棉花白

After Coldcard was compromised, $15 billion worth of Bitcoin was moved to a safer place.

My first reaction was: this sounds like an escape.

But for Casa, which offers self-custody, its CEO said that this is precisely the Bitcoin immune system at work—not a vulnerability in self-custody.

I agree. Self-custody isn’t a guarantee that nothing will ever go wrong; it’s that when something goes wrong, you can still take action yourself. The money is in your own hands—when you notice that something in the process is off, you can move it immediately. If it’s with someone else, you can only wait for announcements, for recovery efforts, for reimbursement—a result that may or may not come.

Moving $1.5 billion isn’t that major holders suddenly lost faith in Bitcoin; it’s that someone discovered a risk in a certain piece of hardware, then quickly moved it. Migration itself is defense.

For ordinary people, the takeaway is this: a cold wallet isn’t something you buy and then just keep it around. It’s something you need to occasionally check on, update the firmware, and store in a distributed way.

The point of self-custody isn’t a one-time, set-and-forget solution—it’s that you still have the power to make the right decision.

Seeing Tencent’s Q2 cash flow turn negative, my first reaction wasn’t panic—it was that this AI industry is really heavy.

After excluding payments for computing prepayments, its books still show RMB 37.6 billion. Put the other way around, that negative RMB 13.8 billion is basically what got plugged back in as prepayment for computing capacity. Liu Chiping put it even more bluntly: in the worst case, you can choose to rent out infrastructure. Translated: whether the AI-native business can actually make money isn’t certain; if things don’t work out, then just be a landlord renting out data center space.

Shanghai’s issuance of computing vouchers, model vouchers, and dataset vouchers—looks like subsidies, but underneath it boils down to one sentence: using AI isn’t cheap. Apple is negotiating with publishers to pay them, because in the future Siri will have to read content from them and can’t just keep getting it for free. Google is also restructuring again, focusing entirely on pushing Gemini.

After looking around, it’s clear that every big company is betting on a future—but cash flow is what matters now. What ordinary people can benefit from is the small slice of usage cost that comes down when those computing vouchers reduce prices. My most direct takeaway for the moment is this: even if AI is amazing, computing capacity still requires real money upfront. If someone tells you an AI project is guaranteed profit, first think about who’s going to take the hit.

The incident where ST Nanxin was fined: the amount is a little absurdly small. In 2021, it overstated revenue by 5.2563 million and profit by 1.5732 million, accounting for less than 1%. For just these numbers, the company was labeled ST, and the relevant responsible parties were issued warnings.

But if you look at it from another angle, it’s even more unsettling. A return-related matter was handled in accounting a year late, and that could shift revenue from 2021 to 2022. If it weren’t discovered by an investigation, outsiders would not be able to tell this move. With revenue at 0.77% and profit at 0.71%, on the financial statements it’s just the weight of a single leaf.

Many investors read financial reports by first looking at growth, gross margin, and whether there are any major anomalies. Something under 1% simply won’t catch their eye. But the real problem often hides in places that are “too small to seem like an issue.”

Financial fraud rarely starts with a huge scheme all at once. Most of the time, it begins with moving a small account. Once they taste the benefits, their hand reaches further and further.

The takeaway for ordinary people is simple: when you see a company penalized for a “minor issue,” don’t rush to say it’s “overblown.” Think about it—if they’re willing to tamper with even 0.77%, what portion might still be undiscovered?

Have you been feeling lately: those “story-backed” coins in your hands—no matter how you try to tell the story, you just can’t get people to listen anymore?

It’s not that your coins are broken. The market has simply changed its temperament—back then, everyone listened to stories; now everyone only cares about whether something is “certain or not.”

First, money has gotten less. In the second quarter, the total crypto market cap slid downward, and trading volume shrank too. Even “bullets” like stablecoins have been running lower. With too few bullets, you can’t spread the fire evenly—you can only concentrate on a handful of targets.

And money has become pickier. Rates, geopolitics, oil prices—these things keep getting stirred up. Big money would rather stay put in “safe” places like cash, gold, and the U.S. stock market than rush in to buy small coins with high volatility. Institutions are even more so: they only hold tight to BTC and ETH, and they won’t proactively “open the spigot” for the altcoins.

