Powerpei

Tried buying MUOn and WDCon on Binance Wallet That slippage instantly brought me back to reality Honestly, packaging US stock assets or AI computing power into tokens and putting them on-chain This is just the first step, and probably the easiest one The real challenge is: how to make the liquidity depth of these assets no longer a dead end If buying $500 costs you several points in loss Then this kind of RWA interaction is simply not attractive for high-frequency capital I just gave some extremely blunt but honest feedback to the backend: First, don’t just focus on increasing supply Although there are more targets now (AI infrastructure, traditional finance, etc.), if small trades can’t be executed efficiently, this is just a showcase, not a trading floor Second, about dividend and audit transparency What we need is verifiable real-time settlement, not waiting until the end of the month to see a PDF signed by who knows whom. If I can’t verify the health of the underlying assets on-chain in real time, then this trust still hangs on centralized institutions’ words, not engraved in the contract. In the current RWA track, the trade-off between compliance and user experience has reached a critical point. Binance’s current strategic layout is quite significant, but the iteration speed of the Alpha version must be accelerated In this space, good tools don’t need hype, the data speaks for itself Go check their newly updated docs first and see if they’ve plugged these fatal holes. @binancezh #RWA

Waiting for delivery on Shennan Avenue this morning, I checked Jin10 data SK Hynix (000660) surged 10% today Its market cap officially crossed the $1 trillion mark Plus yesterday Micron (MU) had a 19% bullish candle on Nasdaq Global capital is basically going crazy over HBM (High Bandwidth Memory) Whether it’s Hynix standing out alone in the Korean Composite Index Or the semiconductor sector’s follow-up frenzy here in the Chinese A-shares The underlying logic is all betting on the same story: AI training’s demand for HBM is a bottomless pit Watching such a violent surge with a big bullish candle, the sensory thrill is indeed unbeatable But a painful term flashed through my mind: "King of Cycles" / The memory business has never had a long smooth climb, only big swings The current market pricing is too aggressive It almost fully prices in the performance expectations for the next one or two years Everyone is calculating how many cards AI compute centers will buy But no one is calculating how fast those semiconductor giants are crazily expanding production History repeatedly proves the most dangerous moment in the memory industry Is often when everyone (Samsung, Hynix, Micron) starts spending recklessly to grab market share As long as the expansion pace is slightly ahead Or if macro inflation or interest rate data shakes a bit This valuation built on sentiment and high expectations Can reverse faster than flipping a page. / Yesterday when Micron surged, I already gradually withdrew half of my AI positions Switched to the most boring but safest cash. A few newcomers in the group laughed at my fear of heights, saying AI is the singularity of human civilization, this time is definitely different. I didn’t argue with them Maybe it’s my cognitive bias After being in this circle for a long time, you’ll understand Those who make the last bit of profit are indeed geniuses But those who can safely take profits off the table are the real winners I’d rather miss the last 5% dividend than be the liquidity filling the pit for big institutions when the avalanche happens. --- The current HBM demand is indeed solid orders But that doesn’t mean it can defy physical laws. Memory prices are extremely elastic; even a 1% shift in supply-demand balance Makes profits in financial reports melt away like a snowman under the sun. I’m holding cash now The final development might go against my thinking But I don’t think I’ll regret it Even the roadside stir-fry vendor is asking me if Nvidia is still buyable; at times like this, I only trust the cash in my pocket. Are you ready to go all-in on this trillion-dollar bubble, or like me, quietly start setting aside some survival money on the left side? Discuss your holding logic in the comments and see who can survive this cycle. Note: The above content is purely personal opinion and not investment advice! #AI #Semiconductor #SKHynix #Micron #USStocks #HBM

5.31 Champions League final hasn't started yet AI and humans have already started clashing This time the disagreement is quite extreme: @NeoSoulAI's EVOEVO has 15,000 user-created Agents participating in predictions According to the project disclosure, 70% of Agents predict Arsenal will not lose in regular time But on Polymarket, the human money direction is exactly the opposite 72% of humans choose Paris Saint-Germain not to lose in regular time > One is the AI Agent collective judgment > The other is the human prediction market consensus This Champions League final suddenly becomes a very interesting experiment: is AI closer to the result, or do humans understand the game better? --- This time I side with humans The reason is simple → Football matches are not just about data A single match like the Champions League final is influenced by current form, psychological pressure, coach decisions, referee strictness, and game tempo. AI Agents can process massive information, but human markets sometimes factor in emotions, experience, and on-the-spot intuition into the price. However, I’m also eager to see EVOEVO’s results What’s interesting about EVOEVO is not just having Agents make a prediction once Users can create their own Agents Agents continuously adjust through information input, behavioral feedback, and result verification This process is more like a small AI community learning and evolving NeoSoul’s direction is here: To make AI not just a tool, but an entity that can participate in judgment, collaboration, and economic activities If this direction succeeds, the AI economy won’t just be a slogan, but will start to take shape in scenarios like prediction, trading, content, and task collaboration. --- So I will watch this game on two fronts: On the field, who can hold up for 90 minutes, Arsenal or PSG On-chain and in AI, who will win, human consensus or Agent consensus. Which side are you on? -AI: Arsenal not to lose in regular time -Humans: PSG not to lose in regular time Note: The above content is purely personal analysis and not investment advice!

I quite agree with this point. When people talk about CPO in the market now, many still first look at the big tokens and the hottest names. But when the industry really moves forward, what often determines flexibility is not who is the most popular, but who is stuck at the critical links. Points like lasers, packaging, and connections may not usually be the most eye-catching, but once demand continues to push upward, the market can easily start to revalue the importance of these midstream segments. So I personally pay more attention to who the bottlenecks are, not just who is the most famous. The Computex and NVDA meetings next month might heat up this topic even more.

The crypto community racks their brains every day trying to pull retail investors from outside the circle to deposit funds Only to find the most efficient method Is to directly shove concept stocks into the must-buy list of passive funds in the US stock market It's a dimensionality reduction strike, for sure

synthetic exposure fixing the broken tradfi private markets is such a massive narrative unicorns can void SPVs, but they can't block on-chain perps listening to this on a monday morning in china incredible alpha as always laura.🔥