%, while MEV and other fees only account for 0.2%. This proposal effectively cripples their future yield-generating ability; AAVE and other legacy DeFi lending and revolving loans also rely heavily on this portion of issuance rewards for interest. Once curtailed, the impact is similar. Institutional treasuries represented by @fundstrat are even more affected; depriving them of this native staking income and expecting them to hold tokens with zero yield purely on faith is simply unrealistic; 3) Objectively, these short-term resistance voices can be regarded as "noise." As mentioned above, their anger mainly stems from concerns that their businesses might fail, but they overlook that this proposal is only under public discussion. First, the 50% staking rate has not yet been reached, and the proposal's existence can prevent the market from purely relying on staking rewards to extract value (blocking the achievement of the 50% staking rate). Second, the proposal's implementation will have a transition period of up to 18 months, so rewards will not be cut off overnight, giving the market ample time to adjust independently. Moreover, by turning off the switch for unlimited ETH issuance subsidies through this proposal, Ethereum is forced to move toward sustaining itself through real economic activity, which is a positive development. Anyone who thinks about the proposal's original intention would not oppose it outright. 4) Returning to the motivation behind the proposal, it must be said that the authors, especially researchers including @drakefjustin, have a somewhat "ivory tower" perspective because they wrote extensively about the short-term harm to staking rewards but neglected to illustrate how long-term benefits will be realized to boost confidence among ecosystem participants. Broadly speaking, reducing inflation is in the common interest of all ETH holders, including those opposing the proposal. However, without explaining how the execution layer's income will incentivize the network in the future, it is hard to prevent ecosystem contributors from feeling anxious. @VitalikButerin 5) @Solana_zh actually has similar proposals, such as SIMD-0550/0228, which directly reduce inflation rates and staking rewards. But Solana's inflation base is higher, and with some parameterized design plus an implementation curve, the market acceptance is somewhat higher. The original intention and future necessity of Ethereum's proposal are definitely there, as no network should rely on continuous high issuance subsidies to sustain itself. But it comes down to balancing short-term and long-term interests. No matter what, this is far more than just issuing a proposal.

I guess I should say something about the token Look. I worked my ass off to the point I got a frozen shoulder and severe health issues from overworking and typing, and it was never enough We built cool shit but it was completely ignored because number down It felt like the things I cared about were completely unappreciated by all but a small group who were drowned out by endless complaining, endlessly calling me a scammer and hating on everything Burwick sued us and we settled with a group of holders for the rest of the treasury and all the money we had. Their claim was ridiculous, but we didn’t have the capital to legally fight it so we settled on giving them the rest of what we had. I never sold my ai16z, never made money other than a modest salary, making as much as our other engineers from essentially a day job. Through all I said we’d keep building Eliza. I’m living on savings, building Eliza. We made some versions of Eliza like Milady, people tokenized, we got some fees, those went back into Eliza. Along the way we tried to make money and make things people told us they wanted. Those things were incredibly hard to build, with developers quitting to launch tokens, getting addicted to drugs, getting in fights with each other, building their own projects on our payroll and then leaving to make their own startup, I mean I’ve seen it all. I wanted collaborators with high agency, I didn’t want to be a manager, and I leaned a lot of lessons the hard way in human alignment. My mission has always been the same, unwavering, toward a personal and social agent who could help us do all the things we didn’t want to do, to put the A in DAOs, to enable the internet itself to generate value through open source coordination. The most expensive thing I own is my MacBook. I don’t own a car, I live in a small bedroom in SF with my wife, I write code and work almost all day every day. I don’t desire luxuries or wealth, and if I had more money I’d put it all back into Eliza and funding other developers. Not to virtue signal, but to give you some perspective— if I am a scammer then I am the dumbest and worst scammer ever, because I had $25m in my wallet and I coulda dumped on all of you and I ran it to to 0. And now as I post cool free things that I made just to make, just show off with no benefit to myself except some social likes, it’s just hate, pressure, cynicism and complaining dominating the comments. A bunch of people will surely take this opportunity to point out some hypocrisy or some small grievance, but it all boils down to a culture of people who don’t take responsibility for their gambling habits. You cannot appease the mob. So I won’t. The token is dead. Completely. The foundation is winding down. I am starting over, since I own the IP, and I am never letting a token come close to Eliza again maybe I’ll farm some fees from a 24 hour shitter to feed my team because 5k is 5k but as long as the casino is a bunch of entitled cry babies I’m never going to support an Eliza token. If you have some, you should either sell or get a cabal to crime it up, but there’s no foundation and no supply coming to save you, there’s no money for buybacks, it’s completely ngmi. A bunch of cry babies