One of my biggest takeaways was the discussion on open-source AI models and memory demand. I hadn't fully considered that larger context windows and always-on reasoning could actually increase memory requirements. It was a good reminder that greater AI efficiency doesn't necessarily mean lower demand for advanced memory. The comparison with Kimi K3 also gave me a different perspective on the AI memory cycle. If open-source models continue gaining adoption, demand for HBM and high-performance DRAM could remain stronger than many investors currently expect. That supports my long-term bullish view on the memory sector. As an investor, I try to focus on structural trends rather than short-term market reactions. This session reinforced my conviction to stay patient, keep following the fundamen

This session strengthened my conviction that the AI memory story is far from over. What stood out was why this cycle differs from previous ones. Instead of short-term speculation, demand is being driven by structural AI adoption and rising memory requirements across hyperscale data centers. I also found the discussion on HBM and DDR5 supply very insightful. Advanced memory capacity takes years to expand, and if demand continues to exceed supply, leading memory companies could benefit for much longer than many investors expect. As an investor, this reinforces my focus on long-term fundamentals over short-term price swings. If the underlying demand story remains intact, market corrections can become opportunities to build positions with greater conviction.

One of my biggest takeaways is that this looks like a healthy correction rather than the end of the AI memory cycle. The discussion on fund flows, inventory expectations, and market sentiment helped me understand why memory stocks have pulled back despite strong long-term AI demand. It reinforced my view that short-term price action doesn't always reflect long-term fundamentals. I also found the explanation of the HBM supply gap very valuable. AI infrastructure demand continues to grow, while advanced memory capacity still takes time to expand. That gives me confidence in the long-term outlook for leading memory companies. As an investor, I focus more on industry cycles than daily market noise. This session reminded me to stay patient, monitor the fundamentals, and use corrections to stre

I bought the dip instead of reducing my exposure. One weak session doesn't change my long-term thesis. To me, this was more of a valuation reset than a collapse in AI demand. I still believe enterprise AI and hyperscaler spending have plenty of room to grow. Corrections like this can also create opportunities to accumulate quality companies at better prices. I'm becoming more selective, focusing on semiconductor & AI infrastructure companies with strong demand, visible orders, and improving cash flow. I continue to DCA into my highest-conviction positions instead of reacting to short-term volatility. Risk management remains important, so I'm keeping my position sizes under control. Over the next few months, I'll watch whether higher AI capex translates into stronger revenue and free c

$ServiceNow(NOW)$ ServiceNow's recent pullback hasn't changed my long-term conviction. In fact, it has given me another opportunity to continue dollar-cost averaging into a company that I believe is becoming one of the biggest beneficiaries of enterprise AI. While short-term market sentiment has turned cautious because of valuation concerns and broader rotation within technology stocks, I think the market is underestimating how deeply AI is being embedded into enterprise workflows. As long as the business fundamentals remain intact, I see volatility as an opportunity rather than a reason to panic. What gives me confidence is that ServiceNow is no longer just an IT service management company. It is evolving into an AI-powered enterprise platfor

The point that resonated with me most was the distinction between a short-term top and the end of a long-term bull market. Many investors mistake every correction as the collapse of the AI thesis, but valuation resets are a normal part of secular growth cycles. That matches how I've been viewing the recent semiconductor pullback. I also found the discussion on the HBM supply gap insightful. As AI models continue to grow, advanced memory remains a key bottleneck. Despite short-term volatility, I believe the long-term demand for HBM and high-performance DRAM remains strong. My biggest takeaway is to separate price action from business fundamentals. Rather than reacting emotionally, I prefer to focus on industry trends and supply-demand dynamics. Corrections often create opportunities for pa

Congratulations on achieving such an impressive return in just a few months. I also appreciate your honesty about the July drawdown. Sharing the importance of stop-losses and position management makes your experience much more relatable and valuable. I found your use of ChatGPT and Claude for research especially interesting. AI can greatly improve research efficiency, but discipline and execution are still what determine long-term success. I also like your focus on staying within your circle of competence instead of chasing every trending sector. My question is: @Yongjian Zhang if the semiconductor cycle weakens more than expe

I’m voting bearish, expecting $Intel(INTC)$ to finish down around 5% to 10% after earnings. The stock has rallied sharply this year, raising expectations significantly. Even a decent quarter may disappoint if management doesn't provide stronger guidance or confidence for the second half. My focus will be on Intel’s 18A foundry progress and data center business. Server CPU demand has improved, but investors want proof that AI investments can drive sustainable profits, better margins, and meaningful external foundry customers. Any disappointment could trigger selling. I remain positive on Intel’s long-term turnaround, but I think this earnings report could become a "sell the news" event. With expectations running high, my prediction is a 5%-10% pu

$SpaceX(SPCX)$ SpaceX has pulled back to its lowest price since the IPO, and I decided to start accumulating a small position. This isn't one of my core holdings, nor am I betting the farm on it. Think of it as a "meme trade" with asymmetric upside—a small allocation that could deliver outsized returns if the long-term thesis plays out. The biggest reason is simple: I believe in Elon Musk's ability to execute ambitious visions. Whether it's Tesla transforming the EV industry or SpaceX revolutionizing reusable rockets, he has repeatedly proven that he can achieve what many once considered impossible. Starlink continues to expand globally, Starship is making steady progress, and the company's leadership in the commercial space industry remains

