On Habits, Tools, and the Complexity We Create

The world changes.

Tools change. The ways we work change. But the habits through which we understand the world do not necessarily change at the same pace.

A habit is not necessarily a flaw. Often, it exists because it worked.

When a new tool appears, some people see an opportunity to explore beyond familiar boundaries. Others adopt it because business demands it. Some are required to use it by their work. Others turn to it because they no longer understand how to solve a problem using what they already know.

The starting points differ.

So do the changes that follow.

One person may rebuild an entire workflow around a new tool. Another may incorporate it into an established routine. Someone else may decide that the tool offers too little practical value to justify the effort.

None of these responses, by itself, tells us whether a person is adapting well.

The more interesting question is what happens when the environment changes around them.

The Tool, and the Need

Imagine a simple cleaning task.

A person picks up a broom and sweeps the floor.

The task is completed with little equipment, energy, or preparation.

Now imagine building an advanced autonomous robotic system to perform the same task.

The system might require sophisticated hardware, computation, energy, maintenance, and infrastructure. If all of these are introduced merely to clean a small area, the additional capability may offer little practical value.

Of course, the calculation changes when the task involves dangerous environments, enormous areas, or repeated work at a scale that makes automation worthwhile.

The point is not that advanced tools are unnecessary.

It is that capability alone does not establish value.

A system may be able to do something that a person cannot. It may also be able to do something that a person can already accomplish more simply.

What matters is not only what the tool can do, but what its use changes in the real world.

And what it costs to make that change.

When Simplification Becomes Complication

Consider a different example.

A customer has a problem and needs help.

Originally, the process might be straightforward: explain the situation to someone who understands it, and work towards a solution.

Now introduce an automated customer-service system.

The customer may need to select categories, answer scripted questions, repeat information, navigate several stages, and eventually explain the same problem to a person.

Each step may function exactly as designed.

Yet the original problem remains unresolved.

Automation may have reduced the workload of a particular department, improved the handling of routine requests, or lowered operating costs. It may also have introduced friction elsewhere.

The system's internal efficiency and the customer's actual experience are not necessarily the same thing.

A process can become more efficient according to one measurement while becoming more complicated as a whole.

At some point, we may find ourselves following a process designed to solve a problem, while the process itself becomes another problem to solve.

The question is no longer simply whether the tool works.

It is whether the result still serves the purpose for which the tool was introduced.

What Changes When We Adopt a Tool?

The introduction of a new tool does not produce a single, universal response.

Some people use it to extend their existing abilities. Others change the way they work. Some discover possibilities they had never considered. Others find that the tool offers little that they actually need.

The same tool can produce very different outcomes because people begin with different questions, expectations, skills, resources, and constraints.

Artificial intelligence makes these differences particularly visible.

AI can expand access to information, generate unfamiliar approaches, and help people explore possibilities beyond their existing experience.

But access to information does not automatically produce the ability to evaluate it.

A person may receive an answer without knowing how to verify it. They may obtain a new method without knowing whether it fits the real conditions. They may use a powerful tool to complete a familiar task without ever reconsidering how the task itself is defined.

AI can reflect existing habits, reinforce them, or introduce something that challenges them.

What happens next depends on more than the tool alone.

The question is whether the new possibilities are understood, tested, and translated into useful action.

The Distance Between People

When a new tool becomes available, differences between people may become more visible.

Some may use it to multiply their output. Others may use it occasionally. Some may find that their work does not require it at all.

These differences do not necessarily indicate a permanent hierarchy of ability.

A person who does not need a particular tool today may have no reason to invest in it. A person who relies on it heavily may be solving a problem that others do not face.

But if the surrounding environment changes, the value of those choices may change too.

A method that once worked well may remain effective. Or it may gradually become less suitable as costs, expectations, competition, and working conditions shift.

The appearance of a new tool does not mean everyone must adopt it.

Nor does the fact that an old method still works today guarantee that it will remain sufficient under every future condition.

The distance between people may widen, narrow, or simply move to a different dimension.

We cannot determine the outcome from tool adoption alone.