killed whatever hope there was left by giving Burwick their name to rape yet another project. The smart traders sold on the heads of the losers and exited long ago, rotating into Trump and such, just leaving the scooters of the world who’s entire business is to complain on the internet about how it’s everyone’s fault but theirs that they lost money. So yeah it’s done. I don’t own any tokens. I don’t support any of it. I love the technology and I’ll come back one day when the culture has grown past this point. But right now there is nothing in it for me busy wasted time, pressure to do dumb shit, people trying to scam me or get me to help scam others, etc. it’s trash. Absolute trash. I suggest you buy a token with real fundamentals, like Ansem’s coin. I’m sure he’ll give you the 10x I couldn’t. We’re still building Eliza and the underlying OS, faster and better than ever, unburdened by all the bullshit and sidequests and pressure to launch launchpad #83837. If that’s not enough for you… I don’t owe you anything. Feel free to complain so I can block you. We’re still gonna be here in 10 years, pushing at the edge of what agents can do and going where the big corps won’t go. We’ll make sure that local, private, crypto-enabled agents are available for everyone, and if Eliza gets outrun by another open source OS project we’ll probably merge into them and help them. But right now we’re out in front. Blockchain is a beautiful technology and I hope the industry can move away from the casino and obvious coordination failures and build real stuff instead of larping harder into chasing someone else’s prediction market market and memecoin success. But I just see more of that, even from the institutional companies, so I’m bearish. AI has none of these problems. People are optimistic, empowered, building the future instead of cutting everyone down. Those are my people. And the audience for what we’re building isn’t on here. The people who need this tech the most have no idea what a token is, either kind of token. So yeah. Gonna keep building no matter what, every single day, and I’m not gonna stop until we live in a world where we each own our own data and we don’t have to pay to be smart. That’s the mission. Eliza is dead. Long live Eliza.

I have noticed a wave of anxiety spreading on Twitter recently, with everyone complaining accurately about various issues: mass wipeouts of altcoins, leading exchanges doing nothing, AI and the US stock market draining talent and capital completely. But what are the deeper underlying causes of these problems? Here are three points of view: 1) This cycle is plagued by severe technical concept bloat, with structural imbalances between infrastructure and application layers. Too many developers and projects are creating new chains, building layer 2 solutions, cross-chain bridges, working on DA, ZK, and parallel EVMs. The tech stack is becoming increasingly complex, and valuations are getting more absurd, yet the actual application side is almost a blank slate. Looking back at the previous crypto cycle, there was a rotation among infrastructure, community, and application sectors like DeFi, NFT, and GameFI. But in this cycle, it’s all about building chains, faster chains, and the communities and applications that could bring incremental growth have almost disappeared. MEME might have briefly carried this utility, but its inherent lack of fundamentals, speculative nature, and high extraction characteristics mean it cannot fill the application layer void. Instead, it causes short-term emotional volatility and long-term liquidity drain. 2) The short-sightedness of vested interests has broken the industry’s intrinsic value transmission mechanism. Rather than blaming leading exchanges for lacking responsibility and commitment, it’s better to recognize their "broker" nature. Ideally, facing a shortage of quality projects and the dilemma of listing and dumping, exchanges should choose to discover, screen, and guide truly valuable projects. Instead, they embraced dark token-issuing groups and completely opened the floodgates for MEME tokens. As a result, a large number of fundamentally baseless, purely emotion- and extraction-driven memes were mass-launched, mercilessly draining market liquidity. The original layered value and liquidity transmission mechanism from on-chain to small, medium, and large exchanges has been completely destroyed. In the short term, exchanges did gain trading volume and fees and enjoyed support from short-term FOMO communities. But in the long run, valuable projects were pushed to the margins, resulting in a lack of tokens to list and reliance on the US stock market for survival. This has long been described as a surrender of discourse and pricing power, effectively a self-castration for the native crypto industry. 3) The cohesion of the crypto community is gone, and the industry’s innovative vitality is drying up. The most fascinating aspect of the last cycle was the constant on-chain creativity and sector rotation, with funds flowing from technical infrastructure to applications, then to community-driven games, social sectors, and so on. Activities like yield farming, open-source contributions, exploring niche tracks, and DeGen trading meant that deep work in a certain field could one day yield unexpected rewards. In contrast, this cycle’s technical narratives are highly homogenized and competitive, primary market VCs can’t find exit paths and have stopped investing, and many developers and communities receive no positive feedback for their on-chain efforts. Over time, the once-proud on-chain innovation experiments and vitality of crypto have no nourishment, and many excellent developers are gradually being drawn away by the AI field. While many complain that AI has weakened crypto’s appeal, the real reason is the internal exhaustion of innovation within the crypto industry.

