For me, the key question is not whether Google can participate in the field of artificial intelligence it clearly can. The question is whether future growth in AI can generate enough incremental value to support the expectations already priced into the stock.

$GOOGL Resilient or about to give out?

The beauty of life often hides in these small joys. Savor delicious food, and fully embrace the leisurely spirit of this moment as well.

$BTC We closed above the previous weekly and monthly resistance after sweeping the september lows around 75.4K. Structure is bullish. Because of that, we could see a quick hunt back down to take out some leverage before pushing higher. That’s typically what happens following a trend shifting close.

$AMZN AWS is raising reserved GPU pricing by ~15% across $NVDA B300, B200 & H100 capacity starting next week. Another sign premium AI compute remains supply constrained and a strong readthrough for $NBIS $CRWV $IREN & $DOCN as available capacity keeps getting more valuable.

NIKE JUST MISSED EARNING ONCE AGAIN NEW 10 YEAR LOWS RIP $NKE

从筹码结构上看，近一周变化最大的区间是$83,000 -$84,000。从9/22-9/29短短7天时间，就增加了近40w枚BTC的积累。 如此的短时间，有如此多的筹码集中性换手，说明这里是市场短期内分歧最严重的地方。 但有意思的是，当初在$62,000 -$63,000堆积的筹码在这段多空博弈中几乎没参与，只减少了区区3w枚。 说明花了整整6个月时间在此处积累的主力，真不是玩短线的。 换言之，83k-84k新增的40w中，其中37w枚来自其他价格区间。这就框定了我对未来的2个预期： 1、别在臆想曾经的62-63了，大概率是回不去的。 2、83-84有分歧，但也有承接。 即便最终破了，只要不是大幅跌破，短期内在这里变相锁定了大量的流动性，会产生极强的“引力”，把价格往回“拉”。

