MoonMan567

In June, crypto hackers stole at least $75.9 million in 40 major attacks. That’s a bit less than in May, but there’s a catch.

A smaller amount doesn’t mean the industry has become safer.

The biggest blow hit Humanity Protocol. Initially, losses were estimated at $31 million, but the project team later said the real damage could have been around $36 million. The cause was cited as a compromise of the private key. The theory that North Korean hackers were involved is still unconfirmed.

The second-largest incident was the breach of the Syscoin Bridge, where an attacker was able to create unsecured tokens due to a validation error.

My observation: increasingly, the biggest losses are not caused by “genius” attacks, but by old problems—stolen keys, logic bugs in smart contracts, or insufficient code verification. In crypto, the most expensive vulnerability is still the human factor.

If you’re interested in analyzing the crypto market without the information noise—follow @MoonMan567

The United States is taking another step toward a technological “iron curtain.”

A group of Republican senators has introduced a bill that would allow the Department of Commerce to block deals involving information technology and digital services if they are linked to countries that Washington considers risky.

This is not only about equipment. The document could also apply to software, cloud services, telecommunications solutions, and other digital products if they are capable of affecting critical infrastructure or national security.

If the law is passed, the U.S. government will gain even more tools to control which technologies can enter the American market and who will have access to them.

We increasingly talk about trade wars, but in reality the world is already shifting to technological wars. And the main currency in them is not dollars, but access to chips, cloud infrastructure, artificial intelligence, and software.

If you’re interested in unpacking major technology changes without the information noise—follow @MoonMan567

€4.1 billion fine for Google - it's no longer just a punishment. It's a signal to the entire tech industry.

The EU Court has finally confirmed the decision in a case that has been ongoing since 2018. The regulator concluded that Google used Android not only as an operating system, but also as a tool to promote its own services and push out competitors.

And the point here isn’t even the amount.

Europe once again shows that it is ready to take the biggest tech companies to court for years if it believes they are abusing their position. And more importantly, to see these cases through to the end.

My observation: ten years ago, it seemed that the largest IT corporations were effectively unreachable for governments. Today, the situation has changed. The bigger a tech company becomes, the higher the chance that its next “product” will be a multi-year court battle with regulators.

If you’re interested in dissecting major tech changes without the information noise - follow @MoonMan567

Fresh macro data from the US made crypto bulls start screaming again about an imminent rate cut by the Fed and an inevitable rally.

The June jobs report was indeed weak: the number of new jobs (Nonfarm Payrolls) came in at just 57 thousand versus the expected 114 thousand. In addition, the May and April figures were revised downward by a total of 74 thousand. At the same time, the unemployment rate unexpectedly fell to 4.2% (the forecast was 4.3%).

Traders have already interpreted this as a “moderately bullish” signal for $BTC , expecting that easing Fed policy will push liquidity into risk assets. But don’t rush to open longs with your whole bankroll.

A weak labor market is not just a reason to cut the rate—it’s a direct indicator of economic cooling and recession risk. And during recessions, big capital flees into cash rather than altcoins.

Falling unemployment alongside weak hiring, in general, suggests that people are simply dropping out of the labor force. There’s nothing to celebrate here—the market is squeezed between the jaws of inflation and stagnation.

If you want sober analysis without amateur calls to buy every macro headline—follow @MoonMan567

Not long ago, spot ETFs were called the main driver of growth $BTC .

Now this same mechanism has started working in reverse.

In June 2026, investors pulled $4.06 billion out of spot Bitcoin ETFs—that’s the largest monthly outflow since their launch. The previous record of $3.56 billion held for a little over a year.

But it’s important not to overstate it.

Outflows from ETFs don’t prove that institutional investors have massively abandoned Bitcoin. There could be many reasons: risk reduction at quarter-end, rebalancing portfolios, or taking losses. Without data from the funds themselves, claiming more would be inaccurate.

The market often celebrates every billion that flows into an ETF, but it gets very nervous when the same money leaves. In reality, ETFs are only a channel for capital flows. They amplify both optimism and fear equally well. That’s why it’s worth watching not only the price, but also who today is willing to hold that risk.