Everyone is focused on the myth of Changxin Technology hitting a 3.3 trillion market cap on its first day of listing, but digging deeper, the layers of logic behind the capital frenzy are more worth discussing: 1) Changxin Technology’s true valuation by the capital market is not simply a “domestic substitution story.” Essentially, it has broken the near-monopoly held for decades by Samsung, SK Hynix, and Micron, forming the fourth pole in the global DRAM market. Previously, the three giants collectively held over 90% market share for a long time, with pricing power highly concentrated; the cyclical ups and downs were almost entirely driven by their capacity rhythms and inventory strategies. Changxin Technology started from zero and by Q1 2026, its global share has risen to about 7.7%–8%. Moreover, as the three giants have shifted their most advanced capacity massively toward high-margin HBM, the supply of general-purpose DRAM (for phones, computers, ordinary servers) has been systemically squeezed, and Changxin just happens to fill this structural gap. Combined with the Chinese market’s growing ability to influence supply-demand and price elasticity, the emergence of Changxin as the fourth pole can compress the joint price control space of the three giants. This is the fundamental logic behind the capital market’s valuation of Changxin Technology; 2) There is a popular story that Hefei’s state-owned capital’s ten-year industrial investment returns have outperformed ten years of land sales, but more importantly, Hefei’s successful investment example may rewrite the capital valuation and entry appeal of semiconductor stocks on the A-share market. Previously, the STAR Market’s storage sector long had “design but no manufacturing leader,” and institutions (pension funds, index funds, foreign capital) lacked a large-cap anchor for global framework allocation. Changxin’s example proves to the market that the A-share hard tech sector is not only about high-volatility, small-cap stories but can also nurture manufacturing leaders like Changxin that benchmark against global peers. This undoubtedly lowers the financing difficulty for subsequent large projects, making the A-share STAR Market, which is mainly retail and theme-driven, tilt more toward institutional and performance-driven investment, thereby fostering many new investment opportunities in the A-share STAR Market and attracting more long-term capital continuously; 3) Of course, some worry that Changxin Technology’s short-term exaggerated valuation might cause a sustained bloodletting shock like historical cases of PetroChina and SMIC. After all, the impact of @SpaceX on the entire US AI tech sector still sends chills. However, Changxin’s freely tradable shares on the first day were only about 6.7% (around 4.5 billion shares), with over 90% locked up, and combined with no price limit for the first five days, the tiny tradable float chased by massive capital naturally forms a high-premium “scarce asset” pricing. The fact that GigaDevice, which previously enjoyed Changxin’s premium, plummeted today shows that market expectations are shifting from “concept” to “real manufacturing leader.” This will certainly cause some bloodletting for older, more concept-driven stocks in the short term, but in the long run, capital shifting from stories to manufacturing leaders with real capacity is not a bad thing. Perhaps under this new structural revaluation trend, a batch of value stocks with real capacity and supply chain demand will be rediscovered.

.@BitMEX and @BitMartExchange both ran into trouble one after another. I've seen quite a few people gloating, thinking that some exchange "explosions" are needed as fuel to drive the next bull market. Well, there are indeed some indicators of a bear-to-bull transition, but the reshuffling logic behind it might be different from what most people think: 1) Under the overall compliance trend, competition among CEXs is far more brutal than imagined. Licensing, reserve proof, KYC/AML/KYT, client asset segregation, and other compliance issues have become the entry ticket for exchanges to survive. This greatly compresses the space for "explosions" in CEXs now. The past zero-sum game logic of "one whale falls, everything grows" is gone. So, rather than calling it an explosion, it’s more like a proactive shutdown under huge competitive pressure, which is a result of healthy market competition; 2) The competition among CEXs in the tokenized US stock track is actually a strategic move by exchanges to actively expand channel business. It also signals that the old platform operation model relying mainly on listing fees and trading commissions is no longer viable. Introducing tokenized US stocks, ETFs, Pre-IPO assets, and other traditional financial asset targets urgently requires new revenue sources and growth scenarios. However, the replacement of crypto-native assets by traditional TradFi assets means CEXs lose pricing and settlement power. In the short term, Perps seem to maintain trading volume and income, but in the long run, the cost of the pricing center becoming just a channel and entry point must be faced. Therefore, the fiercer the tokenized US stock competition, the greater the survival pressure on CEXs. You see, those unqualified to compete basically can’t survive; 3) Currently, CEXs, especially small and medium exchanges, must find a differentiated positioning to survive. Just like small exchanges in the last cycle attracted traffic and users through IEOs and quality on-chain assets, now it seems there is only one way out: Either deeply cultivate specific regional licenses and localized services to exploit regulatory arbitrage, or focus on a particular niche product such as TradFi assets, Perps, RWAFi, etc., or fully embrace crypto-native innovation narratives including DeFi, Agentic Economy, MEMEs, and use the power of crypto-native communities to endure the cycle. In any case, continuing homogeneous internal competition will only accelerate the elimination wave. That said, clearing out some weaker competitors is not necessarily a bad thing.