Based on a tutorial by Tiago Forte
Ever wondered if AI could replace your expensive personal coach? You’re not alone. With coaching sessions running $200+ per hour, many of us are priced out of getting the personalized guidance we desperately need.
That’s exactly what productivity expert Tiago Forte decided to test. In this fascinating experiment, he loaded 40+ personal documents (over 150,000 words) into Google’s Notebook LM to see if AI could provide meaningful coaching insights. The results will surprise you.
Quick Navigation
- The AI Coaching Experiment Setup (00:00-03:45)
- Preparing Your Personal Data Sources (03:46-08:20)
- Loading Sources into Notebook LM (08:21-11:15)
- Testing AI’s Understanding of You (11:16-16:30)
- Uncovering Personal Contradictions (16:31-22:45)
- Finding What Makes You Happiest (22:46-28:10)
- Future-Focused Coaching Questions (28:11-35:20)
- AI vs Human Coaching: The Verdict (35:21-End)
The AI Coaching Experiment Setup
Tiago starts with a bold premise: good coaching requires context. A coach needs to understand your goals, strengths, weaknesses, and how your mind works. But what if AI could access unlimited context about you?
This experiment tests whether loading extensive personal documentation into Notebook LM can create an effective AI coach. The scale is impressive – 40 separate documents containing nearly 150,000 words of personal context.
Key Points:
- Good coaching requires deep personal context
- AI now has essentially unlimited context windows
- The test involves 40+ documents and 150,000+ words
- Uses Notebook LM Plus ($20/month) for the experiment
My Take:
The context argument is brilliant. Most of us can’t afford ongoing coaching partly because of the time investment required to bring a coach up to speed on our lives. If AI can instantly absorb years of context, that’s a game-changer.
Preparing Your Personal Data Sources
The magic happens in the data preparation. Tiago organizes his sources into three strategic categories that you can replicate for your own AI coaching experiment.
Category 1: Private Notes
- Notes from actual coaching sessions with human coaches
- Mind maps with life goals updated annually
- Big Five personality test results
- Time perspective analysis from Zimbardo’s research
- Strengths compilation from customer testimonials
- Annual gratitude lists
- Myers-Briggs profile (INFJ in Tiago’s case)
Category 2: Public Writing
- Course takeaways and learning summaries
- Personal experience essays (like Burning Man)
- Book summaries and reviews
- Autobiographical writing
- List of favorite “open questions”
Category 3: Year-End Reviews
- Annual reflection documents from the past 5 years
- High points and low points analysis
- Lessons learned and key takeaways
- Goals and intention setting
My Take:
The beauty of this system is that most of us already have these materials scattered across different platforms. You don’t need to create new content – just gather what already exists. Even simple text messages or email drafts to friends can provide valuable context.
Loading Sources into Notebook LM
The technical process is straightforward but time-consuming. Tiago spends about 17 minutes manually uploading each document into a new Notebook LM project.
Key Points:
- Create a new project in Notebook LM Plus
- Upload sources one by one (no bulk upload option)
- Total process takes about 17 minutes for 40 documents
- System can handle 150,000+ words without issues
My Take:
While 17 minutes feels tedious, compare that to the months or years it would take to build this level of context with a human coach. The time investment is minimal for the potential payoff.
Testing AI’s Understanding of You
The first test involves basic comprehension – can the AI accurately identify strengths and weaknesses? The results are surprisingly thorough and accurate.
AI-Identified Strengths:
- Knowledgeable with deep understanding
- Visionary and futuristic thinking
- Creativity and innovation
- Insightful perspective
AI-Identified Weaknesses:
- Self-criticism and perfectionism (linked to INFJ personality type)
- Sensitivity to criticism and conflict
- Tendency toward idealism, overlooking details
- Social skills as perceived challenge (not reality-based)
- Emotional vulnerability and resistance to processing the past
- Potential for being overly private
- Taking on too much responsibility
- Fear of being ordinary
- Overthinking and analysis paralysis
My Take:
What’s remarkable here isn’t just the accuracy, but the specificity. The AI connects personality patterns (like INFJ traits) to specific behaviors and challenges. This level of synthesis would take a human coach months to develop.
Uncovering Personal Contradictions
This is where the AI coaching gets powerful. By analyzing patterns across years of writing, it identifies contradictions between stated intentions and actual behavior.
Key Contradictions Identified:
- Work-Life Balance: States intention for balance but consistently prioritizes work
- Delegation: Wants to delegate but struggles to let go of control
- Financial Habits: Desires financial prudence but tends to be a free spender
- Focus vs. Shiny Objects: Values focus but often chases new opportunities
- Family Presence: Wants to be present with family but gets distracted by work
The AI even provides psychological context, noting that free spending was partly a reaction against his father’s frugality – a level of insight that demonstrates sophisticated pattern recognition.
My Take:
This contradiction analysis is pure gold. It’s the kind of honest feedback that’s hard to get from friends (too close) or strangers (lack context). The AI has both the context AND the objectivity to call out these patterns without emotional baggage.
Finding What Makes You Happiest
Moving beyond problems, the AI identifies specific moments and situations that brought genuine happiness, providing a roadmap for more joy.
Happiness Patterns Discovered:
- Day 9 of a meditation retreat with waves of pleasure throughout the body
- Past experiences with psychedelics (ayahuasca) for therapeutic purposes
- Beautiful moments with children as babies
- Consecutive hours of deep reading and writing
- Deeply immersive spaces with fellow entrepreneurs and creators
The insights led to immediate actionable decisions: considering returning to regular meditation, exploring psilocybin or MDMA-assisted therapy, and seeking more immersive creative experiences.
My Take:
This is brilliant coaching technique – instead of focusing on what’s wrong, identify what’s worked before and do more of it. The AI’s ability to surface forgotten positive experiences from years of notes is incredibly valuable.
Future-Focused Coaching Questions
The most powerful coaching happens when addressing future goals and potential regrets. The AI delivers some hard-hitting insights about life priorities.
Potential Life Regrets Identified:
- Not fully prioritizing meaningful relationships and balanced life
- Overemphasizing work and external achievements
- Trading deeper connections for professional success
- Sacrificing personal well-being for business growth
When asked for overall life advice, the AI provided a comprehensive framework focusing on self-knowledge, emotional intelligence, integration over compartmentalization, and defining “enough.”
My Take:
The “deathbed regret” analysis is particularly powerful because it synthesizes years of stated values against actual behavior patterns. This kind of long-term perspective is exactly what good coaching provides.
AI vs Human Coaching: The Verdict
After extensive testing, Tiago reaches a nuanced conclusion about AI coaching capabilities and limitations.
What AI Coaching Does Well:
- Holds vast amounts of context (150,000+ words effortlessly)
- Instantly switches between big-picture and specific details
- Provides objective analysis without emotional baggage
- Identifies patterns across years of data
- Available 24/7 for coaching conversations
- Significantly cheaper than human coaching
What AI Coaching Lacks:
- Cannot read micro-expressions or body language
- Misses subtle emotional cues and reactions
- No intuitive human connection
- Cannot provide real-time emotional attunement
- Lacks the relational aspect of coaching
The Recommended Approach:
- Use AI and human coaching as complements, not replacements
- Leverage AI for pattern recognition and context analysis
- Use human coaches for emotional processing and relational insights
- AI coaching notes become valuable source material for human coaches
My Take:
This hybrid approach makes perfect sense. Use AI to do the heavy lifting of pattern analysis and context building, then bring those insights to a human coach for the emotional intelligence and intuitive guidance that only humans can provide. You get the best of both worlds at a fraction of the cost.
