What’s happening: institutions are selling while whales are buying
After yesterday’s rebound driven by the non-farm payroll data, I checked today’s figures and found something even more worth pondering. First, look at two sets of data—completely opposite directions: One side is the U.S. Bitcoin spot ETFs—over the past two weeks there have been net outflows of more than $4 billion, with net outflows of $4.51 billion in June alone, setting the worst monthly record since these products were launched. On the surface, it looks like institutional capital is withdrawing in large quantities. On the other hand, on-chain data shows that during the same time period, large holding addresses (the so-called "whales") net bought $16.7 billion worth of Bitcoin.
After a weekend, I feel much better. Of course, it might also be because I’ve developed a nervous condition. Ever since I got this nervous condition, my spirits have been pretty good. This past Sunday, I chatted with an old friend I’ve known for more than ten years. I told him $NEWT . Who knew he said he’s currently doing a new token launch and already bought 500U—turns out he’s got more courage than I do. He said #newt has a unique design: developers need to pay using $NEWT to connect AI models/agents into the registry. Model developers can also get a share of the royalty income from it—basically giving AI model developers an on-chain monetization path. Others don’t have this; maybe it’s an advantage that will help take off in the future. @NewtonProtocol #Newt
After working with the team to sort out potential opportunities, to verify my judgment I personally bought some $NEWT and will keep observing. Sometimes investing is that simple—you just decide whether you think it will go up; if so, buy and hold.
One of our team members pointed out that Newton’s transaction generates cryptographic proofs (attestation), which can demonstrate that this operation indeed complies with the predefined strategy rules. This kind of verifiability design is valuable for institutional funds management scenarios. Perhaps recognizing this value is what drove the token’s surge.@NewtonProtocol $NEWT #Newt
#newt $NEWT Web3 has a relatively high learning threshold, and foreign languages are also difficult. The steps involved—cross-chain operations, staking, and yield strategies—are too complicated. Let me recommend something: with Newton Protocol, AI agents simplify this entire workflow, making it much easier. Users only need to set a goal, and the rest is handled by the agent. I tried it myself—it’s pretty interesting.$NEWT @NewtonProtocol #Newt
When the Non-Farm Payrolls came out as a surprise, why did Bitcoin still rise? Many people got the underlying logic wrong
Last night’s employment report directly pulled the crypto market back from “hitting new lows” to “green all over.” Many people’s first reaction is: if the economic data is bad, the market should be panicking—so why is the coin price still going up? Let’s start with the data: the U.S. added only 57,000 jobs in June, while the market had expected 115,000—nearly half as many. In theory, a signal like this “economic brake” should make all risk assets fall together, but$BTC that same day the price directly surged from $61,492 to $61,853, up 2.5%; $ETH even more aggressively, it jumped more than 5%, hitting $1,731.
Your account may be "running in the nude"—3 security settings that are easiest to overlook
Recently, a friend’s wallet was stolen, resulting in a loss of 30,000 U, and it went to zero overnight. After sighing and reflecting, our team conducted a self-audit and put together an educational knowledge piece. Open your phone and your encrypted-asset account may look fine, but many people’s security settings are still at the "as long as it works" level. In most cases of stolen withdrawals, it’s not because hackers are so skilled—it’s because the account itself has vulnerabilities. And 90% of these vulnerabilities can be fixed in advance. First, let’s talk about how to set up 2FA (two-factor authentication). Many people think that binding a mobile phone verification code counts as two-factor authentication. In fact, SMS verification codes are one of the easiest methods to compromise. SIM card theft, intercepting SMS messages—these tactics are nothing new.
Yesterday, I thoroughly reviewed the opportunity points with the team, and also revisited the revenue performance for June. One of the team members pointed out that Newton has an advantage: its operator network generates a cryptographic proof (attestation) for every transaction that passes verification, functioning like a “compliance credential” that shows the operation indeed follows the predefined strategy rules. This verifiability design is very valuable for DAO and institutional fund management scenarios.@NewtonProtocol $NEWT #Newt
I spent half a day with my team brothers doing a brainstorming session. Two actionable suggestions:
🌕If you decide to hold long-term, staggered DCA is better than going all-in at once—it can cope with changing conditions more effectively. 🌕BTC/ETH vs. altcoins: The former has a stronger institutional and compliance-backed rationale, while the latter relies more on specific narratives and team execution.
Two altcoin investment suggestions:
🧧Future unlock pressure is relatively smaller Compared with tokens in the same tier, the circulating supply of $NEAR has reached nearly 96%+, meaning there are almost no large team unlocks hanging overhead; While $AVAX still has ~40% to be released, it follows the predetermined emission/unlock schedule. The market has already priced in these expectations, so it’s less likely to be hit by a sudden “unlock-and-sell” attack like newly emerging tokens.
🧧Institutional backing is relatively solid Both are projects backed by clearly established foundations (NEAR Foundation, Ava Labs), not purely community-driven or anonymous team projects (in contrast to BERA’s “community-driven” nature; Sonic is even more speculative). In a bear market, projects with stronger institutional backing typically have better downside resilience and higher survival rates.
🧧Complementary narratives, no duplication High-performance consumer chain/DeFi narrative AVAX: subnet/modular architecture More enterprise/IBC-style, institution-customized chain applications NEAR: an AI narrative chain, betting on the direction where “AI + blockchain” converges.
Today I looked into the technical architecture of $NEWT . The Newton Protocol uses two technologies: Trusted Execution Environment (TEE) and Zero-Knowledge Proofs (ZKP). It can both ensure the security and verifiability of automated operations, and protect users’ private data from being exposed. This approach is a fairly solid technical roadmap in the DeFi automation track. @NewtonProtocol #Newt
$OUSD was officially released on June 30, 2026. USDT, USDC, and USD1 are joining into one chat—did we get new members in the village?
