Headlines this week - Nov 30, 2025
A look at how capital is being deployed across future opportunities
This week in the future:
1 - Debt markets continue under pressure as AI labs issue bonds to fund AI infrastructure deployments
Prices of newly issued bonds are falling due to a flood of AI-related titles. Prices for newly issued corporate bonds are slipping as the market struggles to absorb a surge of debt from AI companies. This “flood” of AI-related paper is adding significant pressure to broader credit markets, causing spreads to widen and complicating future fundraising.
Investors agree that the risk is growing, despite Nvidia’s earnings performance. Despite Nvidia’s strong earnings, the broader market remains anxious, signaling that the “easy money” phase of the AI trade might be over. Investors are increasingly focusing on rising risks and execution challenges rather than just potential upside, leading to heightened volatility.
Some point to Intel’s history as a cautionary tale for overspending. Observers are drawing parallels to Intel’s past capital spending binges, warning that “spending too much” is a real possibility. The fear is that today’s massive AI infrastructure investments might lead to diminishing returns and financial strain, similar to Intel’s historical struggles.
An additional problem is that risks are coupled, with OpenAI’s success critical to repaying massive debt. Systemic risk is rising as OpenAI’s partners amass a staggering $100bn debt pile to fund infrastructure. Repayment of these loans hinges critically on OpenAI’s continued success and ability to monetize, creating a precarious chain of financial dependency across the sector.
The U.S. economy is now also heavily dependent on AI capital expenditure. The U.S. economy has become dangerously “hooked” on AI capital expenditure to drive growth. Macroeconomic data reveals that without this massive wave of tech spending, broader economic indicators would look significantly weaker, highlighting a fragility in the current economic expansion.
This dependency is extending globally, e.g. driving Taiwan’s economic surge. This trend is global, as shown by Taiwan’s economy roaring ahead almost solely on the back of insatiable AI demand. The island’s export growth and industrial production are surging, underscoring how the global tech supply chain relies on this single boom.
2 - The expectation is that big AI labs, under cash flow pressures, will keep going to the debt markets
OpenAI must raise massive sums just to sustain its projected losses. HSBC estimates OpenAI needs to raise at least $207bn by 2030, in addition to the massive amounts that they have already secured, and assuming a very fast growth in number of users (up to 3bn by 2030) and revenues. The staggering cash requirement would suggest a long-term dependence on external financing rather than organic cash generation.
Although in better shape, Big Tech is also under stress, and even using “aggressive accounting”. Facing severe cash flow pressure from gigantic data center builds, Meta is resorting to aggressive accounting measures, such as extending the useful life of its servers. These accounting tweaks help boost reported earnings while the company burns through billions in capital expenditures.
Alphabet appears to be the exception, defying bubble fears. Unlike its peers, Alphabet is defying bubble concerns, with its stock soaring on the strength of its custom chip infrastructure (see below) and leadership in generative models. Its diversified business and vertical integration offer a buffer that other tech giants currently lack.
Asian chipmakers could be the first to crack if the bubble bursts. Sitting at the top of the supply chain, Asian chip foundries and memory manufacturers are highly cyclical and most vulnerable to a downturn. Analysts warn that any cooling in AI demand will likely trigger an inventory correction here first.
3 - Alphabet / Google is seen as the new AI leader, after the success of Gemini 3.0 (but not only because of that)
Beyond models, Alphabet may be winning the AI chip race too. Nvidia shares fell 2.6% last Tuesday as signs emerged that Google is gaining the upper hand in AI infrastructure. Nvidia’s investors worry Google’s TPUs can effectively compete with Nvidia’s GPUs, challenging its long-held monopoly in the data center market.
Even arch-rival Meta is considering a massive deal for Google’s chips. In a surprising twist, Meta is in talks to buy billions of dollars worth of Google’s AI chips. This potential partnership highlights the industry’s desperate need to diversify suppliers and reduce reliance on Nvidia’s costly and supply-constrained hardware.
The Gemini 3.0 launch signals Google is now “fully awake”. After having been seen as a “sleeping giant”, analysts are now declaring Google “fully awake“ following the successful launch of Gemini 3.0 and its chip advancements. The tech giant is effectively leveraging its vertical integration of data, models, and custom silicon to reclaim its position as the global AI leader.
So not all AI stocks are the same, anymore.The previously monolithic “AI trade” is splintering as Google challenges Nvidia’s dominance. Investors are no longer lifting all boats; instead, they are picking winners and losers, actively rotating capital from chip incumbents to vertically integrated cloud giants.
4 - The AI regulatory battle in the US could be heating up. It could turn into an “existential” fight for AI labs
A major debate on AI regulation may become central to the midterm elections. Wealthy donors are organizing Super PACs to make AI safety a central issue in the upcoming midterm elections. This political push aims to impose stricter oversight, challenging the industry’s unchecked growth and potentially curbing its rapid expansion.
However, regulation could hinder the revenue growth labs urgently need. Regulation could conflict with growth, as seen in OpenAI’s struggle with users forming unhealthy emotional bonds with ChatGPT. Implementing safety measures to prevent users from losing touch with reality risks reducing the engagement levels vital for the company’s financial success.
Big AI labs are amassing war chests for a potential “existential” fight. Tech titans are raising multimillion-dollar war chests to fight what they view as an existential regulatory war. While systemic economic risks make drastic restrictions unlikely, companies are aggressively lobbying to ensure future rules don’t stifle their survival or growth.
