Editor’s Note: As of Monday, February 3, 2025, President Trump’s latest economic move has triggered volatility. Sweeping emergency tariffs targeting Mexico, Canada, and China‌ — ‌ranging from 10% to 25% across various sectors‌ — ‌are adding to uncertainty. Automotive, consumer goods, and tech industries stand to be impacted heavily. If tariffs stand, economists warn of higher consumer prices (inflation), lower GDP growth, and the risk of extended higher interest rates. While tariffs have long been a negotiation tactic, there seems to be a lack of clarity on what the US is actually asking for. For investors, the playbook remains unchanged: a well-balanced, diversified portfolio will be best positioned to navigate against uncertainty.

‍

The start of 2025 has brought an unexpected plot twist in the AI narrative, reminding investors that even the most compelling investment themes can face sudden disruption. But despite the turmoil, the broader markets held steady, with the S&P 500 finishing the month up 2.7% and the Nasdaq up 1.6%.

Let’s unpack a month that showed us why concentration risk matters and how innovation can come from unexpected places.

The AI Chess Game Gets a New Player

Just as markets were digesting the slew of (sometimes confusing) presidential executive orders, a surprising development from China sent shockwaves through the tech sector. Deep Seek R1, an AI model developed at a fraction of traditional costs to build advanced AI models, emerged as a serious challenger to established players like OpenAI (which built chatGPT), triggering a 16% drop in Nvidia’s stock and forcing investors to reassess their assumptions about AI infrastructure investments. If someone can build models at a fraction of the cost, why would they need expensive Nvidia chips?

What makes Deep Seek particularly intriguing isn’t just its performance — it’s the economics. Built under resource constraints due to US export controls, the team achieved comparable capabilities to leading models at roughly 1/20th of the cost. Even more significantly, they’ve made their research open source, allowing global verification and replication of their methods. Researchers at Berkeley (and others) have already had some success.

Rethinking the AI Investment Thesis

This development has sparked a fascinating debate about the future of AI infrastructure. While some see this as a threat to the current AI ecosystem, others, including Microsoft’s CEO Satya Nadella, point to Jevons Paradox — the counterintuitive notion that improving resource efficiency often increases overall consumption rather than reducing it.

Think about how cheaper data and faster internet didn’t kill tech infrastructure investments but instead enabled streaming services and drove more demand for data. Could we see a similar pattern with AI? Cheaper, more efficient models might unlock applications in industries previously priced out of AI, potentially driving even greater semiconductor demand across a broader spectrum of uses.

Winners and Losers

The market’s reaction has been notably nuanced:

• Infrastructure players like Nvidia and TSMC faced immediate pressure
• Energy and nuclear stocks tied to data center power consumption saw spillover effects
• The Magnificent Seven showed mixed responses, with companies like Amazon potentially benefiting from lower AI deployment costs

Perhaps the most vulnerable players aren’t the chip makers but the closed-source AI companies. If Deep Seek proves that powerful models can be built and freely distributed at a fraction of the cost, the business case for premium-priced proprietary models becomes increasingly challenging. Yet, even after all this news, we saw renewed confidence in OpenAI which is potentially raising its latest funding at an even higher valuation.

Broader Economic Implications

Beyond financial markets, cheaper AI could have disinflationary effects. Increased competition typically leads to lower prices and better products, potentially accelerating AI adoption across industries. While there could be shifts between big tech/AI winners and losers, businesses and consumers stand to benefit from more affordable, widely available AI tools.

From an economic standpoint, this could be disinflationary‌ — ‌not immediately, but in the long term. There are obvious concerns about what this might mean for the job market; however, history shows that technological advancements often lead to economic growth and job creation, despite initial fears of job displacement.

And in an even broader sense, ​​the potential for AI to address the systemic barriers facing women is extraordinary — this includes giving women access to guidance without judgment; helping them master new skills at your own pace; the ability to automate endless tasks that have drained women’s time and energy for decades. As journalists Erin Grau and Karin Klein explain, “this technology could be more than just another tool; it could be the great equalizer we’ve been fighting for, giving women the support, efficiency, and confidence that the prior systems have consistently failed to provide.”

Investment Lessons

The S&P 500 entered 2025 with its top 10 stocks making up 37% of the index‌ — ‌nine of which are Big Tech. US technology funds saw $23 billion in net inflows last year, the highest since 2000 (just before the dot-com bubble burst).

Monday’s market action was a reminder of concentration risk: while the S&P 500 was down, 70% of its stocks finished positive, and the equal-weight S&P 500 outperformed the cap-weighted index by the widest margin since July 2024. Investors should take note‌ — ‌high valuations for Big Tech assume near-flawless execution, and any disruption (like an AI breakthrough) can trigger volatility.

AI is evolving rapidly, and markets will be choppy as the dust settles. History is littered with examples of companies losing their first-mover advantage‌ — ‌Blackberry vs. iPhone, Blockbuster vs. Netflix, Yahoo vs. Google. AI isn’t immune.

For investors, the key takeaway is this: Be mindful of portfolio concentration and follow the principles of diversification. Markets reward innovation, but they also punish complacency. AI’s next chapter is just beginning.