$Direxion Daily Semiconductors Bull 3x Shares(SOXL)$ Every time SOXL experiences a sharp correction, I see the same question being asked: "Is the AI trade over?" My answer is usually the opposite. Corrections are a normal part of investing, especially in one of the market's most volatile sectors. Instead of chasing rallies, I prefer adding during periods of fear. My confidence comes from one simple belief: semiconductor demand is still in a long-term uptrend. AI data centers, cloud computing, autonomous driving, robotics, and edge devices all require more advanced chips. While individual companies may go through different cycles, I believe the industry as a whole will continue to benefit from these structural trends. That's why I use SOXL to

$Palantir Technologies Inc.(PLTR)$ Palantir remains one of my highest-conviction long-term holdings, and every pullback is an opportunity for me to keep collecting. My investment thesis has never been about chasing short-term momentum—it is about owning a company that is becoming critical AI infrastructure for governments and enterprises. The main reason I stay bullish is Palantir's Artificial Intelligence Platform (AIP). Rather than simply offering AI models, Palantir helps organizations deploy AI into real-world operations across industries such as manufacturing, healthcare, defense, and finance. As enterprise AI adoption accelerates, I believe demand for platforms that securely integrate AI with business data will continue to grow. Palanti

I'm bullish on $Tesla Motors(TSLA)$ going into earnings, but I'm looking beyond the delivery numbers. Q2 deliveries were solid, and the bigger question is whether margins have stabilized and free cash flow is improving after the recent period of price pressure. A positive surprise on either metric could help rebuild investor confidence. What I'm most interested in is management's update on Robotaxi, FSD and Optimus. These businesses will have a much bigger impact on Tesla's long-term valuation than quarterly vehicle sales. Any progress on commercialization or production timelines could become the biggest catalyst for the stock. I'm prepared for short-term volati

For me, $Alphabet(GOOGL)$ Google Cloud is the key metric. TSMC and ASML have confirmed strong AI infrastructure demand, but Alphabet now needs to prove that its heavy AI investment is translating into sustainable Cloud revenue and healthy margins. I'm also watching AI Search monetization. It's not enough to increase user engagement—Google must maintain advertising revenue while keeping AI inference costs under control. That will determine whether AI strengthens its core business. Overall, I think Alphabet is becoming a leading end-to-end AI platform. If it delivers strong Cloud growth, solid AI monetization and confidence in free cash flow despite high capex, the long-term AI story becomes much more convincing. I’ll also be listening closely to

$ARM Holdings(ARM)$ The recent semiconductor correction has been uncomfortable, but for me it has also created an opportunity. Instead of trying to predict the exact bottom, I've chosen to dollar-cost average (DCA) into ARM during this pullback. The recent selloff has been driven more by concerns over AI capital spending, higher interest rates, and broad deleveraging than by any major deterioration in ARM's long-term business. When quality companies are sold alongside the rest of the sector, I see that as a chance to gradually build a position. What continues to attract me to ARM is its unique business model and dominant ecosystem. Unlike traditional chip manufacturers, ARM licenses its architecture and collects royalties from billions of devi

I've been holding the Singapore banks for a while, and they've been one of the steadiest performers in my portfolio this year. Personally, I prefer $DBS(D05.SI)$ for its strong franchise, consistent dividend growth, and leading digital banking platform. While the stocks are at record highs, I believe the rally is supported by improving fundamentals rather than just market optimism. That said, I'm not chasing prices after such a strong run. Valuations are definitely higher now, so I'd rather wait for pullbacks or volatility to add more. If earnings continue to surprise through stronger net interest income, wealth management fees, and healthy loan growth, I think the long-term trend remains intact. For me, Singapore's big three banks are core lon

$ServiceNow(NOW)$ The recent correction in ServiceNow (NOW) has only strengthened my conviction, which is why I'm continuing to dollar-cost average into the stock. Could it drift lower in the short term? Of course. But after the latest pullback, I believe the market has become a little too pessimistic. For a company with ServiceNow's quality and long-term growth potential, this correction looks slightly overdone to me. My bullish view is centered on the long-term outlook for enterprise software. As businesses accelerate digital transformation and AI adoption, software platforms that improve productivity and automate workflows should remain critical spending priorities. ServiceNow has consistently expanded beyond IT service management into AI-p

$Direxion Daily MU Bull 2X Shares(MUU)$ The recent selloff in MUU, the leveraged ETF tracking Micron, gave me the opportunity to open a new position. Could the weakness continue? Definitely. Leveraged ETFs are inherently volatile, and trying to catch the exact bottom is nearly impossible. But after such a sharp correction, I believe much of the short-term pessimism has already been reflected in the price, making the risk-reward increasingly attractive. What gives me confidence is that Micron's long-term fundamentals remain solid. The AI revolution continues to drive demand for high-bandwidth memory (HBM), DRAM, and NAND, while hyperscalers are still investing heavily in AI infrastructure. Memory pricing may experience short-term fluctuations,