AI Agent · Enterprise Adoption · Public Viewpoints Summary

From Answers to Action: How I Think About Enterprise AI Agent Adoption

When AI moves from answering questions to assisting in completing tasks, enterprises truly need to design a workflow that can be delegated, checked, taken over, and continuously improved.

What This Article Is About

I have long appreciated Professor Jian Lifeng's perspective on AI. He can examine Taiwan's broader industry and technology trends while also returning to the practical issues enterprises face every day, including work, talent, and adoption. His macro view and detailed practical observations coexist.

I especially appreciate his recent explanation of AI agents. AI used to stop at answering questions; now it is moving toward planning, using tools, and completing tasks. Professor Jian's phrase, 'From Answer to Action,' captures this shift clearly. I reviewed several of his public talks and reports and compiled this article around a practical enterprise question: when AI starts doing work, how should we integrate it into a workflow?

Clarifying Boundaries: This is my summary and practical extension based on public sources. It is not a verbatim transcript of a lecture and has not been reviewed by Professor Jian Lifeng. The article distinguishes between his public viewpoints and my framework for enterprise adoption.
Who is this for
  • The company has allowed staff to use generative AI, but the results are still limited to summaries, Q&A, and personal efficiency.
  • Managers want to adopt AI Agent but don't know where to start with the first workflow.
  • Teams are worried about data, permissions, errors, and responsibility, and want to start with a controlled pilot.
What you can take away
  • A plain-language framework for understanding Answer to Action.
  • A screening table for choosing the first Agent scenario.
  • A six-step process and 90-day roadmap for enterprise small-scale pilot.

Answer to Action changes more than just a slogan.

In the past, when we used generative AI, common actions included asking questions, looking up information, summarizing, and generating suggestions. AI provided the answers, and the subsequent judgment and execution were still handled by humans.

As AI moves toward Action, it begins to participate in a task: understanding the goal, reading the context, planning steps, using tools, producing results, and then handing over the parts that require judgment back to humans. This is where AI Agent becomes truly important for enterprises.

In a public talk at the 2025 Future Commerce Expo, Professor Jian Lifeng summarized the evolution of large language models as a move from Answer to Action. He noted that agents can understand, reason, collaborate, and execute tasks. He also placed AI within workplace relationships by discussing human-machine collaboration and AI colleagues. This advances the enterprise adoption question: we are designing both a good answer and a process that can be delegated, checked, and taken over.

Public Viewpoints AI moves from answering questions to taking proactive action, and Agents begin to participate in understanding, reasoning, collaboration, and execution.

Differences Between Answer and Action Tasks

AspectAnswer StageAction Stage
Main TaskAnswering, searching, summarizing, giving suggestionsPlanning, using tools, advancing the task, reporting results
Human RoleHandling after asking a questionSetting goals, authorizing, checking, taking over exceptions
Required DataPrompts and current contentHistorical data, rules, templates, system status
Main RisksAnswering incorrectly, answering incompletelyMake mistakes, overstep authority, leave incorrect results
Complete evidenceA segment of responseCheckable output, records, and handover status

Agent workflow, which can be considered as six actions

I have organized the understanding, reasoning, tool usage, and human-machine collaboration mentioned in public sources into a workflow that is easy for enterprises to discuss.

STEP 01Receive the target

What needs to be completed and who will receive the results.

STEP 02Read the context

What data, rules, templates, and historical records are needed.

STEP 03Break down the steps

Break the task into checkable stages.

STEP 04Use tools

Query data, update systems, or produce documents within authorized scope.

STEP 05Report results

Clearly explain what has been completed, what is still missing, and any exceptions.

STEP 06Return judgment

Have people confirm, correct, or stop at key nodes.

These six steps are my enterprise implementation framework based on public sources. They are not Professor Jian Lifeng's original wording. Their purpose is to turn “we want to introduce an agent” into a process that can be designed and verified.

A simple example of an administrative process

Taking the initial review of supplier data as an example, the Agent can first read the form and attachments, check for missing items based on existing fields, summarize the findings, and then pass the missing items and questions to the handler for confirmation. The first version will focus on reading, comparing, organizing, and submitting for review, without directly approving the supplier or modifying the official main file.

Companies can start measuring the missing item rate, organization time, manual review volume, and common exceptions. These results will be closer to the decision-making process than just "everyone thinks AI is powerful."

Companies can start with four types of scenarios

Knowledge and Administration

Organizing data, comparing documents, producing drafts, preparing meeting summaries, and answering internal questions with clear sources.

Software and Data Work

Assisting with programming, testing, data cleaning, and repetitive processing, while retaining code review and formal deployment permissions.

Customers and Business

Preparing draft replies, organizing customer needs, comparing product conditions, so that people can focus their time on relationships, negotiation, and exception judgment.

Operations and Processes

Tracking task status, checking for missing items, comparing rules, and preparing procurement or inventory information.

My Organization The first scenario does not need to pursue the most imaginative one. First, select a real, frequently occurring, quantifiable, and manually manageable process.

