Technology investor Paul Kedrosky told Benzinga in an exclusive interview that AI could become a major commercial success while leaving investors with billions tied up in unnecessary data centers. He cites a house painter he knows who uses Muse, but the available text ends there.
Summary written by Upside AI from the original article. Read the source for full details.
Read the original on Yahoo Finance ↗
Keywords
AI investmentdata centersartificial intelligenceNVDA
Open a keyword to see other news on the same topic.
Related tickers and likely impact
NVDA
▼ Likely negativeNvidia
Relevance: low
| Market | US stocks |
|---|---|
| Sector | Information Technology |
| Industry | Semiconductors |
| Market cap | $5.2T |
| Close on 10/22/2026 | 212.21 |
| Change that session | -4.40% |
META
• NeutralMeta Platforms
Relevance: low
| Market | US stocks |
|---|---|
| Sector | Communication Services |
| Industry | Interactive Media & Services |
| Market cap | $1.9T |
| Close on 10/22/2026 | 648.85 |
| Change that session | -6.98% |
QQQ
• NeutralInvesco QQQ Trust
Relevance: low
| Market | US stocks |
|---|---|
| Sector | ETF |
| Close on 10/22/2026 | 704.47 |
| Change that session | -2.86% |
Impact is scored by AI from the headline and may be wrong. It is not investment advice.
Related news
- Nvidia turns the AI factory into a product , and agents into the workload
At GTC Berlin, Nvidia CEO Jensen Huang focused on deployment speed, power, security and enterprise data, describing AI as a manufacturing industry. Nvidia introduced DSX POD, a containerized reference design for complete Vera Rubin AI factories that it says can be deployed at a prepared site in as little as 14 days, versus six months, though the figure assumes the site is ready. Huang said a 1-gigawatt AI factory costs about $50 billion and earns roughly $40 billion a year, so a one-quarter delay loses about $10 billion in revenue. The DSX platform adds DSX Flex, which automates how data centers defer work for grid operators, and cooling that saves an estimated 54 million liters of water per day at 2 gigawatts. Nvidia also launched an OpenShell Reference Architecture for running agents safely and expanded its Agent Toolkit.
- AI's billion-dollar question: When will the spending pay off?
Speakers in this discussion argue AI infrastructure spending faces a timing gap, since months can pass between when computing hardware ships and when it starts earning revenue, leaving a large amount of capital tied up. As an example, a buyer of $10 million in Nvidia chips may need six months to get them running while that money sits on its balance sheet. They also point to a wider gap between total AI spending and the revenue it generates, and note that some reports suggest the size of returns may depend on technological innovation. Easing supply chain bottlenecks, they say, would help companies bring computing capacity online faster.
- Nvidia's Jensen Huang wants to build AI factories in weeks, not months
Nvidia CEO Jensen Huang, speaking at Nvidia GTC Berlin, said the new DGX Pod is designed to build AI factories faster and more cheaply, potentially cutting deployment from months to weeks. He said a gigawatt of data center costs about $50 billion, and the 4 to 6 months from shipping components to operation could be shortened by 3 to 4 months if more manufacturing happens in the supply chain, reducing idle cash. Huang argued this would make financing easier because lenders would have more certainty about costs and when revenue starts, which matters amid market volatility over neocloud companies. He said demand is very strong, citing SpaceX, CoreWeave, Nebius and Nscale, and that the large capital tied up before revenue is the key challenge to solve.
- An AI bust could come from places investors arent watching , an economist says
Concerns about an AI bubble are growing, and TS Lombard's Dario Perkins warns that investors may be overlooking other risks. Debt among smaller cloud providers, model developers and suppliers could reach $2 trillion by the end of 2027. Perkins does not see this as a systemic risk, but warns that earnings disappointments and canceled orders could ripple through the ecosystem, and that a chip glut could follow over-ordering. Much of the industry's revenue growth still comes from spending within the AI ecosystem rather than from businesses and consumers outside it. He also flags political backlash, noting that ordinary people dislike AI and nearby datacenters, and warns that one high-profile accident could turn AI toxic.
- Tech companies need more data centers. So why are nearly half of all planned projects delayed or canceled?
- How Investors Are Reacting To NVIDIA (NVDA) AI Partnership Wave
Adobe, Delta Electronics, Navitas, London Stock Exchange Group and other partners have deepened collaborations with NVIDIA, tying RTX Spark chips, Nemotron models and CUDA-X software into AI content creation, semiconductor workflows and 800 VDC data center power designs. The article argues these partnerships embed NVIDIA's AI platform more deeply in customers' daily workflows, though they do not clearly change the core catalyst of converting backlog into shipped AI factory capacity. The main risks remain export controls on China and large customers building in-house accelerators. Under the investment narrative cited, forecast revenue reaches US$995.9 billion and earnings US$540.6 billion by 2029, based on assumed annual revenue growth of 48.7%.
