Every morning lately, I wake up to a flood of new AI models and benchmark scores. Today OpenAI releases a new version, tomorrow Claude Code has a security incident. As an AI-native founder, I'll be honest—I've been experiencing serious "information anxiety."

Bombarded by all this daily, I can't help asking myself: if AI can truly do anything now, what do we still need to do?

Carrying this anxiety, I revisited a TEDx talk by Bill Buxton called Our Place in This World. After watching it, I felt like I'd been splashed with cold water—instantly sobered up.

His "information value pyramid" was a genuine revelation.

Which Layer Are We Lost In?

Buxton drew on the poet T.S. Eliot and combined it with sociological perspectives to sketch a five-tier pyramid: Noise → Data → Information → Knowledge → Wisdom.

"Where shall we find the wisdom we have lost in knowledge? Where shall we find the knowledge we have lost in information? Information is lost in data, and data is lost in noise."

Doesn't that perfectly describe where we are right now?

Today's large language models are incredibly good at generating "noise," "data," and even "information." Give one a prompt, and within seconds it produces a stack of professional-looking analysis reports. But the question is: does any of that actually have value?

So often, we're drowning in data while losing our direction. We mistake AI-generated "information" for "knowledge," and "knowledge" for "wisdom."

For instance, I once had AI help me write a business plan. It was beautifully written—logically rigorous, data-rich. But holding that plan, I felt no confidence. Why? Because it was just a pile of "information." It never told me what I should actually do.

Being Able to Do Anything vs. Knowing What to Do

This is the central challenge in Buxton's talk: "Now that we can do anything, what should we do?"

In the AI era, technology is no longer the bottleneck. We can realize almost any idea through AI: write code, generate images, produce videos, even auto-generate business models. But this is precisely when we're most at risk.

When "can we?" is no longer the question, "should we?" becomes the ultimate test.

We fall too easily into technological fetishism. See a new tool, and the first reaction is "Wow, this is amazing, I need to use it for something." But few people pause to ask: "What real problem does this tool solve for me? Could it create worse side effects?"

Buxton gave a striking example: if humans disappeared from Earth tomorrow, nature would rapidly self-repair—the world would become greener and healthier. But if fungi disappeared, the entire ecosystem would collapse immediately.

Our competitor isn't digital technology—it's fungi. Fungi simply need to exist to sustain the ecosystem; humans, in Earth's ecology, are essentially dispensable.

This is a form of "technological humility." We must acknowledge that our technology isn't as important as we think it is.

Searching for That 1% of Wisdom

So what do we actually need most in the AI era?

We need the "wisdom" to filter information.

When AI is mass-producing content, the scarcest resource is no longer content itself but judgment—a point that's been made many times but bears repeating.

Judgment to discern whether information is real or fabricated. Judgment to tell whether a tool is boosting productivity or just shifting attention. Judgment to know whether a project is creating value or manufacturing anxiety.

Take this very article. If I'd asked AI to generate a reflection on Buxton's talk, it could have produced something eloquent. But I chose not to. I chose to watch the video myself, digest it myself, and write down these genuine thoughts.

Because writing itself is a form of thinking—a process of distilling wisdom.

If you outsource that process entirely to AI, what you get will only ever be "information," never "wisdom."

Next time you see some hyped-up new AI tool, take a deep breath and ask yourself: "Where does it sit on the pyramid? Can it actually bring me wisdom?"

If not, let it go. What we truly need is to stay clear-headed and know what we "should" be doing.