Those who use AI to churn out perfunctory work hastily just to meet the submission requirements are ruining the workplace.

"The proposal is ready, everyone take a look, we will go through it tomorrow morning."

Right after that, a document of more than 20 pages was shared in the group.

With a complete title, clear table of contents, and sections covering industry trends, user analysis, competitor comparison and implementation plans, it looks like a fairly "professional" proposal at first glance.

But as the person in charge of landing the project, you find problems with it in less than ten minutes.

The market data cited in the proposal has no traceable source, the competitor is described as having launched a feature that does not actually exist, the so-called "user pain points" are just a few generic remarks that fit any industry, and the most critical parts including budget, timeline and personnel arrangement are not mentioned at all.

At the meeting the next day, the person who wrote the proposal says: "This is just the first draft, and all the specific details need to be supplemented by everyone."

It is at that moment that you realize the person did not finish a proposal, but used AI to create a new team task.

01

In the past, we usually judged whether a task was completed by checking if there was a deliverable. Now this standard has changed.

With AI, anyone can generate a full-structured report, a set of professionally worded proposals, or a seemingly information-rich PPT in ten minutes.

The problem is, "looking completed" does not equal "truly completed".

There is a term abroad called "Workslop", which can be understood as "AI workplace garbage": work outputs quickly generated by AI, seemingly complete on the surface, but lacking fact-checking, professional judgment and practical value.

The most troublesome part of it is not its poor quality, but that it transfers the work cost to others.

The person who generates it saves two hours, but the person who receives it has to spend four hours verifying information, figuring out the original intention and correcting mistakes.

The efficiency of one person is improved, but the efficiency of the whole team is reduced instead.

02

A company began to encourage its employees to use AI, hoping to reduce repetitive work. Xiao Lin, an operation staff, soon found his "efficiency boosting method".

Every Friday, he copies the background data to AI and asks AI to generate the weekly report automatically. The report is written very well:

"User activity has increased steadily this week, and the content strategy has achieved phased results. It is suggested that we continue to focus on core user demands and strengthen refined operation in the follow-up work."

For several consecutive weeks, the leader found no problems with the reports.

Until one time, the company planned to increase the promotion budget based on the weekly reports, and the supervisor casually checked the original data, only to find that the so-called "increased activity" was just caused by the change of statistical caliber.

AI did not know the caliber adjustment, nor did it know there was a temporary activity that week. It only generated a set of seemingly reasonable explanations based on the numbers in the spreadsheet.

The supervisor had to recheck all the reports of the past month, and ask for the data source of each item one by one.

Xiao Lin indeed only spent ten minutes writing the weekly report, but in order to judge whether the report is reliable, the supervisor spent a whole afternoon on it.

What AI saves is the time of the writer, but what it consumes is the trust of the reader.

During the mid-year performance communication, Amin, a product manager, received a very "formal" comment:

"It is suggested to further strengthen the global thinking, improve the efficiency of cross-departmental collaboration, and enhance the sense of goal and result orientation in complex projects."

She knows every single word in this comment, but has no idea how to improve herself according to it.

Which collaboration went wrong exactly? What does "insufficient global thinking" mean? To what extent in the second half of the year can be regarded as "enhanced result orientation"?

Later Amin found out that the leader just input several keywords into AI and asked it to polish them into a performance feedback paragraph.

The leader saved the time of organizing words, but Amin had to spend a whole week guessing what the leader really meant.

A truly effective feedback may not be gorgeously worded, but must contain facts:

What happened? What impact did it cause? What specific changes are expected from the recipient?

If there are no answers to these questions, no matter how professional the expression is, it is just transferring the responsibility of thinking to the recipient.

There is another team that uses AI in almost every link of their work.

Planners use AI to generate activity proposals; project managers use AI to summarize the proposals; during meetings, assistants use AI to sort out meeting minutes; before reporting, the person in charge asks AI to expand the minutes into PPT.

The whole process goes very smoothly, and the number of documents keeps increasing, until the client asks at the meeting: "Why do you think young users will like this feature?"

The meeting room suddenly fell silent.

The proposal says "it conforms to the trend of young users pursuing personalized expression", and the PPT says "it accurately responds to the demands of the new generation of users", but no one has interviewed any user, nor has anyone looked at relevant data.

Everyone has been in contact with the project, but no one has really studied the problem.

In the end, the team has dozens of pages of documents, but no one can take responsibility for the conclusions in them.

03

What AI truly amplifies is one's work habits. When using AI, some people will become more capable, while others will only become better at delivering perfunctory results.

The difference does not lie in whether the prompt is well written, but in whether he takes AI as an "assistant" or a "person who takes responsibility for himself".

People who work conscientiously will use AI to sort out materials, compare ideas and check expressions, then verify the facts in person, supplement first-hand information, and judge whether the proposal is feasible.

People who are used to being perfunctory will directly submit the first version generated by AI, and leave a sentence: "This is just the first draft, everyone can improve it together."

To judge whether a task is valuable, we should not only look at how fast it is completed, how beautiful the format is, and how many AI tools are used.

Instead, we should focus on three questions:

Does it provide new facts? Does it make clear judgments? Is there anyone willing to take responsibility for the result?

AI can certainly help us write weekly reports, sort out meeting minutes, make PPTs, analyze materials, and even put forward the preliminary framework of a proposal.

But it cannot replace us to understand the actual situation, judge the pros and cons, or take the consequences of decision-making.

This article is from the WeChat official account Liepin (ID: liepinwang), the author is Mochi Guobaorou, published by 36Kr with authorization.