Related markets
Latest news
- Prediction: Nvidia Will Outperform Broadcom Over the Next 5 Years
The article compares Nvidia and Broadcom, two leading semiconductor companies in the AI boom. Over the past five years, Nvidia's shares have risen about 904% and Broadcom's about 563%, as of writing. The author predicts Nvidia will outperform Broadcom over the next five years, citing its more than 90% share of the GPU market and its CUDA ecosystem, while Broadcom holds roughly 60% of the ASIC market and its AI semiconductor business has lower gross margins than its infrastructure software segment. On valuation, Broadcom trades at 19.4x forward earnings and Nvidia at 24.9x, versus an average of 19.3x for information technology stocks; the author argues Nvidia's higher premium is warranted given its position in AI.
- IRS plans lower tax disclosure penalties, including for crypto: Forbes
The US Internal Revenue Service (IRS) is preparing to overhaul its voluntary disclosure program to cut penalties and speed up the resolution of tax violations, including those involving cryptocurrencies, according to Forbes. IRS Criminal Investigation Chief Jarod Koopman said the revised program is targeted for launch before December, with cases aimed to be resolved within 120 days after taxpayers complete the required submission process. Digital assets would not get a separate disclosure process but would follow the same procedures as other tax noncompliance, and penalties are expected to be substantially lower, though no specific figures were given. Separately, the Digital Asset Tax Certainty Act, whose Section 502 would set a process for correcting past crypto violations with defined penalties, cleared the House Ways and Means Committee in September.
- GE Vernova Fell Nearly 11% in a Day After Meta’s Earnings. Here’s Where the Stock Could Go by 2030
GE Vernova (GEV) closed at $827.42 on October 22, down 10.64% after Meta Platforms (META) reported third-quarter results showing $34.91 billion of capital spending and negative free cash flow of $2.12 billion. The stock was 29.57% below its June 30 record close of $1,174.86, though still up about 27% in 2026. The article links the sell-off to concerns about how AI buyers fund equipment, noting that GE Vernova's gas customers often pay early, with one slot reservation agreement requiring about 25% down. Second-quarter cash from operations reached $5.49 billion, but adjusted EPS of $2.41 missed the $3.01 estimate, and consensus expects free cash flow to fall about 35% in 2027. GE Vernova reports on October 28, when BofA Securities expects $21.0 billion of orders against a $22.2 billion consensus.
- AMD Stock Fell 8% in a Day as Lam Research’s Supply Warning Sank Chip Stocks. Here’s Where the Stock Could Go by 2030
AMD closed at $507.93 on October 22, down 8.10% after Lam Research (LRCX) said supply constraints held down its guidance, a drop larger than the 6.72% fall in the Philadelphia Semiconductor Index. The stock was 21.79% below its October 6 record close of $649.42. The article attributes the decline to a lower multiple rather than to estimates: AMD's NTM P/E fell from about 55x to about 46x between September 30 and October 22, while NTM normalized EPS held near $11. The main risk it identifies is the timing of rack-scale shipments, since delays would cut the estimates the model starts from. AMD reports third-quarter results on November 3 after the close.
- Meta and Google About to Run the Old Playbook Again, Will “Sell Space Inside Their Models.”
Former Stability AI CEO Emad Mostaque says Meta and Google will sell space inside their AI models, so that a request for beer could return a Bud Light recommendation. The article notes that ad revenue lets Meta and Alphabet give AI away free, while rivals that rely on paying developers must charge; Meta reported $201 billion revenue in 2025 at a 41% operating margin, and Alphabet's operating margin was 32%. According to the article, Meta's Muse launch post says conversations are kept out of its ad systems, even though Muse became the #1 free app and has paid plans at $20 and $100 a month. The article argues that putting ads inside answers is a much bigger step than placing ads beside search results and could make users doubt recommendations, and notes Meta reports third-quarter results this week.
- Is Elon Musk Coming for Verizon?
A Motley Fool Hidden Gems podcast discusses SpaceX's wireless ambitions after a reported spectrum acquisition sent Verizon stock down about 10%. One host says his current service was about eight times faster than SpaceX's $55 offering, and doubts many users would accept the slower speeds and higher latency of satellite connections. The guests note the spectrum suits terrestrial networks better than satellites, that building one could cost 10 billion or more, and that mobile carriers are valued at less than two times sales. With roughly 11,000 satellites in orbit and an estimated 40,000 users per newer satellite, Starlink could theoretically serve hundreds of millions, but speed and capital expenditure remain open questions.
Get the briefing on Telegram or Zalo
Join the Upside Telegram channel or Zalo group for the Vietnam stocks and gold briefing at 8:00 and 19:00 (Vietnam time) on trading days, plus a weekly wrap on Saturday morning.