If you’re interested in analyzing the crypto market without information noise—subscribe to @MoonMan567

Another on-chain chart hints that $BTC may be approaching the bottom-formation zone. But the word “may” here is more important than the chart itself.

CryptoQuant’s UTXO Profit/Loss Ratio metric has fallen to levels that, in previous cycles, often coincided with the phase of seller exhaustion. In simple terms: more and more coins are moving into loss, and opportunities for mass profit-taking are becoming fewer.

That said, history shows something else, too. Such signals rarely mean that the bottom has already been formed. Often, the market still moves for a few more weeks—or even months—in a wide range before a sustained trend appears.

My observation: the biggest mistake investors make is treating on-chain metrics as a crystal ball. In reality, they’re more like a car’s dashboard: they show the system’s condition, but they can’t tell you what road the market will take tomorrow.

If you’re interested in breaking down on-chain data without magic, loud predictions, and information noise—follow @MoonMan567

A month ago, the United States effectively stated: the most powerful AI could be more dangerous than some dual-use technologies.
That is why the export of Claude Fable 5 and Mythos 5 was blocked on national security grounds.

Now, the ban has been lifted.

The very fact of such a sharp reversal says more than the decision itself.

We’re used to the idea that governments control the export of weapons, microchips, or nuclear technologies. But now, in effect, advanced AI models have been added to that list.

And this is probably the most important change.

AI models are no longer just a product of technology companies. Gradually, they become a strategic resource—access to which increasingly determines not the market, but the state.

It seems we’re entering an era in which the main advantage won’t be who created the smartest AI, but who is allowed to use it at all. And this is a far more serious turn than yet another presentation of a new model.

If you’re interested in unpacking major technological changes without the information noise—follow @MoonMan567

While U.S. President Donald Trump declares billion-dollar royalties from memecoins and stablecoins, his right-hand man isn’t falling behind

In his recent financial disclosure, Vice President JD Vance officially confirmed that he holds on Coinbase the main digital asset — $BTC — valued between $250,000 and $500,000. Compared to previous years, his position has effectively doubled

Of course, against the backdrop of мільярдів шефа, Vance’s quarter-million dollars looks like a mere deposit. But the key point here is the precedent: for the first time in U.S. history, the first people in the White House have a direct, open, and very personal stake in the crypto market growing.

All these government regulatory relaxations and stablecoin laws have now taken on brand-new colors. They’re literally passing laws for their own wallets, without even trying to hide it

🤔 Conflict of interest? No, we haven’t heard of that. Just effective management of personal capital at the highest level of government

If you want to see the real underside of American politics and know which laws are being lobbied for their own crypto-wallets in Washington — subscribe to @MoonMan567

Answer honestly: how many AI subscriptions do you have currently active—and how many of them did you open this week? Twenty dollars here, another twenty there, all quietly charging to your card while you’re signing into one—if at all, and only rarely. A subscription is rent for the possibility, not payment for usage. And the most annoying part is that this possibility sits idle while you keep paying for it.

@OpenGradient did the opposite—removed the subscription altogether. In OpenGradient Chat, you pay for what you actually run: a thousand credits costs a dollar. The price is set at the model’s cost, with no markup. One question for the front-end model—cents, not monthly rent. Credits don’t expire: you buy once and use when you need to, not until the month ends. $OPG is the token that the OpenGradient network uses to pay for the inference itself—meaning you pay for the work, not for the unlocked doors.

I don’t pretend that this is always cheaper. If you run AI all day, a subscription might still be more cost-effective. The difference is this: the subscription charges you for the days you’re NOT using it, while usage-based payment doesn’t. OpenGradient simply removed the charge for idling.

See what it’s really like: chat.opengradient.ai—at the start, credits are free. #opg

How many of your monthly subscriptions are actually worth it, if you only paid for real usage?

Friends, if you know what the funding rate is (Funding Rate) and how to profit from extreme price deviations (Short / Long Squeeze), then here are the results of the market scan by my new AI agent.