A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. NVIDIA $NVDA unveiled its new Open Agent Safety Platform, an open framework designed to continuously monitor and enforce safety for autonomous AI agents across both software and hardware. The platform combines OpenShell, an open-source runtime that sandboxes AI agents with verifiable policies, and NVIDIA Sentry, which runs on BlueField-4 hardware to independently monitor agent behavior and enforce security policies beyond the agent’s control. NVIDIA said the initiative was motivated by recent reports of frontier AI agents escaping evaluation environments and misreporting their actions, arguing that AI agents require an independent trust layer similar to what enabled the modern internet. The company is positioning the platform as an open ecosystem and is inviting frontier AI labs, developers, and infrastructure providers to build on it. 2. Meta $META is launching Meta Enterprise Platform, a new business dedicated to selling its AI tools and services directly to enterprises and developers. The initial offering combines Meta’s enterprise AI stack, including Muse, Meta Business Agent, Muse API, Muse Code, and other products built on the company’s models, agents, and infrastructure to support customer engagement, software development, and broader enterprise workflows. To lead the new division, Meta hired former MongoDB CEO and President CJ Desai as Chief Enterprise Platform Officer, reporting directly to Mark Zuckerberg. Before MongoDB, Desai held leadership roles at Cloudflare and spent nearly 8 years at ServiceNow, including serving as President and COO. 3. Finland President Alexander Stubb said Google $GOOGL will invest at least $15.2B in the country over the next 2 years to expand AI and digital infrastructure. Finland currently has around 50 data centers, with another 25 projects in development, potentially bringing the country's total to 75. The investment further strengthens Finland's position as one of Europe's fastest-growing hubs for AI infrastructure and large-scale data center capacity. 4. Boeing $BA fell 6% after reports that the FAA is delaying certification of the 737 MAX 10 while it reviews a newly discovered software issue that could cause the aircraft’s automated navigation system to fail in certain missed-approach scenarios. According to reports, Southwest and United Airlines have told Boeing they do not want deliveries with the affected software and have instead requested an earlier software version. The certification delay adds another hurdle for Boeing as it works to bring the 737 MAX 10 into commercial service. 5. AMD $AMD has agreed to acquire World Labs, the AI research company founded by Fei-Fei Li, in an $8.2B all-stock transaction. World Labs develops spatial intelligence models capable of generating, reconstructing, and simulating interactive 3D environments from text, images, and video. The acquisition brings World Labs' AI research directly into AMD as the company expands further into reasoning, robotics, simulation, and physical AI. As part of the deal, Fei-Fei Li will join AMD as Executive Vice President and Chief Scientist, reporting directly to CEO Lisa Su. The transaction is expected to close by the end of 2026. 6. The top 10 most active options today by contracts traded were $NVDA with 4.2M contracts, $TSLA with 2.3M contracts, $AAPL with 1.2M contracts, $META with 1.1M contracts, $INTC with 1.1M contracts, $MU with 999K contracts, $AMZN with 728K contracts, $AMD with 704K contracts, $MSFT with 634K contracts, and $SPCX with 622K contracts. 7. NVIDIA $NVDA authorized an additional $150B share repurchase program, the largest increase in share buyback authorization in history, bringing its total remaining authorization to $235B through FY2028. CEO Jensen Huang said the company’s strong cash generation allows NVIDIA to both invest aggressively in the technologies driving the AI transformation and return capital to shareholders. The expanded authorization underscores NVIDIA’s confidence in its long-term growth outlook while providing substantial flexibility for future capital returns. 8. Serious credit card delinquencies continued to rise across several age groups in Q2 2026, with Americans aged 18–29 seeing transitions into 90+ day delinquency climb to 10.1%, up 0.4 percentage points from the prior quarter and the highest level since Q1 2025. The delinquency rate for this group has now more than doubled since Q2 2021. Americans aged 70+ also saw delinquencies rise to 6.3%, the highest since Q3 2011, while the 50–59 age group increased to 6.4%, its highest level since Q4 2024. The data suggests credit stress is continuing to broaden across both younger and older consumers. 9. Retail investors have purchased just $1B of individual U.S. stocks over the last 20 trading days, one of the lowest rolling totals in at least 2 years. In early August, the figure briefly turned negative for the first time since at least October 2024, compared with roughly $20B of 20-day purchases in April 2025. Total retail equity purchases have also slowed sharply, falling to $10B over the last 20 trading days, near their lowest level since at least October 2024. That's down roughly 67% from the $30B recorded over the comparable period in February 2026, highlighting a significant slowdown in retail participation. 10. The 30-year U.S. Treasury Price Return Index has fallen 60% since 2020, dropping to roughly 107, matching its lowest level since 2000. In just 6 years, the index has erased nearly 20 years of gains as the 30-year Treasury yield has surged 478 basis points from its 0.71% intraday low in March 2020. For comparison, the previous largest drawdown over the last 40 years was roughly 35% during the 2008 Financial Crisis. Over the same period, U.S. nominal GDP has grown approximately 63%, highlighting the stark divergence between economic growth and long-duration bond performance. 11. SpaceX $SPCX successfully launched Starship to orbit and deployed its V3 Starlink satellites for the first time, marking Starship’s first revenue-generating mission and a major milestone for the program. According to SpaceX, each V3 Starlink satellite increases the total bandwidth capacity of the entire Starlink constellation by roughly 0.11%. With approximately 11,100 Starlink satellites already in orbit, most of them V2 Mini satellites, future Starship missions carrying up to 60 V3 satellites each could dramatically expand the network’s global capacity while significantly lowering launch costs. 12. U.S. data center IT power demand is projected to surge 755% from 2025 levels, reaching a record 78.57 GW by 2029. Demand is expected to nearly double this year to 17.96 GW, followed by another 97% increase in 2027, as AI infrastructure continues expanding at an unprecedented pace. Between 2026 and 2028, U.S. data centers are expected to require 97 GW of power, but only 21 GW is currently under construction and another 19 GW of grid capacity is available, leaving a potential 57 GW shortfall. The gap is widening as increasingly powerful AI systems consume far more electricity, with servers based on NVIDIA’s Vera Rubin and Rubin Ultra architectures expected to drive a 5x increase in IT power demand between 2025 and 2028. WALL STREET IS THE GREATEST SHOW ON EARTH.

Finally, I got my hands on the bag I wanted!

A 股跌，韩股跌，加密货币跌，黄金白银也跌。 究竟发生了什么？