Why Doesn’t AI Understand Context? An Insight from The New York Times

Why doesn’t AI understand context? Martin Heidegger famously argued:

“The essence of technology is nothing technological.” — The Question Concerning Technology

What, then, is its essence? Its essence is “enframing”— a way of giving us a particular lens through which to view the world. Modern technology is Gestell: a mode of disclosing reality, not a tool. It reveals reality as “standing-reserve.”

But is this true?

The New York Times recently published an article titled “Companies Are Pouring Billions into A.I. It Has Yet To Pay Off.” And Forbes echoed it with an article: “Companies Are Pouring Billions into A.I. Here’s Why They’re Not Seeing Returns.” Why is this happening?

The New York Times emphasizes the human side: employees resist tools they do not trust. Forbes zeroes in on technical issues: AI still fails to understand the context of work. Surprisingly, the solution proposed in both cases is itself technical in nature —training AI to “understand” context.

But can AI understand context? Context is what surrounds the text — the background that allows us to understand the true meaning of words, events, or ideas. How do you train AI to understand context? Companies tend to propose only one solution: feed it more data.

Yet, reducing context to data is precisely what Heidegger calls enframing. Context is not data; it appears as data when we look at it through a technical lens. But what is context proper? The word context comes from the Latin con (“with, together”) and texere (“to weave”). Literally, context means “a weaving together.”

To understand context, we must be weavers. It’s more art than science. To truly understand my friend’s words, I must artfully weave the individual threads of what I know about them into a single, meaningful picture. A weaver doesn’t simply assemble the picture from bits and pieces — they weave a tapestry from disparate threads based on the vision of the Whole.

A weaver cannot produce a coherent Whole unless they have first seen the Whole. True art is recreating on earth what we saw in heaven. We can only weave what we have seen — the Heavenly Pattern. AI cannot see the Heavenly Pattern. And because it cannot see, it cannot truly weave. That is why it cannot genuinely understand context.

All humans are Platonists by nature — we instinctively grasp the Idea behind every thing. AI cannot see Platonic Ideas, and therefore it cannot weave. Reality is not data; it is textile — fabric. It is not assembled from bits and pieces but woven from Logos-colored threads, revealing something that comes from beyond this world.

The employees in those companies are human, and they instinctively sense what AI cannot. That is why they distrust it. Will this ever change? To be able to weave — to grasp context — you must be able to see the invisible.

Reality is not data. It is fabric. It is a tapestry — a visible image of what cannot be seen by the physical eyes. To dwell in the world contextually means to artfully weave on earth what has been revealed in Heaven.