Billionaire investors and quant analysts are clashing over whether AI stock valuations hold up under scrutiny. Hedge fund manager Appaloosa sold Microsoft in Q2 while buying Apple and increasing stakes in Meta, Amazon, and Alphabet. The concern centers on whether hyperscaler spending on AI infrastructure will translate into strong returns. Michael Burry's traditional value investing stance faces off against quantitative analysts who argue the fundamentals may be closer to justified than critics claim.
Market pressure on AI stocks remains tangible. The Nasdaq fell 2.4% and the S&P 500 dropped 1.3% in one of the worst weeks since July. CoreWeave fell 25% over five days, while AMD, Micron, and SanDisk each dropped more than 5%. Gartner research shows AI providers plan to spend $7 trillion on infrastructure through 2030 but may generate only $3.6 trillion in revenue. About a third of AI spending comes from debt, and global AI debt issuance fell from $113 billion in June to $23 billion last month.
Anthropic's $2 trillion valuation faces questions despite 51% revenue growth in the third quarter. The company is also ramping up human support for customers: Anthropic plans to train 10,000 specialist engineers by the end of 2027. AWS has committed $1 billion and OpenAI has invested $4 billion in Field Deployment Engineers who both sell and deploy AI for customers. Consultancies like Deloitte, Accenture, and KPMG are also hiring heavily for these roles, underscoring how human expertise remains a key differentiator in AI adoption.
On the product front, Alibaba's Taobao upgraded its cross-border shopping experience with AI-powered tools offering product recommendations, price comparisons, and curated international products priced in Chinese yuan. DeepLumen built a system that charges Shopify merchants only when AI orders come through their platform, with nearly 2,000 merchants now using it for agentic commerce attribution. Meanwhile, startup Deploiable raised €500,000 to build AI-native products for small and medium-sized enterprises using a micro-acquisition model.
Research developments offer a more grounded picture. A new study on AI math formalization found that agents produce definitions only 25 to 50 percent compatible with later use, but clarifying informal definitions improved compatibility to 77.5 percent. A benchmark called ReuseBench was introduced to test formalization quality. Separately, a hierarchical graph convolutional network called HGCN achieved 91.20 percent accuracy detecting low-quality documents by identifying internal inconsistencies and headline-body mismatches.
Key Takeaways
- Billionaire investors sold six AI stocks in Q2 amid returns concerns, with Appaloosa selling Microsoft but buying Apple and increasing stakes in Meta, Amazon, and Alphabet
- Nasdaq fell 2.4% and S&P 500 down 1.3% in worst week since July; CoreWeave dropped 25% over five days
- Gartner projects $7 trillion in AI infrastructure spending through 2030 but only $3.6 trillion in expected revenue
- Global AI debt issuance fell from $113 billion in June to $23 billion last month
- Anthropic's $2 trillion valuation questioned despite 51% revenue growth in Q3
- Anthropic plans to train 10,000 specialist engineers by end of 2027; AWS committed $1 billion and OpenAI invested $4 billion in Field Deployment Engineers
- Taobao upgraded cross-border shopping with AI tools including product recommendations and price comparisons in Chinese yuan
- DeepLumen's AI referral system on Shopify covers nearly 2,000 merchants, charging only on AI-originated orders
- Deploiable raised €500,000 to build AI-native products for SMEs using a micro-acquisition model
- AI math formalization study found clarifying definitions improved compatibility from 25-50% to 77.5%; HGCN achieved 91.20% accuracy detecting low-quality documents
Michael Burry and Quant Analysts Clash Over AI Stock Value
AI infrastructure spending continues as hyperscalers pour hundreds of billions into data centers. Some investors worry about circular financing and whether AI capital spending is sustainable. Steven Cress compares traditional value investing views against quantitative analysis methods. The debate centers on whether current AI stock valuations can be justified by fundamentals.
Three AI Stock Picks Despite Bubble Fears
Fears of an AI bubble have caused market panic due to competitive pressure from China and slowing revenue growth at frontier AI labs. However, the bond market suggests stocks may be near a bottom despite inflation worries. The author recommends three AI stocks with strong growth and expanding margins. All three picks reached all-time highs recently, showing technical strength.
DeepLumen Tracks AI Referral Sales on Shopify
DeepLumen has created a system that charges Shopify merchants only when AI orders come through their platform. Nearly 2,000 Shopify merchants already use this system for agentic commerce attribution. The approach uses order-level tracking for AI-originated shopping journeys. This commission-based model reduces merchant risk while creating measurable revenue from AI recommendations.