This OUSD isn’t simple. I’ve summarized 3 key points:
🧧 A new stablecoin project promoted jointly by 140+ companies. These are big names—take a look if you don’t believe me: Visa, Mastercard, Google, BlackRock, Stripe, Coinbase, etc.
🧧 It doesn’t just want to compete with other players. It comes out with a major move right away, making everyone else step aside.
🧧 The market generally believes it will quickly change the landscape of stablecoin competition.
Buffett’s Donations vs. the Han Hong Foundation: Who Should Foot the Bill for Charity?
When it comes to domestic public welfare foundations, if I have to say which one I trust the most, it would still be the Han Hong Foundation. Han Hong herself does not take a salary from the foundation. According to the (Beijing Han Hong Love Charity Foundation 2025 Annual Report) disclosed (as reposted by multiple media outlets including Sina Finance, 21st Century Business Herald, Tencent News, etc., originally from Beijing Youth Daily and Securities Times), in 2025 the Han Hong Foundation recorded total income of 782,687,819.31 yuan and cumulative total expenditures of 291,978,046.35 yuan. Recently, I occasionally see news that Buffett has donated more than $60 billion and that the personal share of his wealth he has pledged to donate is 99%. If you put Buffett and the Han Hong Foundation side by side, you’ll notice a rather striking contrast.
#opg $OPG More and more people have been discussing AI agents lately, but I think the real factor limiting their deployment is not model capability, but trust. If AI can help us manage assets, handle work, or even read personal information, then data security becomes the top priority. Notably, @OpenGradient ’s design approach goes beyond just offering AI Chat—it also strengthens privacy protection using technologies such as TEE and O-HTTP, enabling users to feel more confident using AI services. I believe the key in future competition among AI infrastructure providers will be who can balance performance with trustworthiness. Keep an eye on the continued ecosystem development of $OPG , and hope to see more real-world applications take shape.#OPG
Europe’s Heatwave and Its Impact on Crypto Markets
At the end of June, besides reports of extreme heat in India, all of Europe is also experiencing a deadly heatwave. The World Health Organization directly pointed out that the number of abnormal deaths in recent days caused by extreme heat has already exceeded 1,300. Temperatures in multiple countries, including Italy and the Czech Republic, have repeatedly broken 40°C. You should know that London, Berlin, and China’s Greater Khingan Range and Mohe are almost on the same latitude—yet now they’ve broken 40°C, repeatedly shattering June heat records, and in some places even surpassing the all-year highest temperatures. Global map Many people think this is just weather news. Actually, it's not.
When I look at AI projects, I’m increasingly paying attention to infrastructure rather than a single application. Applications will be updated and replaced, but the underlying network, privacy computing, and trusted execution environments may be where long-term value truly lies. @OpenGradient What I want to solve is the issue of "trusted AI"—so users can verify the computation process while also preserving privacy. This direction is clearly different from traditional AI platforms. If in the future Agents, DeFi, and on-chain identities all begin to integrate with AI, then a trusted and verifiable AI network could become a new piece of infrastructure. Of course, the market ultimately depends on real-world deployment, but this space is worth continued attention.#OPG Keep watching $OPG
I’ve been trying @OpenGradient recently, and I feel the biggest difference from using AI before isn’t how brilliant the answers are—it’s that it feels much more reassuring to use. Before, when I ran into questions involving wallets, investing, or code, to be honest I wasn’t really comfortable asking directly. I was always worried about whether my data might be saved or used to train models. What attracts me about OpenGradient Chat is that it places more emphasis on user privacy and data security. By leveraging technologies like decentralization and TEE, it makes the entire AI usage process more trustworthy. I think in the future, people won’t only look at which model is smarter. They’ll increasingly care about whether their own data is truly safe. The future of AI shouldn’t just be smarter—it should also be more trustworthy. Keep an eye on the developments after $OPG and see what new ways it can bring. #OPG
With the overall market pulling back recently, I’m not planning to wait for the absolute bottom—I also don’t think I’ll manage to catch it. Starting today, I’ll be adding to my position in batches.
This time, I’m investing another 30,000 U. I won’t chase the price up, and I don’t intend to trade frequently. Instead, I’ll buy gradually according to my long-term allocation.
📌 My allocation: 🟠 60% $BTC — core position, holding long-term. 🔵 20% $SOL — bullish on the ecosystem and growth potential. 🟢 20% $ONDO + $PENDLE — both are industry leaders with high risk and high returns. ONDO is an important project in the RWA sector, while PENDLE has built its edge in DeFi through its yield tokenization mechanism.
My plan is simple: After I buy, I’ll hold at least for 4 years.
I don’t predict tomorrow’s price movements—I only believe in long-term value. I don’t follow market sentiment; I stick to my own investment discipline. (Sharing my personal allocation only—this is not investment advice.) #BTC #SOL #ONDO #Pendle #加密投资
Many people are discussing how AI models are getting stronger, but the real thing that is easy to overlook is data ownership. Every conversation with an AI may include personal habits, work content, or even business information. If these data are used by platforms to train models, users have almost no choice. After noticing @OpenGradient , I think the “verifiable AI” approach it proposes is valuable. Through privacy computing and trusted execution environments (TEE), it enables users to use AI with more confidence rather than enjoying convenience while worrying about data leakage. In the future, competition in AI won’t be only about model capability, but also about who can truly protect users’ privacy. I will continue to follow the development of $OPG .#OPG