STi克莱小汤

If you want to make quick money from investment, the basic outcome is to give away money,
If you want to earn long-term value from investment, it often may bring surprises instead.
If you want others to take over your investment, you will ultimately be swallowed by even darker hands.
If you want to learn and share together through investment, it will be easier to accumulate value and merit.

In short, you must be true to your conscience; only then can some invisible forces be activated in the midst of it all. This is also an interesting deep-level philosophy, the way of balance.

$BTC $ETH $BNB #加密市场观察

Google's earnings report was ugly, but still within expectations. Let's see how the broader AI business develops going forward. Argon and Carbon still look pretty good. I have nothing but positive feelings about Google—just my personal opinion, not an investment view. I simply like this company's DNA 🧬, just as I like the blockchain and AI industries.

#股票财报季 #谷歌财报能否守住基本盘 $GOOGL

#AI The era, in fact, whether the content matters anymore #AIGC is no longer important, because AI is gradually gaining the ability to learn the writing style of high-quality authors. As long as the creator authorizes their own Agent to learn the content, then AIGC and MCP and other models are instead a means for creators to significantly reduce costs and increase efficiency. But for content platform operations and management, the most important thing next is:

1. ©️ Copyright maintenance and the establishment of reward-and-punishment mechanisms. Those who use AIGC must have content or data copyright, and must also have the awareness that they can be held accountable for content infringement.

2. An optimized mechanism for judging junk content. For junk content, cut off the traffic; for high-quality content, support the traffic (this requires dedicated staff). This point is crucial, because in the AI era, junk content can be produced in huge amounts. The source can instead start from supporting traffic: good creators will inevitably produce mostly in-depth content, be real-name verified, and in the future may also have blockchain ID verification. The more traffic is supported for high-quality content, the smaller the influence of junk content will naturally become. $BTC

3. A high-quality feedback mechanism. Do not blindly trust the judgment of agent bots, because the probability of hallucination and misjudgment is still too high. Improving feedback customer service that communicates with people is crucial. Reddit's recent financial report and data were a major disappointment precisely because its bizarre operational logic of rather mistakenly killing 1,000 than letting one bot go resulted in many high-quality new users being wrongly killed, and Reddit even does not have a human customer service team. This is truly astonishing. By comparison, people at the founder level at Binance are all on the front line arguing with netizens and absorbing feedback; the gap is obvious at a glance. Sometimes a product can rise because of the accumulation of the smallest details over time.

PS: I originally also wanted to test directly letting Claude and other Agent products use AI MCP to call #UniPublisher 's editing and distribution capabilities, so that the Agent could directly complete my idea, but I suddenly found that MCP had only connected long-form calls and not short-post mode yet, so I could only do it manually in the editor hahaha 💡.

Xiaohongshu has no traffic pool for showcasing mini programs,
while WeChat has no traffic pool for showcasing content, haha.
Binance Square lacks a curated pool of professional content, and its algorithm has become a bit more Twitter-like lately;
Twitter is a playground for the pros—advanced accounts get a great algorithm experience, while newcomers are basically in a graveyard, with zero traffic for quality curated content.

$BTC $ETH

Mbappé, the idiot—just when his form had finally started to improve, his brain short-circuited again. His character is nowhere near Ronaldo’s. Good thing he didn’t win the Ballon d’Or.$BTC

Claude's prospectus says that 5–7% of its revenue comes from banning accounts? Is that basically revenue from “robbery”? Can this kind of revenue really be openly disclosed in a prospectus? And isn't it illegal not to issue refunds after taking people's payments? I feel that even if an account is banned, refunds should still be prorated based on usage. But objectively speaking, this probably also includes amounts Claude API users had already used before being banned #AI #Claude $BTC $ETH $SOL

A PM’s number somehow ended up making this number—it looks so lucky, hahaha ᖰ⌯'▾'⌯ᖳ There are so many predictions for $BTC $ETH on #Polymarket ; they feel just like options.

If you use a Gate card and get banned afterward #Claude , the chargeback application fee could be higher than your subscription fee. 0.0 That means you’d actually lose money, so think carefully before deciding and using it. For example, if your $20 subscription gets banned, applying for a chargeback through their partner could cost $35–45 in fees. 💡 Absolutely unbelievable. I used Claude without a system-wide proxy for six months, then switched to paying with a Gate card and got banned on the very first day. 0.0 I’ve decided what to do; subsequent new accounts have been fine, though. But the fee is higher than the subscription itself—they even quoted $45 the first time and $35 the second. That’s so shameless. So they’re helping users with chargebacks but still trying to pocket $10 as a middleman fee?

I’m gradually realizing why so many people who were incredibly successful in Web2 have struggled to adapt to Web3. After recently experiencing AI products, their workflows, and their promotion and marketing channels, I can really see why. The way Web2 handles launches, user acquisition, and readership is completely different from Web3. #AI is relatively better off, since it’s used on both sides. But for other products, the resources are almost entirely disconnected. The launch and growth strategies that work really well in Web2 simply aren’t discoverable by Web3 users (and vice versa).