The first scenario should be small enough and true enough

Many companies start by looking for the most powerful Agent. I will first check whether this scenario can be safely learned.

Filtering IssuesSignals Suitable for Starting
Does it frequently occur?Occurs every week or every day
Is the input accessible?Files, fields, templates, and rules can be found
Can the output be checked?There is a clear format, standard, or responsible person
Can errors be taken over?Can the process be paused and handed back to human when errors occur?
Can the benefits be measured?Can time, quality, cost, or exceptions be compared?
Can permissions be reduced?The first version only gives access to what is needed to complete the task

If four to five of the six questions can be clearly answered, the scenario is worth entering a pilot. If even the input, output, and responsible person are unclear, introducing tools usually only amplifies the originally vague process.

First complete a small pilot

STEP 01Define the outcome

Write down what you want to see after completion, for example, a summary, missing item list, or pending confirmation replies.

STEP 02Organize data and rules

List files, templates, fields, judgment rules, and update responsibilities.

STEP 03Break down small processes

Each step clearly states what was read, what was done, and what was left.

STEP 04Put into human verification

Decide who confirms, when to stop, and which exceptions must be handed over.

STEP 05Compare the actual results

Review quality, time, cost, exceptions, and manual review volume together.

STEP 06Feedback learning

Feedback errors and exceptions to data, rules, templates, and role assignments.

Pilot acceptance sheet

  • Result quality: accuracy, completeness, format consistency.
  • Time: manual operation time, AI execution time, waiting for review time.
  • Cost: model, system, setup, and maintenance cost.
  • Exceptions: which type of situation is most frequently transferred to manual handling.
  • Permissions: whether unnecessary data or tool access occurs.
  • Traceability: whether it is possible to reconstruct what it read, what it did, and who confirmed it.
  • User experience: whether colleagues understand when to use, when to take over, and how to report issues.

Data, talent, and guardrails determine whether an Agent can actually work.

Data and systems

For an AI to complete a task, it must be able to read the company's data and use tools under clear permissions. Companies can start by inventorying readable content, structured data, necessary tools, update mechanisms, and minimum access scope.

As future searches and business processes involve more AI, companies should also make product catalogs, specifications, inventory, and transaction conditions easier to be consistently understood. This is a practical preparation direction I derived from public discussions on Agentic Web, conversational commerce, and AI procurement.

Human Role

One of Professor Jian Lifeng's public observations strongly resonates with me: when AI starts participating in work, people should focus on asking questions, exercising judgment, integrating information, and assuring results. Public reports describe this shift as moving from Work Faster to Work Smarter.

Six basic guardrails

  1. Quality standards: what results are acceptable, which fields must be checked.
  2. Data permissions: which data can be read, which data should not enter the model or process.
  3. Tool permissions: Can query, draft, modify, or can formally submit.
  4. Human nodes: Which steps must be confirmed by a person.
  5. Operation records: Retain necessary records of tasks, data, tools, and results.
  6. Stopping conditions: What situations require pausing or transferring to human intervention.

Three-layer permission starting approach

First layer: Read-only and organization. Agent reads data and produces a draft.

Second layer: Preparation for change. Agent creates content for confirmation, which is then approved by personnel.

Third layer: Limited formal actions. Requires clear scope, records, and withdrawal methods.

Companies can start from the first layer to accumulate evidence. Once quality, exceptions, and responsibilities are clearer, they can decide whether to open the next layer.

Make a first round of learning that can be judged within 90 days

DAY 01–30Select a process

Inventory high-frequency processes, available data, roles, inputs and outputs, and human checkpoints.

DAY 31–60Run a small-scale pilot

Let the Agent work with limited data and permissions, retain human review and record exceptions.

DAY 61–90Decide on the next step

Compare quality, time, cost, and human workload to decide to expand, adjust, or stop.

If only one thing is taken away Enterprise adoption of AI Agent starts with clearly defining a task. Once the goal, data, rules, permissions, exceptions, and evidence of completion are clear, tools can truly participate in the work.
AI AgentEnterprise transformationAI WorkflowHuman-machine collaboration

References and boundary of compilation

The following data are all from public speeches or articles. This article only provides summaries, translations, and practical enterprise organization:

The six actions, scene screening table, six-step pilot, guardrails, and 90-day roadmap in this article are my practical framework based on public views. They should not be considered as the original words of Jian Lifeng or an official methodology.

I am Coach Jiang

I am Jiang Yude, also known as Coach Jiang. I focus on how AI enters real work processes and help people organize their professional knowledge, judgment, and experience into AI-understandable and usable knowledge.

If you are also thinking about how to introduce AI Agent into your enterprise, you can start with a small process: write input, output, rules, permissions, exceptions, and evidence of completion on the same piece of paper. This paper is often closer to the real starting point than buying a set of tools.

Let's study AI workflows together

I will share new articles, public lectures, and practical summaries in the LINE community.

Join LINE Community