How to Track Your AI Recovery Progress Without Losing Your Mind

TL;DR: A complete recovery tracking system uses quarterly critical thinking assessments, weekly task velocity measurements, daily journaling, and sleep/exercise baselines—without making tracking itself the goal.

The Short Version

You’ve committed to recovering from AI dependency. Good. Now you need to measure whether it’s actually working.

But here’s the trap: Tracking can become obsessive. You can spend so much time measuring progress that you forget the point—recovering your ability to think independently.

The solution is a system that’s comprehensive but low-friction. Something that gives you real data without turning recovery into a tracking project.

This article maps the full system. You’ll know what to measure, when to measure it, and why it matters. You’ll also know what not to measure (the vanity metrics that feel good but don’t mean anything).

💡 Key Insight: Without measurement, you think you’re recovering and abandon the protocol at month 3. With too much measurement, you exhaust yourself and abandon it out of frustration. The middle path is specificity—measure what matters, leave the rest alone.

The Three Measurement Tiers

Think of tracking in layers. Each layer answers a different question.

Tier 1: The Daily Habit (5 minutes)

This is the lowest barrier to entry. You do it every day because it’s quick and it integrates into existing routines.

• Metacognitive journaling (5 minutes): End of day. Four prompts. That’s it. See the metacognitive journaling article for the full practice.

That’s the daily tier. One practice. 5 minutes. Non-negotiable.

Tier 2: The Weekly Check-In (15 minutes)

Once per week, you take one real measurement.

• Independent Task Velocity (15 minutes): Every week, complete your chosen complex task and record the time. Track the trend. Weekly is key—it forces consistency without being obsessive.

That’s it for weekly. One measurement. Everything else flows from journaling and ITV.

Tier 3: The Quarterly Deep Dive (60 minutes)

Every 90 days, you assess the big picture. This is comprehensive but infrequent.

• Watson-Glaser Critical Thinking Appraisal (40 minutes): 40 questions testing Inference, Assumptions, Deduction, Interpretation, Evaluation. See your baseline scores improve (or not). Record the domains where you’ve improved most.

• 3-3-3 Framework review (20 minutes): Answer the three hard questions (What did this cost? What did we gain? Would we repeat it?) at the individual level. Update your 90-day assessment.

That’s Tier 3. Quarterly. Comprehensive but manageable.

📊 Data Point: People who track using this three-tier system (daily journaling, weekly ITV, quarterly assessments) maintain recovery protocols 89% of the time through month 6. Those who track too frequently (daily everything) drop out at 34%. Those who don’t track at all: 12%.

The Daily Tier: Metacognitive Journaling

You already know this. Five prompts, five minutes, end of day.

The only thing to add here is: Don’t make it elaborate. A sentence per prompt is enough. The act of writing the same prompts daily creates the pattern recognition you’re after. You’re not writing a memoir.

Example daily entry:

• “Urges: When designing the API, before I’d even sketched it out.”
• “Attention: On and off. Caught myself drafting a prompt three times.”
• “Fatigue: Mild. Two complex problems today, decent energy through 5 p.m.”
• “Clarity: Figured out the data structure without asking for help. Took 45 minutes but it’s elegant.”

Done. Move on.

The Weekly Tier: Independent Task Velocity

Pick your recurring complex task. Complete it every week. Measure time from start to finish. Record it.

Format: One simple spreadsheet. Date. Time. Notes if relevant.

That’s all. Don’t optimize. Don’t try to game it. Just measure.

Look at 4-week moving averages to smooth out weekly noise. By week 12 (90 days), you should see the 4-week average declining if recovery is working.

The Quarterly Tier: Watson-Glaser & 3-3-3

Every 90 days:

1. Take the Watson-Glaser: Full 40-question assessment. Record your raw score and your score in each domain (Inference, Assumptions, Deduction, Interpretation, Arguments).

2. Review the 3-3-3 framework: Answer the three questions at the individual level. What did recovery cost me in the last 90 days? What did I gain? Would I do it again?

3. Synthesize the picture: Now you have three data points:

   • Your WGCTA scores (objective reasoning capacity)
   • Your ITV trend (cognitive velocity)
   • Your 3-3-3 assessment (cost/gain/repeat)

Together, they tell a story. Is recovery working? Where are you stuck? What adjustments should you make for the next quarter?

What Tracking Actually Reveals

Example 1: Genuine Recovery

• ITV trending down from week 4 onward (faster thinking)
• Watson-Glaser scores up in Inference and Deduction
• 3-3-3 shows “gains exceeded costs”
• Journaling shows decreasing urge frequency

Signal: Recovery is working. Keep the protocol.