As for the stories around AI, RWA, L2, DeFi, and MEME—there are plenty of people telling them, but many tokens are still sitting with unlocks and sell pressure, and their revenue is often nothing more than projections. Money is only willing to go to a small number of coins/projects that genuinely have cash flow and real users.

So it’s not that there’s “no market,” it’s that the market is “choosy now”: only a few projects with real substance will move, while most coins just wobble along with BTC—and then fade away.

For ordinary people, the takeaway is simple: don’t let yourself get pulled in by saying “the story is so good.” Ask first: “Does it actually make money? Is there really someone using it?” A great story is less convincing than real cash on the books.

Is there anyone around you who has been instructed by scammers to go to a crypto ATM and transfer money?

In the past, I always thought that if money in the crypto world was stolen, it was basically gone for good. Recently, a law in Arizona helped 35 people recover $171,000—an average of less than $5,000 per person. It’s not a huge amount, but the money really was returned.

This law doesn’t rely on hacking skills or luck. It requires victims to do two things at the same time: notify the operator of the crypto ATM as quickly as possible, and file a report with law enforcement. If the new customer acts fast enough, they can get a full refund, including the fees.

It sounds simple, but most people can’t do the first thing. After being scammed, the initial reaction is confusion, embarrassment, and searching online for “can it still be recovered?” rather than contacting the operator immediately. By the time the emotions settle, the crypto has already been transferred and the window is closed.

In cases like this, scammers often don’t want you to use an exchange. Instead, they make you go to an ATM—because turning cash into crypto is fast and harder to trace.

$171,000 isn’t a massive sum. But for 35 ordinary people, each transaction could be rent, wages, or money they saved up over a long time. The law can get it back because someone managed to beat the clock.

So this news has only one takeaway for me: after you’re scammed, don’t just file a police report, and don’t just scold people. First, notify the crypto ATM operator involved, keep all chat records and transaction receipts, and then report to the authorities. Not every state has Arizona’s full-refund law, but notifying the platform in the first instance is right everywhere.

Protecting your money is something you do in advance. If something really happens, don’t stay silent and don’t delay. The window for recovery only stays open for people who act fast.

SanDisk is up tonight—let me tell you something more down-to-earth: your next phone, or the next solid-state drive you buy, will most likely be more expensive.

That’s because AI data centers are fighting over high-performance flash memory, pushing up the price of NAND chips. In the last quarter, SanDisk’s revenue was about $9.97 billion, up 51% quarter-over-quarter. It said that two-thirds of this growth came from “price increases”—not selling more, but selling at higher prices.

Before, SanDisk was one of those cyclical stocks that basically moved up and down with storage prices. But in its investor day presentation, it pitched a new story: in the AI era, it’s infrastructure—and it even promised that in the future it will return 100% of excess cash to shareholders.

For ordinary people, the real-world feeling is this: once AI takes off, the first thing to get pricier isn’t the graphics card—it’s the chip that stores data. You either check the prices of drives and memory right now, or you wait and then get hit with the sting the next time you upgrade your phone.

Compliance first embraces the big players, not you and me.

Copper’s (a crypto custody firm) U.S. division has just become a FINRA member of the U.S. financial industry regulators, and is also a broker-dealer registered with the U.S. Securities and Exchange Commission (SEC). In the future, it will be able to provide eligible custody, staking, financing, and over-the-counter trading services.

This sounds very proper. But every item says “qualified,” “institutional,” “large amounts.” It doesn’t solve ordinary people’s problems: Where is your crypto kept safely? If something goes wrong, who pays?

While the industry is getting licenses, ordinary people are still standing in the same place. Institutions have qualified custody, big players have access to financing channels, and OTC trading is not something retail users can touch. For ordinary people, the options are still the same: exchanges, hot wallets, cold wallets—you choose, and you bear the risk.

We used to always hope regulation would land, thinking that once it did, the crypto space would mature. Now it really is landing—but it has drawn a line first: inside the line are protected institutional clients; outside the line are ordinary people who carry their own risks.

I’m not saying compliance is bad. I’m just reminding you not to treat an institution’s license as your own safety. It won’t hold your private keys for you, and it won’t compensate you for phishing losses. The things you need to learn still have to be learned by you.

Today, BTC is around 63696 and ETH is at 1884. The market doesn’t seem to react much to this news. But its impact on ordinary people may take a long time to truly be felt. By then, you’ll find that the difference in security has never been just about price.

A crypto exchange sues a country, and the court actually issues a freeze order.