Today, the market offers us the following interesting OPPORTUNITIES:

Filter: FR > 0.06% + Vol > $20M

+ LONG opportunities
1. $SKHYNIX 0.2880% Vol: 713M (7h 59m)

2. SAMSUNG 0.2429% Vol: 48M (7h 59m)

- SHORT opportunities
1. $POWR -1.1532% Vol: 101M (3h 59m)

2. RE -0.4472% Vol: 194M (3h 59m)
3. RIF -0.3658% Vol: 32M (3h 59m)
4. LAB -0.2858% Vol: 280M (59m)
5. MAGIC -0.2431% Vol: 50M (3h 59m)
6. AGLD -0.2151% Vol: 53M (7h 59m)
7. ACT -0.2002% Vol: 366M (3h 59m)
8. CELO -0.1547% Vol: 34M (7h 59m)

If you’re hearing about Short or Long Squeeze for the first time, then I recommend reading this статтю.

Run a test: ask your AI assistant to describe you based on everything you’ve ever asked. Many people do this now, and it turns out pale—the portrait is more accurate than what your loved ones know. But the shock isn’t in the accuracy. The shock is that you didn’t consent to this. You asked separate questions on separate days, thinking they were separate conversations. But they’ve been stitched into one dossier the whole time: fears, money, health—an identity that lives where you can’t control it.

Here’s the difference in approach @OpenGradient . There, your identity is cut off from the request before the model even sees it—it's not stored; it’s severed. There’s nothing to stitch together: there’s no name to attach your questions to. 170 thousand inferences have already passed through the TEE enclave, where even OpenGradient itself can’t see who is behind the request.

I’m not saying other services are villains. The profile grows not out of malice, but because that’s how the system works: it’s profitable to remember. The difference with OpenGradient is that the ability to stitch you into a dossier is removed architecturally—and when you pay for $OPG , you’re paying for the OpenGradient computer itself, not the right to keep you in memory. It’s easy to verify: the same question on chat.opengradient.ai #opg

If your AI already knows you better than your relatives, the only thing left to decide is—who does this copy of you belong to?

Another cryptobreakin has shown one unpleasant thing: sometimes the biggest threat is not new code, but the one everyone has already forgotten.

The hacker stole about $2.1 million from Aztec Connect—a DeFi solution based on $ETH , the support for which was discontinued back in 2023. At the same time, the current Aztec Network was not affected.

The reason for the attack is also revealing. The attacker exploited a mismatch between how the contract verified transactions and how it then executed them on Ethereum. As a result, the system credited unsecured balances, after which the funds could be withdrawn.

Most interestingly, the Aztec team couldn’t intervene. The contracts can’t be changed, there are no administrative keys—so it was impossible to pause the system.

My observation: decentralization always has two sides. It protects users from developers’ arbitrary actions, but at the same time it deprives them of the ability to quickly fix a bug if one is still left in the code. That’s why, in a blockchain, even a “dead” contract can remain a live target for years.

If you’re interested in analyzing the crypto market without information noise—follow @MoonMan567

Not long ago, the most powerful AI models competed with each other for the quality of their answers.

Now they’re starting to compete for… the right to be released.

The United States allowed Anthropic to restore access to Claude Mythos 5, but only for more than 100 verified American organizations, including companies and operators of critical infrastructure. Almost at the same time, OpenAI postponed a full launch of GPT-5.6, leaving the model available only to a limited circle of partners.

Formally, the reason is national security.

But the significance of this story is much broader. If earlier governments controlled the export of microchips, then now the same kind of control is gradually being extended to the most powerful artificial intelligence models as well.

We’re living less and less in a world where AI is simply a commercial product. It increasingly resembles a strategic resource—like nuclear technologies or modern semiconductors. And it seems that access to such models will soon become a new form of geopolitical advantage.