Successful Web2 executives come to #Web3 and still only know how to copy Web2’s successful channels, approaches, and strategies. Then they find that people on both sides can’t see each other, and the approach doesn’t fit. I used to just feel that some projects with Web2 executives parachuted in were really strange. Now I’m gradually realizing the root cause. Maybe it’s not that they’re incapable; it’s that their past hits and paths to success are completely different from how things work now. $BTC $ALGO $ETH

Seeing this celestial phenomenon and unexpectedly catching a performance by the legendary Dutch 🇳🇱 jazz band Dutch Swing College, haha. Will $BTC get better?

In the AI era, copyright ©️ agents are going to be in high demand. I’ve noticed quite a few people using AI to generate article after article by copying whole paragraphs from my old posts 😂—without even citing the source. Once commercial activity is involved, this could become a very dangerous situation. Many people may get burned by AI without even realizing it. $BTC $BNB $ETH

#币安神算子 Isn’t this a new event? Where did all these people with so many points come from? Could it be that points earned before #Predict were included? What’s the point of playing then? If you’re running a points event, everyone’s points should be reset to zero so we can start from claiming check-in rewards.

Binance Square is starting to remind me of Bihu back in the day. It’s a real shame what happened to Bihu in its later days—they just failed to crack down on garbage posts and low-effort fluff. That drove away a lot of quality creators. A thoughtful, in-depth article might be thousands of words long, but it couldn’t compete for traffic; split it into 10 shallow posts and it would get more traffic.

Back then, I told people around me that Bihu wasn’t far from shutting down. Later, Bihu tried to save itself and came very close to turning things around, but Gulu didn’t want to keep going.

Square was really good when it first started, but lately it’s been getting stranger and stranger.

Anyway, that’s enough. There’s no point saying more.

$BTC $ETH

My connection with $TSLA goes back to 2017, when I helped manage the charging and power systems for the development of the Model 3. Tesla’s lead R&D engineer struck me as someone who was extremely passionate about his work and highly professional. I personally wasn’t very interested in mechanical R&D, even though I had published three first-author SCI papers and, by the time I earned my master’s degree, met the standard for a PhD graduate.

But I really disliked technical drawings. What caught my interest instead were Tesla’s financial reports and the secondary market. So if we’re talking about crypto, my connection with $BTC really began with Tesla. It was also thanks to that collaboration that, around 2019, after the fires dealt the company a major blow, I wrote to several large institutions saying that I believed Tesla had tenfold upside.

Of course, some institutions took it as a joke, especially some domestic ones, which may have been rather arrogant in their mockery. But one top-tier American investment bank replied very seriously and discussed it with me. At the time, I felt they treated opportunities with great care. You can see the same thing in venture capital: European and American investors seem more willing to get hands-on and try new things, while some Chinese investors prefer to flock in once success is clearly within sight. That’s why I feel some Chinese institutions are pretty average at going from 0 to 1 (with the exception of state-backed institutions), but very good at going from 1 to 2.

But against the broader backdrop, our country’s pace of development has already been a miracle. I believe things will get better and better, and that our openness to innovation and spirit of exploration will keep improving too.

I vaguely remember what a well-known financial adviser said to me in 2018: “Why should I bring on a career-switching newcomer who’s just returned from overseas? That’s too risky. I’d prefer someone with a finance background.” (That company has since gone bust.)

By contrast, the exact words of the European-region managing director at the time—who is now the CEO of a top American bank—were: “You have experience and judgment from mechanical R&D that people who studied finance and investment will never have. As for financial and investment knowledge, if you’re interested, two years of self-study will be enough. Don’t ever give up on pursuing what you want. If you want to go to JP in the UK, you can list me as a reference.”

Those words subtly but truly showed two completely different attitudes toward risk. I can’t say who was right or wrong—after all, there are plenty of charlatans here at home too. But the two people clearly seemed to be on completely different levels.

#股票财报季

This is a very important day. Let's give it our all together ⛽ and break through!

#股票财报季 Stocks sometimes lose a lot of their upside potential because of earnings reports, but having those reports often gives us more to think about when positioning ourselves early.

That’s because the data is delayed. For example, the premium memberships driven by A Record of a Mortal’s Journey to Immortality may not show up until Bilibili’s Q4 earnings report. So Bilibili’s stock tends to rise later, with dips along the way.

This is a crucial opportunity for Bilibili this year—though of course, we don’t know whether BLG will drop the ball.

$BILI.US $NVDAB

There are two kinds of people who build products and invest:

One kind does it purely to make money; the other does it to improve their own efficiency or for fun, sharing it with others along the way. If it attracts a large user base, that's a matter of fate; if it doesn't become wildly popular, having some high-quality users and institutions is still great. It's purely about improving one's own efficiency.

My friends and I are the latter. A product can improve our own work efficiency by around 50%; several products together can boost it severalfold, or help us learn new methods. In this situation, it doesn't really matter how many people use them; it's more about practicing the process. It's not a negative-cash-flow model or one that burns through money recklessly. As for investment views, styles, and approaches, I share my judgments months in advance. Whether people believe them before the trend arrives is a matter of their own judgment. It doesn't affect the short-term line, and therefore doesn't affect the overall trend or the changes in the Dao.

Many VCs suffer serious losses because they always like to look at decks full of hype, indulging the dopamine rush of their own fantasies, while being more wary of decks for products that have already been completed. If a VC keeps losing money, perhaps it's time to reflect on its decision-making process.

$BNB $NVDAB $BTC #Aİ