Bybit (a crypto exchange) says that in February 2025, the Lazarus Group (a North Korean hacking group) stole $1.5 billion. Now it has recovered $48.4 million and frozen $30.5 million—together only a small fraction of the stolen amount. The court granted the freeze order, targeting a hacker group supported by a state.

What’s most bewildering is enforcement. Who is the freeze order supposed to be served on? North Korea probably won’t send anyone to appear in court, and the hacking group certainly won’t cooperate with the investigation. The part of this piece of paper that can actually take effect in the real world may be even less than we imagine.

The theft happened a year and a half ago, and only in 2026 does this step occur. It isn’t that the law is intentionally slow—it’s that this is a cross-border, state-backed money-laundering network, so tracing it feels like using a dull knife to cut flesh. Every transfer record may go through more than a dozen countries, dozens of exchanges, and in the end becomes cash, or disappears into privacy coins.

What I’m thinking about isn’t whether Bybit can recover more. It’s whether ordinary people, seeing words like “lawsuit” and “freeze,” will think their money can be saved. In reality, a crypto exchange with massive resources and a team of lawyers might only be able to freeze amounts on the order of tens of millions. If ordinary people lose money, they likely won’t even make it to the first stop.

So for me, this news has only one takeaway: don’t treat recovery as a fallback.

Even if a contract gets liquidated, there may still be scraps left. But if you lose your private key—or it gets compromised—that’s the cleanest ending.

Keep your private key safe, faster than you can ever rely on any court to move.

A Florida man has been sued by the CFTC, accused of misappropriating approximately $48 million in customer funds through a crypto Ponzi scheme, with the entire scheme totaling up to $397 million.

When looking at numbers like these, one thought comes to mind: in the crypto world, news amounts have already grown so large that people have become numb.

$480 million, $390 million, $160 million… Every time I see a report about a Ponzi scheme, my first reaction isn’t anger—it’s “here we go again.” This numbness isn’t because people are cold-blooded; it’s because there are just too many. It’s like a scam template copied and pasted: change a name, swap a face, and run the script again.

Law students might think this makes for great case-study material. But for ordinary people, these stories all read like the same play: promises of high returns, head-hunter/affiliate rewards, flashy in-person events that look luxurious, and then when the whole thing collapses, the boss runs or gets arrested.

And then the victims comment, “I invested too.”

The CFTC’s lawsuit is regulators doing their job, and catching people is a good thing. But to be frank, arresting someone can’t solve the root problem. Once the scheme collapses, the money has already been moved—more than $40 million. So where is the rest? How much can be recovered? Are victims waiting for “the person to be caught,” or “the money to come back”?

Chances are it’s the latter. But news coverage usually only reports the former.

I don’t know the victims in this case, or who they are. But I can imagine that in a community somewhere in Florida, there are people who put their paychecks in, people who did it because a friend recommended it, and people who just wanted to beat inflation. They’re not stupid—they just believed someone they shouldn’t have.

Believing the wrong person can happen in any industry. It’s just that in the crypto world, the cost is often much higher.

Today BTC is around 64,200, ETH around 1,914. The market is quietly calm. No one is panicking because of a $390 million Ponzi scheme, and no one seems to care. People are more concerned with Nvidia’s earnings and whether the Federal Reserve will cut rates next.

But I care about the people who put in their money. They’re waiting for an explanation.

It’s difficult.

Here we go again.

The SEC is holding a meeting this week to discuss crypto regulation, while the CLARITY Act is still stuck in Congress. Put into plain language: lawmakers don’t do the work, so enforcers have to handle it themselves.

What ordinary people fear most isn’t strict regulation—it’s regulation that’s unclear. Today you buy a token and you don’t know whether it’s a commodity or a security. Tomorrow the project founders get sued, and the justification is a legal provision from two years ago—back when Bitcoin was barely born.

I’ve lost count of how long the CLARITY Act has been stalled. In any case, every time Congress says it wants to “clarify the rules,” it ends up being the enforcers who come out and issue warnings. What can the SEC’s meeting actually accomplish this time? Probably draw a few more red lines, issue a few more fines, and keep the industry on edge for a while. But enforcement without a legal framework is like blowing the whistle before anyone’s even marked out the playing field.