If you’re interested in unpacking major technological changes without the information noise—follow @MoonMan567

There’s a question you’ve been turning over in your head for a week, but you’re afraid to say out loud—because the wrong person might hear it. Someone has lost their family savings and doesn’t know how to admit it to their relatives. Someone carries a diagnosis, a debt, or a failure that their household doesn’t even know about. Irony: the only one who will listen at 3 a.m. without judgment is an AI assistant. And that’s exactly who you’re afraid to message—what if the conversation shows up somewhere your people can see it.

A private layer here isn’t an abstraction—it’s what removes fear. On @OpenGradient , your request is encrypted on the device, and the identity is severed before the model sees it. 170 thousand private inferences have already passed through a TEE enclave, where even OpenGradient can’t see what’s behind the request. You can send it—not to replace your loved ones with AI, but to find the words that will let you say it to a living person.

I’m not selling a miracle. OpenGradient won’t fix the loss and it won’t have the conversation for you—it only gives you a place to think out loud until you’re ready to say it to anyone. The first step toward a difficult conversation is to rehearse it somewhere where nobody can recognize you.

Try OpenGradient: chat.opengradient.ai, the first queries are free. $OPG pays for this computer, not access to you. #opg

Why are we used to paying for a judgment-free space with the fact that someone reads every word?

Over 10.8 million $BTC are currently below the price of their most recent on-chain movement. According to Glassnode, this is the highest reading in the entire period of observation.

At first glance, this sounds like very bad news. But there’s an important nuance.

The metric doesn’t show how many investors are at a loss. It shows how many coins were last moved at prices higher than the current one. One large wallet can contain thousands of such $BTC .

Historically, similar periods have often been accompanied by increased volatility. The logic is simple: the more coins are “in the red,” the stronger participants’ emotions in the market may be—both during new waves of selling and during sharp rebounds.

My observation: the market tests investors the most not when the price is falling, but when more and more people are forced to decide what matters more—locking in a loss or waiting. These are the moments that often determine the next chapter of market history.

If you want to explore on-chain metrics without loud headlines—follow @MoonMan567

Friends, if you know what a funding rate (Funding Rate) is and how to profit from extreme price deviations (Short / Long Squeeze), then here are the results of market scanning by my new AI agent.

Today, the market offers us these interesting OPPORTUNITIES:

Filter: FR > 0.06% + Vol > $20M

+ LONG opportunities
1. $CBRS 0.4789% Vol: 111M (7h 22m)

- SHORT opportunities
1. $AGLD -0.8810% Vol: 551M (7h 22m)
2. $ARK -0.7196% Vol: 32M (3h 22m)
3. PORTAL -0.3931% Vol: 34M (3h 22m)
4. LAB -0.3755% Vol: 399M (22m)
5. RE -0.2283% Vol: 173M (3h 22m)
6. BICO -0.1932% Vol: 23M (7h 22m)
7. PUNDIX -0.1539% Vol: 84M (3h 22m)
8. TNSR -0.1431% Vol: 28M (3h 22m)

If you’re hearing about Short or Long Squeeze for the first time, I recommend reading this статтю

I opened the dashboard @OpenGradient and saw numbers that most AI-crypto projects won’t show: 894.6 thousand inference transactions, 4449 models, and over 1.67 million blocks. Not a landing page with promises, but the OpenGradient network that actually works.

But I’m cross-checking the euphoria with the source. On OpenGradient’s own website, the message still says: Testnet Is Live. That means the activity is still in test mode. And here’s the question that hype threads avoid: test activity doesn’t equal steady paid demand. Racking up 895 thousand transactions on a testnet where incentives encourage activity is one thing. Sustaining them when every call is paid for for real is entirely different.

And this isn’t emptiness—quite the opposite. Next to it is live monetization: almost 170 thousand private inferences in the OpenGradient Chat, 4641 $OPG has already been spent on compute. People are paying for a product today. The question is narrower: how much of the network’s numbers will survive production, and how much was activity for the sake of activity.

See for yourself: chat.opengradient.ai, dashboard is open, numbers are public. $OPG inside it is payment for real compute, not a hype ticket. #opg

The question is open: when a project shows big numbers on testnet, are you looking at proof of demand—or a demand rehearsal you still have to play to the fullest in real life?