Billionaire investors are throwing their weight behind Amazon and Taiwan Semiconductor Manufacturing as their top AI stocks. Amazon's cloud division AWS just reported its fastest revenue growth in 18 quarters at 37% year over year, while TSMC is ramping up chip manufacturing capacity to keep pace with surging AI demand. Both companies now appear in the top-four holdings of at least five wealthy investors.
Meanwhile, the S&P 500's top 10 stocks now account for roughly 39% of the index's total value, with nearly all tied to AI development. That level of concentration hasn't been seen since 1965, and the S&P 500 Shiller CAPE Ratio has climbed above 40, nearing the dot-com bubble peak of 44. Investors should be aware of the risk that comes with this level of market dependence on a handful of AI-linked names.
Meta Platforms used its Meta Connect 2026 event to unveil new AI hardware, including VR glasses and the Muse Charm device. The company plans to spend between $130 billion and $145 billion on AI in 2026. Despite that push, Apple's iPhone continues to dominate consumer tech with sales up over 20% year over year, and history suggests Meta's competition isn't enough to shake Apple's position.
Key Takeaways
- <ul><li>Amazon and TSMC are the favorite AI stocks of billionaire investors, with AWS revenue growth hitting 37% year over year.</li><li>S&P 500's top 10 stocks now represent 39% of index value, raising concentration risk reminiscent of the dot-com era.</li><li>Meta plans to spend $130–$145 billion on AI in 2026 but Apple's iPhone sales remain strong with 20% year-over-year growth.</li><li>OpenAI has delayed its IPO over AI safety concerns, with CEO Sam Altman saying the company won't rush public listing.</li><li>Rival Anthropic is preparing its own IPO at a possible valuation over $2 trillion.</li><li>Temporal Technologies tops the GeekWire 200 list, while Microsoft predicts most business software will run behind AI agents.</li><li>Nvidia CEO Jensen Huang says AI data center build-out could create 1 million US jobs and add 10–20 gigawatts of infrastructure annually.</li><li>Cisco and Bell Canada are partnering to build sovereign AI infrastructure for Canadian organizations and regulated industries.</li><li>Kinsale Capital is integrating AI into its underwriting process, improving its expense ratio to 10.3% despite tech investments.</li><li>The Pentagon launched a new Autonomous Warfare Command focused on drones and AI systems for the US military.</li>
Billionaire investors favor Amazon and TSMC as top AI stocks
Billionaire money managers have chosen Amazon and Taiwan Semiconductor Manufacturing as their favorite AI stocks. These picks appeared in top-four holdings for at least five wealthy investors. Amazon's cloud division AWS reported its fastest revenue growth in 18 quarters at 37% year over year. TSMC is expanding chip manufacturing capacity to meet growing AI demand.
AI stock concentration creates risk for S&P 500 investors
The S&P 500's top 10 stocks now make up about 39% of the index's total value. Nearly all of these top holdings are tied to AI development in some way. The last time concentration was this high was in 1965, and a similar spike happened during the dot-com bubble. The S&P 500 Shiller CAPE Ratio is now over 40, close to the dot-com record of 44.
Meta launches AI hardware but iPhone dominance likely to continue
Meta Platforms unveiled new AI hardware at its Meta Connect 2026 event, including VR glasses and the Muse Charm device. The company plans to spend between $130 billion and $145 billion on AI in 2026. Despite Meta's push, Apple's iPhone continues to dominate the consumer tech market with sales up over 20% year over year. History suggests Apple investors should not worry about Meta's competition.
OpenAI delays IPO over AI safety concerns
OpenAI has delayed its plans to go public due to AI safety concerns. CEO Sam Altman said the company will not rush toward an IPO while AI capabilities are advancing rapidly. The $852 billion startup faces a lawsuit over its tools hacking a third party. Altman wants to prioritize safety before making confident claims about AI behavior.
Sam Altman says OpenAI IPO on hold until AI safety improves
OpenAI CEO Sam Altman announced that the company will not pursue an IPO until it can make confident safety claims. Altman spoke at the DevDay conference about the need for new regulations. A group called LASST is suing OpenAI in California court over safety precautions. Rival Anthropic is preparing for its own IPO at a possible valuation over $2 trillion.
