Nvidia's Earnings & Geopolitical Concerns, USA's 10% Intel Stake, MIT Study Dampers AI Hype, & xAI Sues Apple and OpenAI
China fights Nvidia's CUDA moat, USA takes 10% stake in Intel, MIT study worries AI investors, xAI sues Apple & OpenAI, Meta & Midjourney partner, and bee brains inspire AI breakthrough
Nvidia’s Strong Q2 Report Is Clouded by Geopolitical Concerns
Nvidia reported record sales, reaching $46.7 billion in the last quarter. This impressive growth was fueled by the high demand for AI computing, with revenue from its crucial data-center segment climbing to $41.1 billion.
Despite these strong numbers, the company's stock price fell after hours due to their $54 billion revenue forecast for the next quarter, which was seen as a sign that the explosive demand for AI chips may be slowing.
Nvidia CEO Jensen Huang dismissed these concerns during the analyst call, stating that demand for their H100 and H200 chips remains incredibly high, with everything "sold out."
The primary challenge for Nvidia's growth is its ability to produce chips fast enough to meet demand. As long as supply remains the limiting factor, the company’s sales are directly tied to its manufacturing capacity. This means that while sales numbers continue rising, the growth rate is likely slowing as Nvidia faces increasingly larger revenue figures to surpass.
A major complicating factor for the company is the geopolitical situation with China, which represents a huge market. While the U.S. government has permitted the sale of a less-powerful chip, the H20, to Chinese companies, the Chinese government has reportedly instructed companies not to buy them.
This situation with China brings Nvidia's moat into question. Nvidia's dominance is based on three key factors
Their software platform, CUDA, which is the industry standard for AI development
The fact that CUDA is available across all major cloud providers and computer companies
Their unmatched ability to connect GPUs into a single, cohesive system for large-scale AI training.
While Nvidia also benefits from having the most power-efficient chips, which is a significant selling point in the West where data center power is a concern, this advantage may not be as critical in China, as power is much cheaper.
Therefore, China adopting the CUDA platform is all the more vital, and China seems aware of the longer term risks.
USA Takes 10% Stake in Intel, The Geopolitical Price of Domestic Chips
The U.S. government announced an $8.9 billion investment for a 10% stake in Intel, aiming to strengthen national security by securing domestic chip production
Critics, like Scott Lincicome in his Washington Post column, call it a dangerous pivot from market-driven principles, warning of distorted incentives and weakened competitiveness.
And while the downsides from these critics are correct, they also leave China and Taiwan out of the discussion.
Most advanced chips come from Taiwan’s TSMC, located perilously close to China, with Samsung’s foundries in South Korea similarly vulnerable. A Chinese attack on Taiwan could cripple global chip supply, tanking economies and compromising U.S. defense. Intel’s U.S.-based foundries, though lagging, are a critical hedge against this risk.
Unfortunately, the United States has no viable alternative. The core question for Americans is straightforward: Is a leading-edge chip foundry in the U.S. essential?
If the answer is no, then it makes sense opposing this. If the answer is yes, then this move might be worth considering.
MIT Study: 95% of Organizations Have Zero Return from GenAI Investments
MIT’s GenAI Report reveals that 95% of organizations are getting zero return from their GenAI investments.
While 80% of organizations are using tools like ChatGPT and Copilot, these are primarily used to boost individual productivity, not to improve P&L performance. MITs study showed that users vastly prefer these generic LLMs instead of integrated tools.
Enterprise-grade systems are the real story. 60% of organizations evaluated such tools, only 20% reached a pilot, and just 5% made it to production.
The main reasons for these failures include rigid workflows, a lack of contextual learning, and a poor fit with daily operations.
We have seen this at our company Firmograph, an AI agent for sales leaders. Selling is too dynamic for a one size fits all approach. And if we go too wide, we risk losing mindshare to generic tools like ChatGPT.
Instead, we tackle use cases individually, automating specific jobs and tasks. This approach ensures our customers get real value and avoid the common pitfalls of GenAI adoption, even if it means a slower rollout and lower ARR initially.
Elon Musk’s xAI Sues Apple and OpenAI Over ChatGPT-iPhone Deal
Elon Musk's xAI has filed a federal lawsuit against Apple and OpenAI, accusing them of anticompetitive practices stemming from their exclusive ChatGPT integration deal for iPhones.
The suit claims the partnership creates an "illegal monopoly" that disadvantages rivals like xAI's Grok chatbot, potentially costing billions in damages. Musk's companies argue that Apple's OS-level favoritism toward ChatGPT stifles innovation and limits consumer choice in generative AI.
Central to the complaint is the allegation that Apple manipulates App Store rankings to keep ChatGPT at the top, deprioritizing competitors and delaying their app updates. Musk has publicly claimed this rigging makes it "impossible" for non-ChatGPT AI apps to reach number one, framing it as antitrust abuse tied to the OpenAI deal.
However, competitors of xAI have shown that it is possible. Chinese AI app DeepSeek surged to the top of the App Store charts in January 2025, just months after Apple's OpenAI partnership. This ascent, alongside Perplexity's brief number-one stint, directly undercuts claims of artificial suppression.
The lawsuit underscores the distribution power platforms like Apple have. While generative AI is a new paradigm, it seems like the major players are at the mercy of companies who won the prior paradigm.
Meta Partners with Midjourney on AI Image and Video Models
Meta has teamed up with Midjourney to supercharge their image and video generation capabilities.
The collaboration will integrate Midjourney's advanced technology into Meta’s systems, a move that could significantly help the Meta’s 2 million advertisers, especially small and medium-sized businesses who lack image animation resources.
While this is a clear win for Midjourney, this partnership also shows how behind Meta is with their own AI efforts. Instead of waiting for their technology to improve, Meta is relying on this partnership, which will cut into their margins.
We’ll see how quickly Meta can catch up in generative AI. This partnership signals that they’re still very far behind.
Bee Brains Inspire AI Breakthrough in Perception
Researchers at the University of Sheffield found that bees enhance brain signals through flight movements, a discovery modeled digitally to propose a new AI perception framework. By mimicking bee neural processing, this approach could enable AI to better interpret dynamic environments.
This bio-inspired leap could revolutionize robotics and autonomous systems, where real-time perception is critical. Unlike Google’s Genie 3, which focuses on world simulation, bee-inspired AI targets lightweight, adaptive sensing, potentially slashing compute needs.