坚定的持有勇敢的操作

$Echostar(SATS.US) With such high trading volume, is it aggressive accumulation or distribution?

$SpaceX(SPCX.US) Only got two shares in the IPO subscription, let's see the strength of the universe's number one stock.

Recently, it feels like everyone in the market is in a state of FOMO!

If this one doesn't go up, they chase into the next one that does without stopping

I think this is the time to calm down and read some books

Recently, it seems that being a relay station (or intermediary) is particularly popular. After joining an AI group, I found a bunch of people are doing relay station business. With Justin Sun and Trump's son now also entering the field, this track has become overly crowded. What I can currently think of to still make money in this track is providing stable account pools and integrated deployment.

Actually, there's still an untapped blue ocean, which is private deployment for enterprises. All corporate data stays within the company, and the enterprise has sufficient authority to control what data goes to the cloud and what data remains on-premises. At the same time, help enterprises customize some content AI software. Because on the toC side, various products are emerging endlessly, almost all of which are software similar to manus created by nesting large models, so focusing on privatization must be a good path. This route relies on cheap + high-performance chips and increasingly powerful open-source large models.

$NVIDIA(NVDA.US) $Invesco QQQ Trust(QQQ.US) $Sandisk(SNDK.US) $Tesla(TSLA.US) $AMD(AMD.US) $S&P 500(.SPX.US)

$Destiny Tech100(DXYZ.US)The expectation for SpaceX's listing remains, with the IPO estimated to be in June. As long as the IPO timing hasn't been announced, the expectation persists, and the upward logic remains unchanged.

Recently, when I read books, I open Claude in a side window, tell it I'm reading a certain book, and then ask it to accompany me while I study. The method I use is the Feynman technique, where AI quizzes me on knowledge points, I answer, and finally AI scores and summarizes.

Take a recent book "Illustrated Guide to Large Models" as an example. I highly recommend everyone to read this book. If I had read it earlier, I might have been able to catch companies in areas like storage and data annotation. After reading it, you'll have a clear understanding of the essence of large language models, learn how AI companies typically train them, and understand why GPUs are so crucial.

@Value & Investment Value Brother once said that the best investment is investing in your own brain. When you thoroughly grasp a fundamental concept, you can use first principles to analyze any new development that appears. As for what to invest in, such questions become easily solvable. The book was published in September 2024, long before the explosion in storage, chips, and GPUs. So don't wait any longer, start arming your brain now!

$Micron Tech(MU.US)$Invesco QQQ Trust(QQQ.US)$Sandisk(SNDK.US)$NVIDIA(NVDA.US)

Going to Berkshire to hide from the bull market$Berkshire Hathaway B(BRK.B.US)

Because I don't really like using my phone, I haven't been posting. Today, I finally found the Longbridge web posting tool. Next, I'll share on Longbridge how, as a Gen Z, I usually use AI in my daily life, and how I use AI to empower my trading.

The first share today is a super useful tool, Claude + Obsidian. If you don't have the conditions, you can use Claude Code and then switch to a domestic large model.

First, Obsidian, as a Markdown document management tool, has an inherent advantage in combining with AI. At the same time, Obsidian's plugin marketplace allows it to achieve almost the same level of functionality as Notion. Compared to Notion, a huge benefit is its local nature, allowing Claude Code to directly read and write local files. I used to use these tools because I was lazy, but lazy me didn't even want to spend time setting up various tables and links. Now, with AI, it can be done with one click. Just tell the AI to create an Obsidian Vault to help you record your trades. After running for over ten minutes, a fully functional vault is ready.

Next, I'll share the structure of my Vault, which includes:

- Roadmap

- Strategy

- Risk Management

- Trading Records

- Research

- Knowledge Base

- Daily Notes (mainly for recording mood)

- Document Templates (a template library that can be directly called when creating)

- Kanban Board

Claude.md (This is very important because it allows Claude to know the purpose and function of this vault every time it's reopened.)

All of these can be built by AI. Friends who need it can comment below, and I can turn this vault into a skill. You can then have your AI use this skill to get the same template as mine.

Finally, I want to say that it's important to find a balance between building and outputting. Try to let AI manage things, but don't spend too much time on designing the knowledge base, losing sight of the primary goal. Also, ensure daily recording, every trade, every research session, every summary, and learning. I will also create a script later to directly obtain asset allocation through Longbridge MCP (thanks here to @LongbridgeAI )

$Tesla(TSLA.US) $NVIDIA(NVDA.US) $Invesco QQQ Trust(QQQ.US) $Sandisk(SNDK.US)

I did well with other stocks, but with this PLTR, because I didn't know much about it, I saw a bearish comment from a PLTR influencer on the square a few days ago, plus the price pullback, so I sold. I didn't expect it to skyrocket in the next two days, making me dizzy. It seems that no matter what you do, doing your research and believing in yourself is the most important thing. Stick to your own trades because everyone's way of making money is different.

$Palantir Tech(PLTR.US)