And I actually think Michael Bur's got this wrong. And so what Michael Bur is saying is that all of these hyperscalers have extended their depreciation schedule or the useful life of their data centers by roughly 2x which cuts the operating costs in half when they report it in earnings and so it's making their earnings inflate. So he's claiming they're cooking the books.Google first made this change in Q1 of 21 where they said the servers are now going from 3 to four years. The chart shows the combination o ...

allowing all of those AI plants that are being built to build their own electric generating force. And they're building them like nobody's ever seen. They're building a plant for $50 billion.You know, if you build a shopping center for $50 million, it's okay. It's not a bad shopping center. 500 million.You're talking about topof the line. They're spending 50 billion. I said, "No, you mean 50 million." "No, no, sir. We mean 50 billion." And then I saw some of them that were the size of Manhattan. ...

AI Technology & Market - The speaker expresses excitement about AI technology and the investments being made to deploy it in enterprises, anticipating significant productivity gains [4] - The market is currently optimistic about AI, potentially underestimating risks and overestimating the speed of adoption; adjustments and volatility are expected [5] - Despite potential short-term volatility, the long-term secular trend of AI is expected to bring extraordinary economic benefits and productivity gains [13][15] - Some AI-related assets may be overvalued, with earnings not meeting current expectations, which is typical in acceleration cycles [18] Capital & Financing - Hyperscalers are investing more of their substantial cash flow into AI projects and are considered financially strong [9] - Data center financing is viewed as real estate deals, with debt serviced by the output of the data centers, and the creditors are typically large, stable companies [10][11] - Private capital formation in new AI companies involves higher risk, with potential for both significant successes and failures [11][12] - Private credit is a large business, and it's important to differentiate between lending to established companies and financing data centers [21][25] - Underwriting standards and risk management are crucial in credit, and economic contractions will reveal the strengths and weaknesses of different lenders [26][27]

Industry Risk - Rising electricity demand from data centers is raising the risk of blackouts across a wide swath of the US during extreme conditions this winter [1] Regulatory Oversight - The regulatory body overseeing grid stability is monitoring the situation [1]

Market Trends & Economic Impact - Market faces key tests with Nvidia results and the September jobs report [2] - Data center buildout is moving to the "rodeo region" (Mountain West and Southwest), attracting billions in investments from tech giants [4] - Capital investment in semiconductor plants and data centers is an important component of overall US economic growth [7] - Google committed $40 billion to Texas, which will have significant knock-on effects on the regional economy, including construction and permanent employment [9] Technical Analysis & Market Performance - S&P 500 broke through its 50-day moving average, the first close below since just before May 1st [19][20] - Sector rotation is the lifeblood of the markets, with healthcare (XLV) up 9% since the beginning of the quarter, while the S&P 500 is negative [25][26] - Bitcoin is down 17% quarter to date and 2% over the trailing 24 hours, breaking support levels [28] - Ethereum is down 21% and Solana down 28% over the last month, indicating damage within the crypto ecosystem [30] - The US dollar index has risen off of lows but upside is limited, with 96 as a level to watch to the downside [31][32] AI & Technology - Grammarly rebranded to Superhum, with a current valuation of over $13 billion [33] - Superhum has 40 million daily active users worldwide, with about a third being students and two-thirds being professionals [36] - Superhum differentiates itself by focusing on "assist" AI, bringing AI to where people work in various applications [39][40] - Superhum does over 100 billion LLM (Large Language Model) calls per week [50] - Superhum generates over $700 million in revenue a year [51]

Nvidia and AI Infrastructure - Nvidia remains well-positioned in the AI infrastructure market, with demand not slowing down [3] - The shift from the Hopper era to the Blackwell era is underway, driving the need for GPU clusters and compute clusters [1][2] - Over 50 gigawatts of IT load capacity is planned or under review for new data center builds in the next 5 years, representing greenfield growth [2] - Hyperscalers' continued investment in AI infrastructure depends on the profitability of their AI investments [4] - AI data centers will fill the gigawatts of capacity coming online in the next 3 to 5 years [5] Apple Leadership and Innovation - Tim Cook is Apple's longest-serving CEO at 14 years [6] - Apple has a deep bench of potential successors, including John Turnis and Craig Federighi [7] - Apple has an opportunity for another hardware run, including AI wearables, smart glasses, and the evolution of Vision Pro [8] - The next Apple CEO will need hardware and software expertise, especially in AI, to drive the company's growth [10][12] - Maintaining a strong supplier and supply chain position is crucial for Apple, even after Tim Cook's succession [11]

AI Development & Infrastructure - Compute resources are consistently insufficient for AI development [1] - Data centers are critical infrastructure for the digital economy [1] - Open-source AI models represent a significant source of geopolitical influence [1] Geopolitical & Economic Factors - China's government's national commitment to industry is a powerful force [1][2]

Data Centers, Crypto Mining To Push Electricity Costs Higher In 2026 https://t.co/NjBh4qhgjx ...

Core Insights - The rise of data centers in the U.S. is leading to significant increases in electricity costs for consumers, with some experiencing bills rising by 20% in a month [3][7] - The U.S. Department of Energy predicts that by 2028, data centers will consume between 6.7% and 12% of the nation's total electricity, marking a shift from a marginal to a central role in energy consumption [4] - The demand for electricity from data centers is causing a strain on the power grid, particularly in the Mid-Atlantic region, where consumers are already feeling the financial impact [6][8] Group 1: Electricity Cost Increases - Maryland residents are among the first to feel the impact of rising electricity costs due to data center demand, with hidden costs in electricity bills that consumers may not easily notice [6][9] - The price of future electricity capacity has surged by over 1000% compared to two years ago, driven by anticipated demand from data centers [8][9] - The total cost of infrastructure upgrades proposed by PJM to accommodate new data center loads exceeds $110 billion, which will ultimately be passed on to consumers [9] Group 2: Regional Variations - States with high concentrations of data centers, such as Virginia, Illinois, and Ohio, are experiencing electricity price increases significantly above the national average, with Virginia seeing a 13% rise and Illinois a 16% rise [13][15] - Texas, despite having a large number of data centers, has seen a lower electricity price increase of about 4%, attributed to its independent grid system [17] - California's electricity prices have only risen by about 1%, partly due to the removal of wildfire-related costs from customer bills [17] Group 3: Political and Social Implications - The rising electricity costs are becoming a political issue, with candidates in Virginia attributing price increases to data centers and promising accountability from tech companies [15] - There is growing public sentiment against the financial burden placed on consumers due to the expansion of data centers, with many Americans struggling to afford basic necessities [18] - The challenge of balancing technological advancement with social equity is becoming a pressing concern, as ordinary citizens demand fairer energy costs in the age of AI [18]

RT Sundar Pichai (@sundarpichai)Today, we're announcing a new $40B investment in Texas through 2027 to build Cloud & AI infrastructure and support thousands of new jobs.This includes new data centers in Armstrong and Haskell Counties and a major investment to strengthen energy resilience and abundance. We're also providing funding to more than double the projected pipeline of new Texan electricians to power the AI era.We've proudly called Texas home for 15 years. Thank you @GregAbbott_TX for your continued ...