Introduction|The situation of Taiwan’s 00後: fluent with the technology, new to the workplace
On the generational spectrum of Taiwan’s workplace, people born in 2000–2009 (the 00後) are the first generation of “digital natives” to have lived fully inside mobile internet and the platform economy. They completed their socialization through LINE, YouTube, Instagram, TikTok, and Dcard. Taiwan’s 108 Curriculum Guidelines also strengthened training in “inquiry and practice” and in self-expression. But as they step into the workplace, they happen to run into the outbreak of generative AI and multi-agent systems. For them, a tool is only a “language interface” — not a remote technology. What truly makes a person uneasy is that, inside the old rules of process, permission, governance, and responsibility, they find themselves like a sandwich layer that is “neither the lead, nor a supporting role.” In this piece I try, in a Taiwan context, to clarify a place and a path for the 00後: how to translate a native sensitivity to tools, and a multi-tasking rhythm, into a system capability the organization can understand, and is willing to give authority to.
NO° 01|The generational background of the 00後: raised on the internet, entering at intelligence
If the 90後 are native to platform logic, the 00後 are native to the semantic interface. For them, ChatGPT is a tool you can talk to; Midjourney, Runway, and CapCut are expression that can be outsourced; Notion and cloud collaboration are the everyday environment. Input and output run in natural language, and iteration is as ordinary as breathing. Technology, then, is not sacred; tools are not scarce; action is learning, and “try it first” is the basic rhythm.
The problem is that how an organization runs has not become simpler just because tools became easier to use. Most companies in Taiwan still have a clear hierarchy of permissions, a compliance process, and nodes of cross-department reporting. When the 00後 first enter the workplace, they often project “can’t I just finish this myself?” onto a real team: they dislike waiting in line for a sign-off, they lean toward “short-chain output,” and they are used to treating AI as a tacit teammate — while overlooking context, the chain of responsibility, and traceability, elements that cannot be omitted in the language of an organization.
The remote learning of COVID-19 enlarged this break still further. The 00後 are good at teaching themselves and at collaborating online, but they have little experience of the rhythm of “alignment before a meeting,” “trade-offs in the meeting,” and “handover after the meeting,” and still less patience for “why do we have to follow this node.” The result is that they charge fastest at the front of an AI transformation, and produce a great deal, and yet often lose speed on “the achievement the team can see.”
NO° 02|The positive effect of AI: an efficiency amplifier in its element
First, a multiplier effect on high-frequency tasks. The 00後 can pack writing copy, revising a résumé, generating a script, image synthesis, transcribing subtitles, and scheduling distribution into a “mixed task pack,” and, with a few tools and a few prompts, run out a respectable first version. For roles in content, community, customer service, assistance, and junior analysis, this muscle of “start at once — iterate quickly” makes them the most flexible front end of a team.
Second, a natural advantage in multiple threads. The 00後 can open ten tabs at once: talking with a model, arranging Notion, revising a poster, uploading assets, and answering a viewer’s comment. This non-linear, multi-channel collaboration happens to fit a high-concurrency workflow under AI support: a person is responsible for definition, the standard of judgment, and a change of gear; AI is responsible for throughput, tracking, and continuous output. Together they make an overall efficiency that exceeds the optimization of a single point.
Third, the prototype of a new kind of role. In more and more Taiwanese companies, entry roles have appeared such as “prompt / workflow assistant,” “RAG data curator,” and “agent-orchestration helper.” The 00後 may not yet have a grasp of strategy and governance, but they most understand “how to get AI running”: feed the material cleanly, connect the nodes smoothly, and escalate the exception. This is the best forward post from which to enter the practice of an AI transformation.
NO° 03|The risks under AI: a short-sighted loop × structural aphasia × stacked anxiety
The first risk is a hollowing-out of business understanding. Tools lower the barrier to entry, and also make it easier to skip the prior thinking of “whose need? why now? what is the definition of success?” You can make a chart but cannot speak an insight; you can produce a plan but cannot set a metric. Over time, the market defines the capability as “task outsourcing,” and the value is stuck at a replaceable bottom.
