In the AI Era, What You Shouldn't Overlook is the Ability to Distinguish Between 'Two Types of Judgments' and 'Two Types of Knowledge' (Organizational Behavior Science®)

Request Co., Ltd. (Headquarters: Shinjuku-ku, Tokyo, Representative Director: Tomoyasu Kouhata) has released the report "What You Shouldn't Overlook in the AI Era is the Ability to Distinguish Between 'Two Types of Judgments' and 'Two Types of Knowledge'".

Report Download

d68315-195-54795e6f62a3ce0e51dabfd174f2dc54.pdf With the spread of generative AI, corporate work is beginning to change significantly. Tasks such as searching for knowledge, organizing information, referencing existing cases, and processing according to predefined procedures will become increasingly easier for AI to handle in the future. On the other hand, what remains for humans are jobs that require deciding what to confirm, what to prioritize, how much to rely on precedents, and where to change the approach, depending on the conditions specific to each customer, project, site, and stakeholder.

However, many companies have not yet fully organized this change.

The background to this is treating judgment and knowledge as if they were each a single entity.

This report categorizes:

・ Judgments into 'judgments based on precedent' and 'judgments based on facts'.

・ Knowledge into 'knowledge that does not require experience' and 'knowledge that requires experience'.

It points out that much of the confusion occurring in companies in the AI era stems from conflating these four categories. Particularly problematic is trying to handle jobs that inherently require experienced knowledge and fact-based judgment using precedent application or model answers.

As a result, the following situations tend to occur in the field:

・ Increased understanding but not increased judgment

・ Proceeding according to precedent, yet not progressing as well as before

・ Increased rework and additional responses

・ Difficult cases concentrated among a few experienced individuals or managers

・ Work is getting done, but no new strategies are emerging

This report reframes these phenomena not as mere lack of ability, but as 'misallocation' relative to the original conditions for success. In other words, it suggests that the stagnation is due to jobs that should inherently be handled with fact-based judgment and experienced knowledge being treated as if they can be managed with precedent application or knowledge instruction alone.

Key Points of the Released Report

1. In the AI Era, What's Needed Isn't Just Saying 'Judgment is Important,' But Distinguishing Between the Four Categories.

The report indicates that what companies truly need to re-examine is:

・ Which jobs can be handled with judgments based on precedent.

・ Which jobs require judgments based on facts.

・ Which knowledge can be imparted through instruction.

・ Which knowledge requires experience to be useful.

Only with this distinction can we identify which tasks should be delegated to AI and which should be handled by humans, what should be taught in training versus what needs to be experienced in practice, and what organizational capabilities companies truly need to strengthen.

2. Organizing Work into Four Quadrants Reveals Areas to Delegate to AI and Areas to Retain for Humans.

This report categorizes work into four quadrants by combining the 'two types of judgments' and 'two types of knowledge':

・ Quadrant 1: Standard Processing Area

Includes procedures, rules, checklists, standardized explanations, and standardized judgments. This is the area most compatible with AI, automation, and standardization.

・ Quadrant 2: Confirmation and Adjustment Area This is an area where procedures exist, but confirmation of application conditions or exceptions is necessary. The focus is on fact-checking to prevent misapplication.

・ Quadrant 3: Area Prone to Misallocation This is an area where knowledge that inherently requires experience is handled through precedent application, model answers, or searching for correct answers. Understanding increases, but judgment is difficult to develop.

・ Quadrant 4: Core Area Remaining for Humans This is the area where conditions are differentiated, facts are confirmed, judgments are made, and standards are updated based on results. This is the core value that humans will provide in the AI era.

The report emphasizes that it is more important to correctly handle Quadrant 4 work as Quadrant 4 than to simply increase the amount of Quadrant 4 work.

3. What Causes Corporate Stagnation is the 'Quadrant 3-ification of Quadrant 4 Work.'

This report categorizes 'knowledge requiring experience x judgment based on precedent' as an unstable configuration under the original conditions for success.

