America’s Historic Focus on Fundamental Research

One of the most remarkable things about American companies and society is their deep focus on fundamental research. This academic pursuit and tradition have shaped industries for decades. Universities in the United States have long been leaders in cutting-edge research. But what makes America unique is how corporate giants have also played a crucial role in advancing science.

Even decades ago, companies like GE, Bell Labs, and IBM invested in research far beyond their immediate business needs. Bell Labs, for example, pioneered the transistor, which became the foundation of modern computing. IBM pushed the boundaries of computer science, leading to breakthroughs in artificial intelligence and semiconductors. This DNA of long-term scientific exploration became a defining feature of American technology companies.

The Corporate Quest for Scientific Discovery

Technology companies took this to a new level. Instead of funding research relevant to their business, they became involved in scientific quests without immediate commercial application. IBM led in artificial intelligence and semiconductor advancements. Over the last two decades, individual tech leaders like Elon Musk and Jeff Bezos have invested heavily in space exploration, a field far removed from their primary businesses. However, these ventures were not a part of their flagship companies.

But the most striking example of corporate-driven scientific research came from Google. Around 15 years ago, Google aggressively approached research through acquisitions like DeepMind and D-Wave. These investments were not aimed at selling search ads or cloud services. Instead, they led to groundbreaking developments like AlphaGo, the attention mechanism that underpins modern AI, and AlphaFold, which solved the long-standing protein folding problem. Google’s deep commitment to research reshaped industries far beyond tech.

Microsoft’s Shift Toward Fundamental Research

For a long time, Microsoft took a different approach. While it always had a strong research division, its focus was more applied—working on projects that, while advanced, had clear ties to its software and cloud businesses. Unlike Google, Microsoft did not seem eager to explore areas with no immediate commercial value.

This is where NVIDIA provides an interesting contrast. NVIDIA’s research has always been product-driven, focusing on making its GPUs the backbone of AI, gaming, robotics, and biotech. The company developed CUDA, an essential tool for AI computing, and expanded into areas like robotics with Isaac, biotech with BioNeMo, and materials science with simulation platforms. Microsoft, until recently, had a similar applied-research mindset.

However, this has changed dramatically in the last few years. Microsoft is now aggressively investing in fundamental research, much like Google. The shift is evident in its recent work in AI-driven protein design, materials science, and quantum computing. Microsoft is no longer just improving Windows and Azure—it is building a research powerhouse that rivals Google. And the announcements over the last few weeks are a testament that it is getting close across fields.

Microsoft’s Big Moves in Fundamental Research

The clearest sign of this shift came in just the last four weeks. Microsoft made three major research announcements that signal its growing commitment to deep science:

1. MatterGen for Material Sciences – Microsoft announced MatterGen a month ago through a paper in Nature. In its own words: " MatterGen, a generative AI tool that tackles materials discovery from a different angle. Instead of screening the candidates, it directly generates novel materials given prompts of the design requirements for an application. It can generate materials with desired chemistry, mechanical, electronic, or magnetic properties, as well as combinations of different constraints."
2. AI in Protein Design – Over the weekend, Microsoft revealed an AI-powered platform to assist in protein engineering. Unlike most AI tools that predict a protein’s fixed structure, Microsoft’s platform simulates how proteins move, flex, and transform over time. This dynamic perspective—think of it as watching a protein “dance” rather than just seeing a still image—could lead to better drugs and custom-designed proteins for medicine or industry.
3. Quantum Chip Development – Last week, Microsoft announced its first quantum processor, Majorana 1, aiming to build more stable and scalable quantum computing architectures. It has introduced new types of Qubits, called topological Qubits, which are fundamentally different from superconducting transition qubits used by established houses like Google and IBM. They are based on completely new material, coined topoconductors by Microsoft. 

These are neither isolated experiments nor the only announcements. This weekend, Microsoft also announced Azure Foundry Labs with the “World and Human Action Model,” or WHAM, to help video game developers. While this last statement is about a development directly linked to some of Microsoft's existing businesses, collectively, these announcements represent a fundamental shift in Microsoft’s strategy—investing in long-term research that may not have immediate commercial returns. 

Interdisciplinary Efforts a Must in the New Innovation Era

For most of its history, Microsoft was known as a follower rather than a pioneer. It refined business models rather than creating new ones. In AI and cloud computing, it trailed behind Google and Amazon before catching up. But something has changed.

Like other tech giants—Alibaba included—Microsoft is making unprecedented capital expenditures. Despite concerns about the lack of near-term returns, more business leaders now realize that merely using AI to enhance existing offerings, as many software firms do, is not enough. True, lasting revenue streams require exploring new frontiers beyond their traditional expertise.

Innovation is becoming more interdisciplinary. Tech companies are now leading innovation in industries where traditional players have been slower to act. Take the example of the sleeping pharmaceutical giants: when only Google was working on AI-driven protein research, the pharmaceutical incumbents were perhaps right to be unbothered about any business risk. But now, with NVIDIA, Microsoft, and others entering these fields in proteins and across other segments, established industry leaders risk being completely disrupted. This risk is equally true for those making traditional automation equipment, consumer appliances, automobiles, or any product or service industry that can have different offerings with a different type of thinking.  

Finally, the accelerating pace of innovation itself must be noted. Microsoft’s eye-catching innovations are within a few quarters of the announcement into deep fundamental research with substantial resources. A few have noted how quickly xAI, which was not in existence two years ago, is able to develop world-beating chatboxes. A better indicator of the pace of innovation is from Microsoft’s announcements, where it can impact fields where it had no expertise. Unfortunately, analysts and media are more interested in capex numbers rather than such substantial outcome progress (until everyone begins to talk about them like in the case of DeepSeek). At least there is one change as of this newsletter as evidenced by Alibaba’s announcement last week: when a company announces large capex in innovation in Feb 2025, markets like it. It was clearly not the case outside the US last year.