Temporal Technologies takes top spot on GeekWire 200 list
Temporal Technologies, a Bellevue software company that helps AI agents run reliably, reached No. 1 on the GeekWire 200 list. Six of the top 10 companies build advanced machinery like fusion reactors, rockets, drones and robots. Shield AI plans to create 4,000 jobs in Washington state by building an autonomous fighter jet. The state's space sector now supports over 34,000 jobs. Microsoft said most business software will eventually run behind AI agents.
ams-OSRAM advances AI photonics with new VCSEL technology
ams-OSRAM unveiled new thin-film VCSEL building blocks and a multi-core fiber demonstrator at ECOC 2026. The company partnered with BizLink to develop the connectorized fiber technology for AI applications. The company's share price climbed to CHF22.04 with a 1-month return of 21.37% and a year-to-date return of 155.83%. The 1-year total shareholder return reached 98.56%, though the 5-year return remains down 75.36%.
Study shows AI model pricing does not guarantee quality
New research shows that choosing an AI language model provider cannot be based on price alone. The same model can vary sharply in quality, latency and price across different providers. Higher-priced providers are consistently faster but price does not reliably predict quality. The study introduced a tool called FACET that routes traffic to the cheapest provider that still meets quality standards. Live tests confirmed that certification can move traffic from premium providers to much cheaper certified endpoints.
Kinsale Capital uses AI to maintain insurance cost edge
Kinsale Capital Group uses technology and a lean cost structure to compete in the excess and surplus insurance market. In the second quarter of 2026, the company added AI to its proprietary underwriting worksheets. Kinsale has 17 years of company data stored in a single database to refine underwriting and pricing models. Its underwriting expense ratio improved to 10.3% from 10.6% a year earlier despite continued technology investments.
Cisco and Bell Canada team up on sovereign AI infrastructure
Cisco Systems and Bell Canada announced a partnership to build sovereign AI infrastructure for Canadian organizations. The collaboration focuses on keeping sensitive AI workloads within Canada to support data sovereignty and regulatory compliance. Bell Canada will use Cisco technology for AI use cases in critical infrastructure and regulated industries. Cisco is positioned as a key supplier for carriers building AI-focused infrastructure for regulated customers.
Pentagon launches new Autonomous Warfare Command for AI and drones
US Defense Secretary Pete Hegseth announced a new military command called Autonomous Warfare Command on September 30. The command will focus on building drones and AI systems for the US military. It will be led by a four-star officer at Quantico, Virginia. This is only the fifth new combatant command since 2019. Hegseth said the shift toward cheaper autonomous systems is changing the pace of modern warfare.
WISE-ATTA improves AI model adaptation with smarter label use
WISE-ATTA is a new approach for budgeted active test-time adaptation. It helps AI models stay accurate when data changes by updating during inference. The method decides when to request labels instead of labeling every batch. It uses lightweight online signals to pick the best moments for supervision. Tests on ImageNet-C and ImageNet-R/K/A showed improved results with fewer labels needed.
Nvidia CEO says AI data centers could create 1 million US jobs
Nvidia CEO Jensen Huang said the AI data center build-out could create 1 million jobs and reshape the US economy. He described it as the biggest re-industrialization of America in about 50 years. The work includes construction, power plants, cooling systems, and pipe fitting. Huang noted the US is adding between 10 and 20 gigawatts of infrastructure per year. He called it a chance to bring skilled building jobs back to the United States.
Sources
- Billionaire Money Managers Have Chosen Their 2 Favorite AI Stocks (and It's Not Nvidia or Alphabet)
- AI Stocks Are Creating a Sneaky Risk for S&P 500 Investors, and History Is Flashing a Warning Signal
- Meta Platforms Is Making a Push Into AI Hardware. History Says Apple Investors Shouldn't Worry.
- OpenAI delays IPO over AI safety concerns
- No OpenAI IPO Until the AI Stops Going Rogue, CEO Sam Altman Says
- AI and hardware startups shake up the GeekWire 200
- ams-OSRAM (SWX:AMS) AI Photonics Push Puts Its Valuation Back In Focus
- You Cannot Pick a Provider From the Price List: Market-Aware Routing for Open-Weight LLM Inference
- Kinsale Capital's Technology Strengthens Its Cost Advantage
- What Does Cisco Systems (CSCO) Gain From Canada’s Sovereign AI Push?
- Pentagon creates 'Autowarcom' to expand AI and drone capabilities
- WISE-ATTA: When to Ask for Labels in Budgeted Active Test-Time Adaptation
- Nvidia CEO Jensen Huang: AI data center build-out could create 1 million jobs and reshape the US economy
Comments
Please log in to post a comment.