When More Becomes the Default

There is another possibility worth considering.

When a tool gives us the ability to do more, we may gradually begin to expect more.

A simple task becomes an opportunity for optimisation. An established workflow becomes a candidate for automation. A decision that once required experience may become a demand for additional data, prediction, and analysis.

Sometimes this creates genuine value.

Sometimes the additional complexity costs more than it returns.

And sometimes the objective itself changes without us noticing.

We begin by asking how to solve a problem.

Then we ask how to automate the solution.

Eventually, we may find ourselves building a larger system to manage the complexity introduced by the earlier system.

This is not an inevitable consequence of technology. It is one possible outcome when the means of achieving a goal gradually become more important than the goal itself.

Perhaps the difficulty is not that we have too many tools.

Perhaps it is that the ability to do something can make it harder to question whether it needs to be done.

The world will continue to change.

Our habits will change too, although not always for the same reasons, or at the same speed.

Some changes begin with curiosity. Others begin with necessity. Some emerge from pressure, while others begin when an old way of doing things no longer provides a satisfactory answer.

A new tool can expand what is possible.

But possibility does not determine what we choose to do with it.

When we can delegate more, automate more, calculate more, and build increasingly complex systems, what should determine the direction of that progress?

The capability of the tool?

The expectations of the market?

The habits we have already formed?

Or the actual value of the result?

Perhaps the question is not whether we can keep up with the changing world.

Perhaps it is whether, as our capabilities expand, we can still recognise the difference between what has become possible and what is genuinely necessary.

Case 69 | 當更多，變得不再必要

習慣、工具，以及我們創造的複雜度

世界一直在改變。

工具在改變，工作的方式在改變。但我們理解世界的習慣，未必會以相同的速度改變。

習慣不一定是缺陷。很多時候，它之所以存在，是因為它曾經有效。

當一種新工具出現，有人把它視為探索未知的機會；有人因為商業需求而採用；有人因為工作要求而被迫學習；也有人因為原有的方法已經無法解決眼前的問題，才開始尋找新的答案。