I’m not saying the SEC shouldn’t regulate. The problem is that when the rules are blurry, the regulator’s actions themselves become the rules. Today you hold a meeting, release a statement—then the market has to guess. If they guess wrong, retail investors lose money; if they guess right, the wind may change again next time.

BTC is at 63,593, ETH at 1,880, and DOGE is up nearly 3%. LINK is rising even more sharply. The market doesn’t care what meeting the SEC is having right now—the sentiment is elsewhere. But I care about the people being repeatedly worn down by ambiguous rules.

I hope this time the meeting includes a few words humans can understand. At the very least, tell us what counts as illegal, what’s in the gray area, and what you’re temporarily not going to pursue.

Don’t make ordinary people keep guessing.

$1 billion.

Not some big shot “picking up bargains.” Not an institution “calling the shots.” It’s an ETF—over one week, net inflows. The U.S. spot Bitcoin ETF has just had its best week since April.

Some may ask: isn’t the market still falling? BTC is hovering around 64,300, ETH around 1,885—sentiment isn’t great. So where did the $1 billion come from?

Retail investors are in panic, but the money is quietly coming in.

What’s most worth watching in this round of inflows isn’t the amount itself, but the timing. After the major inflow in April, what happened? BTC slowly climbed from the 60,000 range to above 70,000. It may not be that inflows pushed the price up directly—it could be that the people entering at that level saw something.

But even more interesting is another statement. Today, the founder of Nansen said Bitcoin will never fall below $60,000 again. In the crypto world, that’s the kind of remark that’s easy to get slapped by the market. Bitcoin’s history is one of repeatedly breaking through both lower and upper bounds—who would dare say “never”?

Still, if you look at it differently, maybe his confidence doesn’t come from technical analysis. Maybe it comes from the fact that ETFs have changed the holding structure. In the past, Bitcoin’s price was driven by retail sentiment and leverage. Now there’s a steady, compliant, weekly inflow buyer base. The characteristic of this group is that they don’t chase prices, they don’t panic—they buy gradually.

I’m not here to say whether $60,000 will break. But if you’ve been waiting for a dip to $45,000, you might want to consider one question: when $1 billion–level capital is calmly buying in the $63,000 range, is your “waiting for even lower” essentially betting against the direction those funds are taking?

Many people have gotten it right. More people end up with no position when the game is over.

Today I’m not predicting up or down. I’m just thinking that when a market’s loudest voice is panic, but the actual cash flow is completely the opposite, at least one thing is clear: this is not a situation of unanimous bearishness.

Someone is buying—quietly, without making noise.

One person, one wallet, a few trades—then the entire regulatory machine comes chasing after them. It sounds like a plot straight out of Kafka, but in the crypto world, it’s something that actually happened.

A judge in New York has paused the CFTC’s (U.S. Commodity Futures Trading Commission) civil lawsuit against a U.S. Army soldier. The soldier is accused of using insider information to place bets on Polymarket (a prediction market platform), involving an even more intriguing target—the outcome of hostage negotiations involving Venezuelan President Nicolás Maduro.

First, the amount. The complaint shows that the soldier earned more than $400,000 on the prediction market—nothing small for a soldier. It’s enough to change the course of a life.

Next, the logic. A soldier places a few bets on a prediction market, and now faces a civil suit from a federal agency. This isn’t just about a fine—it’s about the possibility that your life trajectory could be rewritten.

I’m really curious how this ends. The judge has stayed the case because a criminal investigation is ongoing at the same time. But who—or what—is the focus of that criminal investigation? Is it insider trading? Does a prediction market count as securities trading? Legally, there are no clear answers to these questions.

Most ironic of all, this prediction market space has long been watched by Wall Street. Compliance-forward prediction platforms have been fighting lawsuits to get legalized, and traditional financial giants have also been studying how to get in. If prediction markets become compliant in the future—if Polymarket gets absorbed or “recruited”—what would that leave as the basis for this soldier’s charges?

This isn’t an attempt to defend him. Profiting from official information is indeed a serious problem. But the issue is that rules in the crypto world change fast, while legal accountability replays like slow-motion footage. Today you’re sued because of a gray area; three years from now, that gray area may already have been legalized. Your stain remains, while the rules change.

For ordinary people in this kind of environment, the safest strategy isn’t to dig through legal text—it’s to understand one thing: in places where the rules are unclear, when the regulatory machine comes grinding over you, it won’t first ask whether you had malicious intent.

