#03. Why the obsession with “time saving” and automation makes you lose your compass

“Save 10 hours a week with AI.”

It’s a seductive promise. We’ve all heard it, dozens of times.

And yet it’s not only misleading, it’s counterproductive.

Because in reality, AI isn’t going to save you any time (yet).

It asks you to invest. To think differently. To rethink your methods, your tools, your processes.

And so much the better.

Professionals who really take advantage of AI are not looking to go faster.

They’re looking to go further.

In this issue, I’d like to take a look at three preconceived ideas that may be holding you back in AI:

  • Myth #1: You’ll earn X number of hours in the first week (you always have to invest before you earn).
  • Myth #2: AI will automate your repetitive tasks (spoiler: they don’t exist)
  • Myth #3: Faster = better (the crucial difference between efficiency and effectiveness)
  • The tool to test: NotebookLM, the AI assistant specialized in understanding and synthesizing your own content.

Myth #1: “You’ll earn X hours in the first week”.

The hidden equation of time

September 2025. Parisian consulting firm. This AI tool saves 5 hours a week on market research.”

3 months later :

– Team training : 15 hours

Data cleaning: 18 hours 

Error correction: 15 hours

Systems integration: 10 hours

Process adjustments: 12 hours

TOTAL INVESTMENT: 70 hours

To “recover” these 70 hours at 5h/week… you’d need 14 “perfect” weeks.

No bugs. No learning. No adaptation.

The realistic formula

Time “saved” = Initial task time – (Learning time + Prompt time + Correction time + Monitoring time + Maintenance time)

The vendors of AI solutions and tools only show you the first term: “Initial task time”.

Yet very often, this equation is negative for the first 6 months.

And that’s normal.

The problem isn’t investment. It’s hiding it behind unrealistic promises.

The catch? We want to believe in the mirage because it responds to our frustration: the feeling of running out of time and the fear of wasting more of it.

Myth #2: “AI will automate your repetitive tasks”.

Or why repetitive tasks are like unicorns

Since the arrival of ChatGPT, an obsession has taken hold: identifying repetitive tasks and entrusting them to AI.

Search for repetitive tasks

Google Trends data, a free tool that shows the evolution of Internet users’ interest in a word or expression in the Google search engine.

This graph illustrates that the automation of repetitive tasks was a niche topic before 2022, but with the arrival of ChatGPT and the generative AI wave, interest has exploded on a global scale.

In reality, repetitive tasks are popular on the Internet, but hard to find in everyday life.

The sudden popularity of automation advocated by AI vendors disguises a problem: most of these tasks are not repetitive at all.

Let’s take a seemingly trivial example: writing a reminder email.

Repetitive? On the face of it, yes.

Now answer:

  • Same tone for a difficult new customer or a loyal customer?
  • Same approach for €500 and €50,000?
  • Same style for a start-up and an institution?

If you answered “no” only once, your task is not repetitive. It is contextual.

In practice, each relaunch depends on multiple variables:

  • The tone: formal, friendly, insistent?
  • The context: first message or fifth attempt?
  • History: what exchanges have already taken place?
  • The issue: administrative slowness or risk of insolvency?
  • The personality of the recipient: how does he or she communicate?
  • Timing: emergency or simple follow-up?

Yes, AI can write the email.

But who’s going to give her the right instructions? Who’s going to make sure she sticks to the tone? Who will adjust it?  You.

The grain of sand in automation: recurring ≠ repetitive

  • Recurring: A recurring task comes up often. You often repeat the same action (writing emails, analyzing contracts, preparing presentations).
  • Repetitive: A repetitive task is the same every time. You repeat exactly the same action, in the same context, with the same parameters.

What’s the difference? Context.

And context is 80% of intellectual work.

What we call repetitive actually varies according to context.

And yet, context is precisely what AI struggles to grasp on its own, and why humans remain the determining factor.

Myth #3: “Faster = better

Reminder: “Speed doesn’t matter if you’re going in the wrong direction.” (Gandhi)

AI is a formidable gas pedal. It enables us to produce faster, generate more ideas and respond more quickly.

But just because AI makes it possible to go fast doesn’t mean we should lose sight of the right objective.

Effectiveness vs. efficiency the nuance that changes everything

In French, both are often translated as “efficiency”, but this is a mistake.