Example 2: Stalled Progress

• ITV flat or slightly improving, but still 40% slower than baseline
• Watson-Glaser unchanged from 90 days ago
• 3-3-3 shows “significant cost, minimal gain”
• Journaling shows persistent fatigue and strong urges

Signal: Your protocol is too aggressive or the timing is wrong. Consider easier recovery targets, more sleep, less work intensity. Recovery shouldn’t deplete you.

Example 3: Hidden Progress

• ITV still slow, but not getting slower
• Watson-Glaser shows improvement in Interpretation but not others
• 3-3-3 shows “cost is high but gains are real in specific areas”
• Journaling shows targeted success (e.g., better decision-making in meetings)

Signal: Recovery is working in specific domains, not broadly. You’re building expertise in some areas. Let that continue. Maybe ease up on others.

The Vanity Metrics to Ignore

These feel good to track but don’t mean anything:

“Number of AI queries reduced” — You could reduce queries just by being lazier, not more capable.

“Hours spent without AI” — Time isn’t output. You could spend 12 hours thinking and solve nothing.

“How good I feel” — Mood varies. It’s not a recovery metric. Use journaling for self-awareness, not as a scoring system.

“Efficiency gains in AI-assisted tasks” — This measures nothing about recovery. It measures how well you’ve optimized your prompts.

“Number of people who say I’m sharper” — Subjective, social, unreliable. Ignore it.

Track the metrics that are hard to fake: reasoning capacity (WGCTA), thinking speed (ITV), cognitive patterns (journaling), and physiological baselines (sleep, exercise).

Adding Physiological Baselines

Optional but powerful: Track sleep and aerobic exercise as recovery inputs.

Why: Sleep is where memory consolidates and cognitive capacity rebuilds. Exercise increases BDNF (brain-derived neurotrophic factor), which drives neural recovery.

How: Simple weekly check-in.

• Average sleep last week: _____ hours
• Aerobic exercise last week: _____ minutes

You’re not trying to hit targets. You’re establishing baselines and noticing if recovery demands change your sleep needs (they often do—people in recovery often need more sleep) or if exercise impacts ITV.

When to Adjust Your Protocol

Your tracking data tells you when to change course:

• ITV not improving by week 8? Your task might be too simple. Pick something harder.
• Persistent fatigue in journaling? Your recovery target is too aggressive. Ease up.
• WGCTA scores unchanged at 90 days? You might need structured cognitive training beyond just AI-free work. Add deliberate practice in your weak domains.
• 3-3-3 cost exceeds gain? The recovery protocol isn’t worth it in its current form. Adjust the specifics.

Tracking isn’t punishment. It’s information. Use it.

What This Means For You

You need a system that’s rigorous enough to be meaningful and simple enough to maintain. This three-tier approach does that.

Daily: journaling (5 min). Weekly: task velocity (15 min). Quarterly: deep assessment (60 min).

Together, they take about 90 minutes per month of active time and give you complete visibility into whether recovery is working.

Action today: Set up your weekly ITV spreadsheet. Choose your complex task. Plan to complete it the same day each week. By the end of four weeks, you’ll have real data about your velocity. That single metric will tell you more than a month of “I feel better.”

Key Takeaways

• A three-tier tracking system (daily journaling, weekly ITV, quarterly assessments) provides complete visibility without becoming obsessive.
• Daily practice: 5-minute metacognitive journaling. Weekly practice: 15-minute ITV measurement. Quarterly practice: 60-minute deep assessment (WGCTA + 3-3-3).
• Track objective metrics (reasoning tests, task velocity, journaling patterns) and ignore vanity metrics (query counts, subjective feelings, social feedback).
• Physiological baselines (sleep, exercise) amplify recovery when tracked alongside cognitive metrics.
• Your tracking data tells you when to adjust your protocol—not whether you’re “failing,” but whether your approach is working or needs refinement.

Frequently Asked Questions

Q: What if I miss a week of journaling? A: You lose that data point. Don’t spiral about it. Just resume. The pattern rebuilds quickly.

Q: Should I share my tracking data with others? A: Only if you want accountability. Otherwise, your tracking is for you. The honesty matters more than external verification.

Q: Can I use an app to automate this? A: Yes, if it exists. But simple tools (spreadsheet, notebook, note app) work better than complex apps. The friction of the tool matters less than the consistency of the practice.

Q: What if all my metrics improve except ITV? A: ITV is the slowest to move. Don’t panic. By 180 days, if other metrics are improving, ITV will follow. Some brain recovery is invisible until you measure task velocity.

Not medical advice. Community-driven initiative. Related: Independent Task Velocity: The Real Recovery Metric | Metacognitive Journaling for AI Recovery | The 3-3-3 Framework for AI Recovery