BTC is hovering around 64,000, and ETH around 1,880. The market has no mood to care about a soldier’s fate. But I still want to say: this case isn’t just one person’s trouble—it’s an awkward test run of regulation in a decentralized world. And in a test run, the easiest people to get hurt aren’t the driver—it’s the ones passing by on the roadside at that moment.

Tonight’s U.S. stock market looks pretty calm on the surface. The S&P 500 (U.S. stock index token) is down 0.14%, and the Nasdaq 100 (U.S. tech index token) is down 0.11%, but there are some interesting undercurrents beneath.

Intel (Intel, a chip company) is down as much as 4%. SK Hynix (Korean chip company), AMD (chip design company), and Micron (U.S. memory company) are all sliding together. On the other side, Microsoft (Microsoft, tech giant) is up 0.68%, and Meta (Facebook’s parent company) is up 0.79%.

Retail investors are panicking, while institutions are rebalancing.

With semiconductors selling off like this, it’s pretty clear someone is retreating. MicroStrategy (a company holding Bitcoin) is down nearly 3%, and Nvidia (AI chip leader) can’t stay green either—suggesting the money isn’t just hiding in risk aversion; it’s repositioning. From “hardware AI” to “software AI”—Microsoft has Copilot, Meta has open-source models, and the money is shifting toward the application layer.

I looked through the data and found it interesting that the 3x Nasdaq bullish ETF (TQQQ, a leveraged ETF token) is still up 0.08%. That suggests someone is buying the dip during the selloff—and using leverage. Retail investors may be watching semiconductor panic, while institutions may be watching the Nasdaq’s support levels.

Before the U.S. market opens on Monday, there are three things worth keeping an eye on. First is the futures open sentiment—if tonight’s weakness in semiconductors continues, the open may still see another leg down. Second is whether Microsoft and Meta can hold up; that will determine whether funds are only temporarily taking cover or truly switching the main theme. Third, Nvidia’s earnings expectations for Wednesday have already been “priced in,” so volatility this week won’t be small.

The biggest mistake ordinary people make when looking at the U.S. stock market is focusing only on up/down moves and ignoring the structure. The steepest drop could be institutions unloading inventory, while the most stable rise could be the formation of a new main storyline.

Monday’s direction doesn’t depend on who is down tonight—it depends on who holds steady first after the open.

When I first saw this news, my only thought was: here we go again.

The U.S. Department of the Treasury’s OFAC has sanctioned two more Iran-related crypto exchanges, citing money laundering facilitation, with an involved amount of roughly $5 million. The methods are nothing new—mixers, hop-by-hop wallets, and cross-border layered transfers, everything you’d expect.

So what does $5 million mean in the crypto world? It’s basically spare change for a medium-sized rug-pull project.

But that’s not the point. The point is that this kind of news shows up every few months, and the plot is nearly the same: a certain country gets sanctioned, a few exchanges get named, and some addresses are added to a blacklist. Then one group in the industry shouts, “Regulation is here, run!” Another group says, “Compliance is great, compliance is wonderful!” Two weeks later, the hype fades and everyone goes back to doing what they do.

What I’m curious about is something else: what, exactly, does this cycle change?

The addresses on the sanctions list just put on a new disguise and come back online. If a mixer gets shut down, three new ones appear. The people who truly use the crypto network to move funds across borders update their technology much faster than regulators do. But the ones who get hit are always ordinary users—your withdrawal suddenly gets frozen, your account is flagged for risk controls, and you may not even know why.

Some will say this is the necessary price. Compliance means sacrificing a bit of convenience.

The problem is that it’s we who sacrifice convenience, while the sanctioned parties don’t care at all about this “convenience.” They have technical teams, cold wallets, and well-structured money-laundering networks. Ordinary users have to worry whether their normal transfers on an exchange might be mistakenly targeted; meanwhile, they’ve already bypassed every centralized checkpoint.

I’m not against regulation. I just think that when sanctions turn into a performance-like political statement, it hurts not the bad guys the most—it hurts the people who stay quietly within the rules.

Today, BTC is hovering around 64,880 and ETH around 1,910. The market is already numb to this kind of news—prices don’t move at all. But don’t forget: every time sanctions tighten, the on-chain compliance threshold gets raised another layer. The cost isn’t paid today; it’s paid later, when one of your transfers gets stuck in the future.

I hope that when the time comes, the one who gets stuck isn’t you.