The second risk is the absence of the ability to collaborate. A real organization does not reward an “efficiency island.” Reporting clearly upward, aligning sideways, and taking responsibility in a division of labor are all tickets to an upgrade. When the 00後 hand all communication to a tool, and outsource all judgment to a model, the team finds it hard to form an expectation of, and a trust in, their output.
The third risk is a thinning of the sense of meaning. AI makes “task—delivery—end” so fast that it loses focus: ten posters today, twenty short films tomorrow, a hundred FAQ entries the day after. When output is like a production line, and feedback is fragmented, it is easy to fall into a nihilistic loop of “doing a great deal, and feeling nothing.” What is not being trained is the narrative muscle that connects the work to a value.
NO° 04|A suggestion for a breakthrough: from user of a tool, upgrade to “designer of the work”
The first step is to restore the task to a problem. Before every need, first answer four Ws: who wants it? why? when is it due? how will it be accepted? Write these four answers into the prompt and the workflow, and let AI run by “the context of the business,” not by “keywords.” When you can offer a success metric and a boundary for exceptions on your own, you are already standing in the place of a teammate, not a “production add-on.”
The second step is to practice the language of the organization, and the rhythm of collaboration. Fix a cycle: before a meeting, use one page to align the goal and the style of the output; in the meeting, make trade-offs with an observable metric; after the meeting, write success and failure back into the knowledge base and the workflow. Upgrade yourself from “the person who makes the thing” to “the person who can receive the matter”: able to receive a need, able to receive a challenge, able to receive the next baton. This is scarcer than being fluent in ten new tools.
The third step is to treat AI as your growth operating system. Do not only use it to revise a slide deck or write a report; let it become your learning coach and your review assistant: use it every day to arrange reading notes, ask it every week to help you review output and blind spots, and every month arrange a portfolio and a workflow into a “system résumé.” When your rhythm of growth is systematized, your bargaining power in the market will be more persuasive than “the number of works.”
NO° 05|Closing: the 00後 are not a transition. They are a generation of restart
The age of AI has cancelled the navigation many people were used to: no longer waiting for the company to arrange it, no longer relying on a single path. Everyone has to become the designer of their own career, their own tool engineer, and the manager of their own rhythm. For the 00後 the pressure is, to be sure, not small. But you also hold, at the same time, the two scarcest resources of this age: an affinity for tools, and a tolerance for iteration. The next step is not a contest of who uses it faster. It is a contest of who thinks more deeply, who receives more steadily, and who can write it into the system. The step from “using up the tool” to “designing the system” is exactly the starting point from which you walk out of the sandwich layer, and take the right to speak.
Topic | Five Generations under AI (AI洪流下的五代人)
Five Generations under AI, Future
Five generations in concert: how five workplace operating systems can be written into one cross-generational protocol Five Generations under AI, Future
Five Generations under AI|The 10後: finding a place in a “world without jobs” Five Generations under AI, Future
Five Generations under AI|The 00後: how does the “sandwich layer” of the AI transition break through? Five Generations under AI, Future
Five Generations under AI|The 90後: meet AI head-on, and shape the future with the machine Five Generations under AI, Future
Five Generations under AI|The 80後: standing on a fault line, how to cross the mid-career bind? Five Generations under AI, Future
Five Generations under AI|How can the 70後 place their experience in the age of AI?
The judgment, restated
My judgment is that the 00後 are stuck in a sandwich layer — they have to understand the system, and also make the system able to see their contribution.
Three things this week
- Make a stretch of “system résumé”: which reusable workflows have you designed.
- Practice context communication once: write a need so that an agent and another department can both read it.
- Pick one sandwich-layer task to automate, and leave a human–AI review point.
The series overview is in Five generations in concert. You are welcome to subscribe to Future Insights Lab, or to reply and discuss the sticking points of the sandwich layer.