Here’s the fundamental difference:

Efficiency (efficiency) = doing things well

  • Definition: Initially, the word “efficient” was mainly used in philosophy to designate an “efficient cause”, i.e. one that produces an effect. Since the 1950s, under the influence of the English word ” efficient”, its use has spread into everyday language, mainly in management, to designate that which produces an effect with a minimum of resources, without wasting time or effort (in other words, that which is “effective” but optimized).
  • Question: “How can I do this job faster/cheaper?”
  • Focus: Method, speed, optimization of resources
  • Measurement: Time saved, costs reduced, processes accelerated

Effectivenesseffectiveness) = Doing the right things

  • Definition: “One who produces the expected effect; one who succeeds in reaching the set goal or fulfilling his task”. Here, we’re talking about achieving the set objective, no matter how many resources (time, money, effort) are involved. We are efficient when we do the right things and achieve the expected result;
  • Question: “Am I doing what it takes to reach my goal?”
  • Focus: Results, impact, relevance of action
  • Measurement: Objectives achieved, results obtained, value created

Why this nuance is crucial in the field of AI

Efficiency with AI:
“ChatGPT helps me write my LinkedIn posts five times faster.”
→ Result: 20 mediocre posts that reach no one.

Effectiveness with AI:
“AI helps me understand why my posts aren’t engaging.”
→ Result: 3 targeted posts that generate conversations and leads.

The efficiency trap: AI makes us very efficient: we produce faster, we automate, we optimize. But we can become super-efficient… doing the wrong things.

The power of efficiency: AI can also help us to identify what really works, to focus on high-impact actions.

To be effective is to reach the goal, to be efficient is to reach the goal with the least effort and resources.

As management guru Peter Drucker said, “There’s nothing more pointless than doing efficiently what shouldn’t be done at all.”

3 principles to follow

  1. Look for effectiveness before efficiency

Ask yourself, “Is this task worth optimizing?”

Tip: Use AI to identify what’s going wrong. Then look for ways to improve.

  1. Invest in understanding before optimization

Learn where AI excels (and where it fails).

Develop your intuition before looking for speed

  1. Measure the value created, not the time “saved

Focus on the impact of your actions

Prefer 1 well-used hour to 3 “optimized” hours

Concrete examplesèts

The wrong approach

The right approach

“How can I write my reports faster?”

“How to identify key information?”

“How can I automate my emails?”

“How can I better understand my customers?”

“How to produce more content?”

“How do you create content that resonates?”

It’s better to be slow to do the right thing than quick to do anything at all.

So let’s use AI less for efficiency (doing things faster) and use it more for effectiveness (doing things that really matter).

NotebookLM: The AI assistant that all consulting professionals and knowledge workers urgently need to try out

NotebookLM is an AI assistant from Google that summarizes, analyzes and organizes your documents, relying solely on your sources to give more reliable answers.

Unlike AIs like ChatGPT, NotebookLM doesn’t rely on all the data found on the Internet (or in the AI’s training data), but only on the documents, notes, videos, PDFs, web links, audios, images or Google Docs you provide.

NotebookLM requires an initial investment (organizing your sources, structuring your documents) to offer you real added value: in-depth understanding of your content.

The main advantages of NotebookLM :

  • Multiply your sources and amplify your knowledge: NotebookLM accepts an impressive variety of formats – PDFs, Google Docs, websites, YouTube videos, audio files (MP3/WAV), and Google Slides files. With up to 300 sources per notebook and a context of 25 million words, the tool can process huge volumes of information.
  • Precise quotations : Each answer includes direct quotations with links to the exact passages in your source documents, greatly reducing the risk of AI hallucination.
  • Confidentiality: Uploaded documents remain private, are not used for model training, and data management ensures traceability and security.
  • Synthesis and writing: Generate Q&A sheets, outlines, structured summaries, audio podcasts from your documents, etc. All directly linked to your document database and exportable.
  • Audio and Video Overviews : NotebookLM is attracting a lot of attention for its ability to transform your documents into podcast-style audio discussions between two AI hosts of 5 to 30 minutes offering an engaging synthesis of your content (audio overviews), or into a presentation combining audio narration and visual presentation (video overviews).
  • Collaboration made easy: NotebookLM lets you work with several people, sharing project folders with quotes, annotations and version tracking, ideal for professional or advanced academic use.

Resource: The World Ahead 2025 (The Economist & NotebookLM)

Every year, The Economist publishes The World Ahead, its forward-looking report on the year ahead. The 2025 edition tackles 10 major trends: the return of Trump and his geopolitical repercussions, protectionist escalation, the cleantech boom, global aging… and the moment of truth for AI with over $1,000 billion invested in data centers.

The Economist - The World Ahead - NotebookLM

New: this edition is available as a Featured Notebook in NotebookLM (a first IA collaboration for The Economist).

With this format, you can :

  • Read the original articles,
  • Ask your questions in chat and get informed answers,
  • Explore major themes using Mind Maps,
  • Or listen to Audio Overviews to quickly grasp the key points.

Explore the resource : Featured Notebook The Economist – The World Ahead in 2025

Other Featured Notebooks to discover: