Running-Different-Races

Key Capital Private, Investment Note - Issue #35

Running Different Races

Late last year, we wrote about how the scale of the AI buildout is tying the disparate parts of the US stock market together in unexpected ways. It's a theme we're likely to return to more than once this year, given how central it has become to markets.

Unsurprisingly, most market outlooks for 2026 spent considerable time on AI, including ours. At our annual outlook event (linked below), we tried to put some context around AI in the US and how OpenAI could be the largest single point of weakness. One area we didn't touch on was China, and how that market has evolved post-DeepSeek. We address that below.

Just as it is everywhere, assessing how the AI industry in China is evolving can be difficult. In some ways, the Chinese market is incredibly transparent. Since 2023, every publicly facing AI model must be filed into a publicly available registry before launch, giving a window into the AI ecosystem. But equally, the CCP have apparently “asked” DeepSeek engineers to hand in their passports. What is clear, however, is that China is taking a markedly different approach to AI than the US.

The US, via Silicon Valley, is obsessed with the concept of Artificial General Intelligence or AGI. In 2019, OpenAI CEO Sam Altman gave the following quote in an interview about OpenAI’s business:

“We have no current plans to make revenue. We have no idea how we may one day generate revenue. We have made a soft promise to investors that once we’ve built this sort of generally intelligent system, basically, we will ask it to figure out a way to generate an investment return for you”

One Bloomberg columnist described this as “the greatest business plan in the history of capitalism: 'We will create God and then ask it for money.”

The framing of AI as a panacea is pervasive in Silicon Valley and helps explain why tech CEOs are betting the farm on building AI; they fundamentally believe that AGI will be the invention to end all inventions. AI 2027, a dystopian AGI manifesto created by former OpenAI and other AI researchers, is a vivid example of this fixation within US AI circles. Outside of Silicon Valley, the fixation is the same; the White House’s AI action plan describes AI as “An industrial revolution, an information revolution, and a renaissance - all at once”.

Contrast that with Chinese Premier Li Qiang’s speech at the World AI Conference in Shanghai, which focused on practical AI applications, and the Chinese State Council’s ‘AI+ initiative’, which aims to integrate AI into 70% of the economy by 2027 and 90% by 2030. That narrower focus is reflected in the Chinese AI registry. According to an analysis of 6000 records of generative AI models by analysts at the Oxford China Policy Lab, Chinese AI companies are more focused on building specialised models for specific commercial applications rather than cutting-edge frontier models: with healthcare, retail and media seeing large numbers of domain-specific models in the registry.

That narrower focus is partly defensive, given China’s constraints. Venture funding for AI startups in China halved year-on-year in early 2025. The broader market in China isn’t firing on all cylinders, real estate is still recovering from its collapse, youth unemployment statistics are now a state secret, consumer confidence fell off a cliff in 2022 and has yet to recover.

Facing those constraints and a US chip embargo, the Chinese government has tacitly acknowledged that it cannot win a frontier-model arms race. China have instead adopted a whole-of-nation approach, attempting to make up for its shortcomings by integrating AI more deeply into the economy. This integration isn’t just about AI chatbots and software; it is enabled by hardware dominance in energy and manufacturing.

China’s ability to deploy ‘good enough’ AI into complementary industries, like industrial robots, electric vehicles and drones, is world-leading. In 2024, China had more than five times as many factory robots in operation as the U.S. (2,027,200 to 393,700), and almost half of Chinese manufacturing equipment sold that year incorporated some form of AI (be it machine vision, predictive maintenance or autonomous-control functions).

Electricity, on which AI relies, is another area where China leads the world. In 2024, China generated 31% of global electricity output - double the US’s output. They are currently building more nuclear plants than the rest of the world combined, and, unlike the US, China has ample spare power capacity. According to Goldman Sachs, spare capacity in China will grow through 2030 despite the rapid increase in power demand.

Even in technological innovation, China may well catch up to the US. China graduates 50% more STEM PhDs than the US and is generating 70% of global AI-related patents.

Conclusion

As evidenced by this week’s software crash, AI will continue to shape markets. The prevailing narrative, though, that the US and China are in an “AI race”, is misplaced. It's an understandable misplacement; the White House’s AI action plan is literally titled “Winning The Race”, and it echoes back to the Cold War space race. The space race, however, had a clear finish line, putting humans on the moon, while the end goal of the AI race will differ depending on who is being asked.

The US is increasingly betting on AGI as the ultimate goal, while China is focusing on broad and deep integration into the real economy – absent geopolitics, these are complementary, rather than competing, paths.

Key Capital Annual Outlook Event:

Key Capital's Annual Market Outlook in Dublin was held over two sessions, breakfast and lunch, on January 28th, and was presented by Ian Kilcullen, Managing Director of Key Capital Private.

We reflected on the past year and gave attendees a sense of how we interpret the market movements that unfolded in 2025. It was also an opportunity to look forward to the year ahead and the key issues that are likely to impact markets.

This year’s presentation covered a wide range of themes, but the central narrative that dominated was the role AI now plays in the US equity market and the US economy.

The US economy is heavily reliant on AI Capex spending, the US stock market's performance is heavily reliant on AI-related names, and US consumer spending is being supported by the top 10% who are feeling wealthy, largely due to the US stock market's performance. This creates a significant level of interdependence, and a lot of it depends on AI capex spending being maintained and ultimately a clear earnings boost from all this spending. A high level of interdependence within any system increases fragility.

Please find attached a pdf of the slides, a couple of topic specific video edits and also the full video link.

Video 1 - 8 minute edit. LINK: AI CapEx Spending and AI Financing

Video 2 – Full presentation. LINK: Full Outlook 2026 Presentation

PDF LINK: Outlook 2026 Slides

Key Capital & Schroders Event Video:

Key Capital and Schroders recently hosted an event on the Key Capital Balanced Multi Strategy (BMS) Fund.

• BMS Overview: Ian Kilcullen, Managing Director of Key Capital Private, gives an overview of the BMS Fund and how it differs from traditional multi asset funds. - 14 minutes.

• Panel Discussion: A discussion with Rishi Sivakumar, BMS Fund Manager at Schroders, and David Rees, Head of Global Economics at Schroders, on the diversification illusion at the core of most multi-asset funds and how the BMS Fund seeks to address this for Irish private clients. - 14 minutes.