Improving AI Math Formalization With Definition Grounding
AI systems can now generate formally verified Lean code at high speed for mathematical research. However, definitions and helper lemmas must be general enough for others to use in further research. The study found that agents often produce definitions that later statements cannot use, with only 25 to 50 percent compatibility. Clarifying informal definitions improved compatibility to 77.5 percent. A new benchmark called ReuseBench was created to test formalization quality.
Graph Network Detects Low-Quality Documents With 91 Percent Accuracy
Researchers developed a hierarchical graph convolutional network called HGCN to detect low-quality documents. The system identifies internal inconsistencies within documents and mismatches between headlines and body content. The model achieved 91.20 percent accuracy on the Inconsistency Dataset and 92.00 percent on a public dataset. The study also found that using metadata significantly improved detection performance.
Billionaire investors sell six AI stocks amid returns concerns
Billionaire investors sold six AI stocks in the second quarter due to worries that returns on AI infrastructure investments may be weak. Hedge fund managers like Appaloosa sold Microsoft but bought Apple and increased stakes in Meta, Amazon, and Alphabet. The market can quickly change direction if hyperscalers show AI is improving their financial results. Retail investors should use billionaire moves as idea sources but not blindly follow them. Billionaires are not abandoning AI entirely but are concerned about specific companies.
AWS and Anthropic deploy thousands of engineers for AI customers
Tech giants including AWS, Anthropic, and OpenAI are spending billions to deploy specialist engineers who help customers use AI in real-world situations. Anthropic plans to train 10,000 specialist engineers by the end of 2027, while AWS has committed $1 billion and OpenAI has invested $4 billion in Field Deployment Engineers. These engineers handle both selling and deploying AI for customers. Consultancies like Deloitte, Accenture, and KPMG are also hiring heavily for these roles. The most successful AI companies may be those that use human engineers best to attract customers.
AI stocks fall amid debt and data center spending fears
AI and tech stocks dropped sharply on Friday, with the Nasdaq falling 2.4% and the S&P 500 down 1.3%, marking their worst week since July. CoreWeave fell 25% over five days, while AMD, Micron, and SanDisk each dropped over 5%. Gartner research shows AI providers plan to spend $7 trillion on infrastructure through 2030 but may generate only $3.6 trillion in revenue. About a third of AI spending comes from debt, and global AI debt issuance fell from $113 billion in June to $23 billion last month. Anthropic's $2 trillion valuation is being questioned despite 51% revenue growth in the third quarter.
Taobao upgrades cross-border shopping with AI tools
Alibaba's Taobao platform upgraded its cross-border shopping experience with curated international products and AI-powered tools. Users can now browse international brands with prices and shipping shown in Chinese yuan. The AI tools offer product recommendations, price comparisons, and customer reviews to help shoppers decide. Taobao partnered with international brands to offer exclusive products and promotions. The upgrade aims to meet growing Chinese consumer demand for international products in a market expected to reach $4.47 trillion by 2025.
Deploiable raises €500,000 to build AI products for businesses
Deploiable, a startup founded by Roberto Macina, Valentina Iannucci, and Alessandro Piccinini, secured a €500,000 funding round from the Cannonball fund and business angels. The company builds AI-native products for niche markets, focusing on generating measurable results for small and medium-sized enterprises. Their approach uses micro-acquisitions as a path to innovation rather than traditional startup growth cycles. Macina previously founded Qurami, a queue management app acquired by Ufirst in 2018. The team plans to expand this product lab model from Italy to the broader European market.
Sources
- Top AI Stocks: Michael Burry Vs. Quant Fundamentals
- My AI Bubble Shopping List: 3 Easy Strong Buys
- AI Referral Sales Trackers
- Definition Grounding in Theory-Level Auto-formalization
- Hierarchical Graph Convolutional Network Approach for Detecting Low-Quality Documents
- Billionaires Are Selling These 6 Artificial Intelligence (AI) Stocks. Should You Follow?
- Thousands of specialized engineers deployed by AWS, Anthropic to prove AI is worth the investment
- AI stock sell-off ramps up over debt-fueled data center spending fears: ‘Full-blown crisis of confidence’
- Taobao upgrades cross-border shopping with curated products, AI tools
- Deploiable secures a €500,000 funding round for AI products for businesses
Comments
Please log in to post a comment.