The DeepSeek Upheaval: Why Human Wisdom Still has the Edge

Having spent decades in IT leadership, I’ve learned that innovation thrives not under endless budgets, but under constraints. Whether it was optimizing cloud costs during the spending frenzy of the 2010s or rebuilding security frameworks without breaking the bank, our teams succeeded by refusing to accept “how it’s always done.” We challenged assumptions and focused on outcomes, not tools. That’s why DeepSeek’s recent AI breakthrough—training a ChatGPT-level model for $5.6 million (1% of OpenAI’s costs)—feels so familiar. Like my teams, they turned limitations into fuel, using architectures like Mixture-of-Experts to slash computational waste. Proof that strategy can still beat brute-force spending—no trillion-dollar budgets required.

Yet even as Sam Altman argues “figuring out what to ask will eclipse knowing answers,” Silicon Valley often misses a critical truth: direction matters more than speed. Without resourcefulness, you’ll never course-correct when plans falter. When COVID forced us remote overnight, we didn’t just deploy VPNs; we asked, “How do we keep trust and efficiency intact?” We turned crisis into opportunity—streamlining workflows and emerging stronger. When ERP upgrades loomed, we didn’t outsource blindly. Instead, we asked, “What defines success for our people?” and rebuilt processes from the ground up. DeepSeek’s engineers mirrored this mindset: “How do we make AI efficient, not just big?” Their answer—open-sourcing models like DeepSeek-R1—exposes the fallacy of the “dollar wars.” The real race is for wisdom.

But here’s where machines fall short: AI lacks the human experience. DeepSeek’s models, while cheap, can’t weigh ethics, build trust, or rally teams in a crisis. They can’t spot hidden risks in unpredictable environments or ask, “Does this align with our values?” As AI commoditizes intelligence (DeepSeek’s API costs $0.14/million tokens), the human edge—judgment, empathy, grit—becomes irreplaceable.

The market’s panic over DeepSeek crashing Nvidia’s stock by 15% feels eerily familiar—another “cloud cost hysteria” moment. The U.S. Stargate AI project’s success hinges on something deeper: work ethic forged by purpose, not panic. True leaders see upheaval as a mirror: What do we value?

We need to be ok with a new framework. A framework where AI is AI, machines are machines and where humans… human. Now as humanity, we better figure out what that is, and fast.

So let’s not fear AI. Let’s fear losing sight of what makes us human: teams built on trust, guided by integrity, and relentless in asking, “Does this align with who we are?” The greatest tools I’ve seen weren’t the smartest or shiniest—they were shaped by humans solving the right problems, together.