In other words, Quadrant 3 is not a distinct job type but rather a state of misallocation where work that should inherently be handled in Quadrant 4 is processed through precedent application, model answers, or searching for correct answers.

When this misallocation occurs, frameworks tend to become the answer, cases become model answers, and principles become rules to be followed. While understanding increases, judgment does not, leading to increased rework, additional responses, concentrated confirmation, and stronger reliance on experienced individuals.

4. The Often-Overlooked but Significant Area in Practice is Quadrant 2.

Quadrant 2 is an area where, although procedures and standards exist, fact-checking is necessary to determine if they can be applied as is. Many jobs, including customer-facing roles, management, planning, site supervision, and back-office functions, include this area.

Even if systems and procedures are in place, if it's not clear what needs to be confirmed to avoid misapplication, problems arise where the system is followed but the field stops, or procedures are followed but rework increases. The report indicates that clarifying what needs to be confirmed and which conditional differences should not be overlooked in Quadrant 2 is crucial for overall quality and reproducibility.

5. What Companies Should Identify First is 'Where in Our Company Has Become Quadrant 3.'

The report outlines the first steps companies should take:

・ First, inventory the company's jobs across the four quadrants.

・ Second, identify Quadrant 3 areas.

・ Third, design jobs in Quadrant 4 to retain judgment.

・ Fourth, thoroughly delegate Quadrant 1 to AI and standardization.

The key is not to simply instruct people to 'think more,' but to design the work to clarify how much fact-checking is necessary to move forward and what assumptions must be made.

Why This Report is Being Released Now

Statements like 'judgment is important in the AI era' or 'experience is important' are correct. However, without further clarification, AI utilization, human resource development, and job design quickly become abstract discussions. What is truly needed is to recognize that there are two types of judgment and two types of knowledge, and to discern how these four are combined within jobs.

Request Co., Ltd., based on Organizational Behavior Science®, which analyzes data from 338,000 workers across 980 companies, has been analyzing what happens in jobs and why it persists from the perspectives of business environment, history, and experience. This report organizes the 'misallocation, not lack of ability' structure identified through this analysis as a point of discussion for job design in the AI era and releases it as a perspective for companies to review their own jobs.

For Companies with Such Concerns

・ Even after introducing AI, difficult cases remain concentrated among a few individuals.

・ Despite increasing training and knowledge sharing, the number of people who can make judgments in the field does not increase.

・ Although proceeding according to precedent, rework and additional responses are increasing compared to before.

・ Want to organize what should be delegated to AI and what should be retained by humans for each job type.

・ Need to structurally review the development of managers, customer-facing personnel, and the improvement of judgment skills in the field.

Judgment Design Laboratory's View

What will determine corporate competitiveness in the AI era is not increasing the number of people who know many correct answers. It is about developing people who can judge by discerning differences and designing jobs so that their judgment experience increases within their work. We believe that it is not the amount of AI utilization itself, but the depth of human resources capable of handling judgment-based tasks that remain for people, and the structure that increases their judgment experience within their work, that will determine the future response capabilities of companies.

Report Download

d68315-195-70073b78fc58b1a1fc300f44bec16377.pdf

Company Overview

Request Co., Ltd. Company Information: https://requestgroup.jp/corporateprofile Representative Director Tomoyasu Kouhata: https://requestgroup.jp/profile

E-mail: request@requestgroup.jp

Request Co., Ltd. (Headquarters: Shinjuku-ku, Tokyo, Representative Director: Tomoyasu Kouhata) operates under the slogan "Aiming for Better" and is a company supported by 980 companies through its 7 research institutions, based on Organizational Behavior Science®, which utilizes data from 338,000 working individuals. Organizational Behavior Science® clarifies 'why our thoughts and behaviors in organizations occur and persist' from the perspectives of the business environment and experience, and provides means to reproduce them better.