5 - More arguments in favor of a vision of AI as a complement to human workers, rather than a substitute. But many uncertainties remain
McKinsey envisions future work as a partnership between people and AI. A new report from the McKinsey Global Institute predicts that the future of work will rely on “skill partnerships” between humans, agents, and robots. Rather than total automation, they argue that AI will primarily augment human capabilities, requiring deep collaboration between people and technology.
The “jagged” nature of AI performance supports the complementary view. This partnership model is supported by the technical reality of AI’s “jagged frontier,” where models excel at some complex tasks but fail at simple ones. This unpredictability necessitates human oversight, making AI a poor candidate for full substitution in many roles.
Early adopters are using AI as a tool, not a replacement. Real-world adoption patterns show workers are treating AI as a “power tool” rather than handing over the reins. Employees often hesitate to delegate even basic tasks completely, preferring to remain in the loop to ensure quality and maintain control.
However, adoption is skewed away from those who can extract most value of AI. The benefits of AI are currently uneven, as early adopters tend to be highly skilled workers who need the help the least. Because lower-skilled workers are adopting the tech more slowly, the broad, transformative productivity gains economists expect may be delayed.
Meanwhile, companies continue to use AI as a rationale (excuse?) for layoffs. Despite the “augmentation” narrative, companies like HP are still linking AI to workforce reductions. HP announced plans to cut up to 10% of its staff, explicitly citing its strategic shift toward an “AI push” as a driver for the restructuring.
6 - Robots keep attracting capital as a potential “next frontier” in AI
China is rapidly deploying robots to build the “future of factories”. China is aggressively integrating AI and advanced robotics into its manufacturing sector to counteract a shrinking workforce and property slump. This state-driven push aims to modernize the economy by creating highly automated, efficient factories capable of producing higher-value goods.
Humanoid robots are attracting massive funding for “mundane” physical tasks. Physical Intelligence (an American startup) has raised massive capital to develop “universal brains“ for robots, aiming to enable them to perform everyday tasks like folding laundry or clearing tables. This “next frontier” focuses on creating software that allows various robot types to generalize and master diverse physical interactions.
But China is already warning of a potential bubble in the sector. Despite the hype, Chinese state media and industry officials are warning of “blind expansion” and overheating in the humanoid robotics industry. They fear a bubble is forming as investment frenzies outpace the technology’s current commercial viability and reliability.
7 - China’s dominance in open-source AI models could turn into a key geopolitical advantage
Countries are pursuing “Sovereign AI” to avoid dependence on superpowers. Governments worldwide are launching “sovereign AI” initiatives to reduce reliance on U.S. and Chinese tech giants. By building domestic infrastructure and training local models, nations aim to secure control over their data and strategic autonomy in the AI era.
As a result, China’s leadership in open-source models offers a new “soft power” edge. China has surpassed the US in releasing high-performance “open” AI models, such as Alibaba’s Qwen, granting it unexpected geopolitical leverage, as these models can be used by countries as a platform to develop their own “Sovereign AIs”. This “model diplomacy” would allow Beijing to embed its technology in global software, expanding its influence and soft power across the developing world.
8 - Money keeps flowing into nuclear energy projects, that look to remove a key bottleneck for AI progress. Could Small Nuclear Reactors accelerate deployments?
The US plans massive investments in large reactors, but execution risks loom. As discussed in previous editions, the U.S. is planning an $80bn partnership with Westinghouse to build eight large reactors to meet soaring electricity demand, especially from AI. However, analysts are pointing to past projects with the same technology which suffered major delays, cost overruns and workforce and supply-chain challenges
Meanwhile, innovative SMRs continue attracting massive capital from top investors. Despite having no commercial plants yet, Small Modular Reactor (SMR) developer X-energy raised approximately $585m in a round backed by Jane Street, Citadel, and Amazon. This massive bet values the startup at over $2.5 bn, underscoring the intense investor interest in next-generation nuclear solutions.
9 - Europe’s defense industry is running to adopt AI
Leonardo is building an AI-powered “Michelangelo Dome” air defense system. Italian defense group Leonardo is developing “Michelangelo Dome,” a drone-based air defense system that uses artificial intelligence to rapidly identify and intercept aerial threats. The system aims to provide a comprehensive shield against drones and missiles.
German startups Helsing and Stark are racing to fill the British Army’s drone gap. German AI defense startups Helsing and Stark Defense are bidding to supply thousands of attack drones to the British Army. They aim to fill a critical capability gap by providing cheap, mass-producible unmanned systems optimized for modern warfare.
Spy drone maker Quantum Systems secures a €3bn valuation. German drone manufacturer Quantum Systems has raised new funds at a €3bn valuation, cementing its status as a defense unicorn. The company, which supplies surveillance drones to Ukraine, is capitalizing on the surging demand for autonomous military technology.
10 - The battle for “Quantum Supremacy”… in the markets continues
IonQ’s CEO claims their technology is ahead of Google and IBM. IonQ’s CEO argues their “trapped ion” technology beats the superconducting chips used by Big Tech rivals, which he says suffer from manufacturing defects. This highlights a typical industry dynamic where founders aggressively promote their specific technology (often beyond what it can actually do at the moment) to capture the investor attention needed to fund development. Nothing new…