What Has Large Language Models Really Changed?
What Has Large Language Models Really Changed?
Lately, I’ve been pondering a question: What have large language models really changed?
Sometimes, I feel their impact runs deeper than what meets the eye. In the past, there was always a certain “rhythm” to how we worked. A proposal took days to prepare, a process took weeks to complete, a document was written slowly by one person. Time and space acted as natural constraints, framing everyone within their boundaries.
But with the arrival of large language models, many of those rhythms have been abruptly disrupted. They can generate a first draft in minutes, handle multiple tasks simultaneously, and push forward tasks that once required queuing, waiting, and repeated confirmations. The old logic of “time must be spent” seems to have been rewritten.
This reminds me of a saying: Nietzsche declared “God is dead,” meaning the old order had collapsed and humanity had to face the world anew. Similarly, the emergence of large language models seems to be telling us: many old orders have been leveled, and we need to rethink everything.
In my work, I’ve already felt this shift. Previously, a project proposal required several people to discuss, iterate, and finalize. Now, I first feed the model a version of the material, get a direction immediately, and then bring the output to the team for discussion—the pace is completely different. Before, writing a formal email meant carefully weighing every word; now, the model can outline a framework for me, and I only need to polish it based on the context. That “rhythm” has truly been reconstructed.
Sometimes, I’m even taken aback: processes we once thought irreplaceable, time we believed uncompressible, have suddenly become flexible. This forces me to reconsider: if efficiency is no longer the bottleneck, how should we define our own value? If machines can handle all “quantifiable” work in parallel, what can we, as humans, still contribute?
I don’t have answers to these questions yet. But what I am certain of is that large language models are not just about saving us a bit of time—they have completely overturned our understanding of “time, space, and work rhythm.”
They’ve made me realize that we are standing at a new threshold. Many things now require us to look again, to understand anew.
Originally written in Chinese, translated by AI. Some nuances may differ from the original.
