AI Is More Than a Writing Tool
Many people fall in love with AI through writing.
Drafting reports, outlining structures, polishing expressions, building slides—when words start flowing and you can’t stop them, the sense of momentum is exhilarating.
Looking back at my own experience with AI over the past two years, the efficiency gains have been exponential. I’ve never written a grand, once-a-year keynote speech for senior leadership, but for a typical thousand-word internal report, the old workflow—information gathering, cross-checking, drafting, revising—used to take two or three full days. Now, with the same standards and expectations, half a day is often enough.
So here’s the inevitable question:
if AI can write faster and reasonably well, what exactly are we—the so-called “pen holders,” writers, planners, and communication professionals—still here to do?
There is a common misconception that AI-generated writing is unreadable or soulless. But once you truly integrate AI into your workflow as a collaborator rather than a novelty, you quickly realize something else: the clearer you are about what you want, the better the output becomes.
This is because most writing within organizations is not creative writing. It is not about self-expression or emotional catharsis. It is about packaging data collected across different levels and departments, structuring it coherently, aligning it with strategic priorities set from above, and matching the rhetorical style of leadership—then assembling all of that into a document that may only be a few hundred kilobytes, but carries real weight.
What you produce is not your personal will. It is an organization’s interpretation of its past achievements and its future direction.
When we say “writing materials is exhausting,” what we are really talking about is the exhaustion of accuracy. As an individual with opinions and perspectives, how do you balance senior leadership’s intent, peer departments’ positions, frontline audiences’ reception, historical consistency, and the risk of future accountability?
That is why major reports are often filled with carefully calibrated, ambiguous phrases. They are not laziness; they are the result of deliberate risk management. Even the invention of new terminology often serves this purpose—though that is something I won’t expand on here.
After all this, it may sound as if not using AI to write is almost heretical. But my point is the opposite: using AI only for writing is a waste.
I am not suggesting everyone should immediately jump into vibe coding, start building products, or force themselves onto the latest technological wave. Everyone has different problems to solve, and not everyone needs to chase trends at full speed.
What matters more is that AI allows us to reduce cognitive load—and redirect our attention toward things that actually shape the real world.
That sounds abstract, so let’s return to reports.
In the United States, there is the State of the Union address. Historically, similar practices existed long before: in ancient China, feudal lords reported their governance to the king; in imperial dynasties, officials submitted performance accounts to the emperor. In 1790, George Washington delivered the first State of the Union. While inspired by the British Speech from the Throne, its audience was Congress—and, more broadly, the public.
What were they really doing?
What problem were they solving?
Times change. Media change. But the core task of writers has never changed: persuasion.
I write so that executors can identify key signals, media can extract headlines, experts can interpret meaning, competitors can assess implications, and the public can feel reassured.
Whether we are drafting major reports, explaining strategy to employees, or communicating with the public, the process is essentially the same: structuring messy, unorganized information, then choosing the right way to persuade specific audiences.
Persuasion is not achieved by making text “beautiful.” It requires judgment, an understanding of boundaries, and sensitivity to needs.
AI eliminates much of the labor of expression—but it amplifies the importance of judgment and aesthetic discernment. These are the aspects that actually influence reality.
We are finally able to focus on deciding what deserves to be included in a narrative, what can be omitted, and which form of expression will be most effectively understood.
These decisions come from your interaction with the real world. They are also where AI cannot generate content for you, and cannot bear responsibility on your behalf.
This is where I want to focus future writing:
when polished text—and even multimodal content—becomes instantly accessible, what role remains for those whose job is to produce it?
As organizations, how do we use AI not only to assist creativity, but to inform long-term development paths?
As individuals, are we destined only to marvel at AI’s outputs—or quietly wait to be replaced by something more advanced?
Do we continue competing with AI on execution speed,
or do we step into the role of judgment and choice?