起點不同。

後續的改變，也不會相同。

有人可能因此重建整套工作流程。有人只是把新工具融入原有的習慣。也有人衡量之後，發現它帶來的實際價值不足以抵銷所需的成本，於是選擇不使用。

單憑這些反應，我們無法判斷誰適應得更好。

更值得觀察的，是當周圍的環境改變時，這些選擇會帶來甚麼結果。

工具，以及需要

想像一項簡單的清潔工作。

一個人拿起掃帚，把地面清理乾淨。

整個過程只需要少量工具、體力與時間。

現在，假設我們建立一套先進的自主機器人系統，來完成同一件事。

這套系統可能需要複雜的硬件、算力、能源、維護，以及相應的基礎設施。如果只是為了清理一小塊地方而投入這些資源，新增的能力未必能帶來相應的實際價值。

當然，如果工作涉及危險環境、大面積場地，或需要反覆執行到足以抵銷自動化成本，情況便可能完全不同。

問題不在於先進工具是否多餘。

而在於，能力本身並不能證明價值。

一套系統可以做到人做不到的事情。

它也可以用來完成一些人本來已經能夠更簡單地完成的事情。

真正重要的，不只是工具能做甚麼，而是它的使用改變了現實中的甚麼。

以及，為了這個改變，我們付出了多少。

當簡化變成複雜化

再想像另一個情況。

一位客人遇到問題，需要尋求協助。

原本的處理方式可能很直接：把情況告訴一個理解問題的人，然後一起尋找解決方法。

現在，我們引入一套自動化客服系統。

客人可能需要先選擇問題類別、回答預設問題、重複輸入資料、經過多個處理環節，最後仍然要向真人重新解釋一次。

每個環節都可能按照設計正常運作。

但最初的問題，依然沒有解決。

自動化或許減少了某個部門的工作量，提高了重複請求的處理速度，或降低了部分營運成本。但它也可能在其他地方增加阻力。

系統內部的效率，與客人實際經歷的效率，並不一定相同。

一套流程可以在某項指標上變得更有效率，卻同時令整個問題的處理過程變得更複雜。

有時，我們原本建立流程，是為了處理問題；到最後，卻發現流程本身成為另一個需要處理的問題。

這時，問題已經不只是工具能否正常運作。

而是它的結果，是否仍然服務於當初引入它的目的。

當我們採用一種新工具

新工具的出現，不會令所有人產生相同的反應。

有人用它擴大原有的能力。有人因此改變工作方式。有人發現從未想過的可能性。也有人發現，這種工具並沒有提供多少自己真正需要的價值。

同一種工具之所以產生不同結果，是因為每個人的問題、期望、技能、資源與現實限制都不一樣。

人工智能令這些差異變得格外明顯。

AI 可以擴大我們接觸資訊的範圍，提供陌生的方法，協助我們探索超出既有經驗的可能性。

但接觸更多資訊，不代表我們自然具備評估資訊的能力。

一個人可以取得答案，卻不知道如何驗證；可以得到新的方法，卻不知道它是否適合現實條件；也可以用強大的工具，更快完成熟悉的任務，卻從未重新思考任務本身的定義。

AI 可以反映既有的習慣、強化它們，也可以提供挑戰它們的新觀點。

接下來會發生甚麼，不只取決於工具本身。

還取決於我們能否理解這些新可能性、檢驗它們，並將它們轉化成真正有用的行動。

人與人之間的距離

當一種新工具出現，人與人之間原有的差異，可能因此變得更加明顯。

有人利用它大幅提高產出。有人偶爾使用。也有人發現，自己的工作根本不需要它。

這些差異不一定代表一種永久的能力高低。

今天不需要某種工具的人，可能沒有理由投入時間去學習；高度依賴工具的人，也可能正在處理其他人根本不需要面對的問題。

但當周圍的環境改變，這些選擇的價值也可能隨之改變。

曾經有效的方法，可能繼續有效；也可能因為成本、期望、競爭與工作條件的變化，逐漸變得不再適用。

新工具的出現，不代表所有人都必須採用它。

但原有的方法今天仍然有效，也不代表它在未來任何環境下都必然足夠。

人與人之間的距離可能擴大、縮小，也可能只是轉移到另一個層面。

我們不能單憑一個人是否使用工具，就判斷最終的結果。

當更多，成為預設

還有另一種可能，值得我們留意。

當工具令我們有能力做更多事情，我們可能逐漸開始期待更多。

一項簡單的工作，變成優化的機會；一套既有流程，變成自動化的對象；一個原本可以憑經驗作出的決定，開始需要更多資料、預測與分析。

有時，這確實創造了價值。

有時，新增的複雜度卻比帶來的收益更高。

也有時，連最初的目標都在不知不覺間改變了。

我們最初問的是如何解決問題。

接著，我們問如何把解決方法自動化。

最後，我們可能發現，自己正在建立一套更大的系統，來管理前一套系統所帶來的複雜度。

這不是技術進步必然造成的結果。

但當達成目標的方法，逐漸變得比目標本身更重要時，這種情況便有可能發生。

也許問題不在於我們擁有太多工具。

而在於，當我們有能力做某件事時，反而更難停下來問：這件事是否真的需要做？

世界仍然會繼續改變。

我們的習慣也會改變，只是改變的原因不盡相同，速度也不一定一致。

有些改變始於好奇，有些始於需要；有些來自外部壓力，有些則始於原有的方法已經無法提供滿意的答案。

新工具可以擴大可能性。

但可能性本身，並不能決定我們會如何運用它。

當我們可以把更多工作交出去、把更多流程自動化、進行更多計算，並建立越來越複雜的系統時，究竟應該由甚麼決定進步的方向？

工具的能力？

市場的期望？

我們已經形成的習慣？

還是最終帶來的實際價值？

也許問題不在於我們能否跟上不斷變化的世界。

而在於，當我們的能力不斷擴大時，是否仍然能夠分辨：

甚麼已經成為可能。

以及，甚麼才真正有必要。