Blog Meta Analysis

I’ve been writing on this blog for nearly nine years now, and it occurred to me that I’ve never done what any good data person should do: analyze my own data. So I built some Python scripts to dig into 266 posts, notes, and garden entries spanning 2016 to 2025.

The Numbers

Let’s start with the basics. My blog contains:

• 266 total pieces across three content types
• Posts: 124 longer-form articles
• Notes: 124 shorter thoughts and observations
• Garden: 18 interconnected knowledge base entries (all from 2025)

The content spans from September 2016 to August 2025 - about 3,259 days of sporadic but persistent writing. It’s very fun that the number of posts and notes exactly equal (it will be broken after this post goes live).

Theme Evolution: From Economics to AI

The most striking finding was how dramatically my focus has shifted over time. I categorized everything into eight major themes using AI:

1. Artificial Intelligence & ML: 60+ pieces (23%)
2. Data Science & Analytics: 50+ pieces (19%)
3. Economics & Business: 45+ pieces (17%)
4. Technology & Tools: 40+ pieces (15%)
5. Sports Analytics: 25+ pieces (9%)
6. Personal Productivity: 35+ pieces (13%)
7. Entertainment & Media: 20+ pieces (7%)
8. Professional Development: 15+ pieces (6%)

But the real story is in the temporal evolution:

2016-2017: Economics Foundation My early writing was heavily economics-focused (63% in 2016). Posts like “Apple and Price Discrimination” and “How to Price Sports Tickets” established my analytical approach to business problems.

2018-2020: Technical Expansion
I diversified into data science tutorials, sports analytics, and early AI experiments. This period saw my first GPT-3 post in August 2020 - I was apparently an early adopter.

2021-2024: The Quiet Years Output dropped dramatically (3-12 posts per year vs 40+ in peak years). With navigating Covid and jobs afterwards I got busy, but what content I did produce was increasingly AI-focused.

2025: The AI Renaissance This year has seen an explosion of AI content - 23 out of 45 posts (51%) are AI/ML related. I also introduced the “Digital Garden” format for interconnected thinking.

The Productivity Paradox

Looking at posting frequency reveals some interesting patterns:

Peak Years: 2019 (48 posts), 2017 (46 posts), 2025 (45 posts) Most Productive Month: March 2025 with 17 posts Least Productive Month: October historically (10 posts total)

The 2021-2024 period is fascinating - my output dropped by about 75%, but when I did write, it was increasingly sophisticated AI analysis (if I do say so myself).

Early Adopter Patterns

One thing that’s been fun was how consistently I’ve been early to new technologies:

• GPT-3: First blog post within 2 months of getting access
• Various dev tools: Often among the first to review new productivity software
• Digital Gardens: They’ve actually been around for awhile, but I would say most people haven’t heard of them yet.

Content Sophistication Over Time

The analysis revealed clear sophistication progression:

• Early years: Basic economic explanations and tool tutorials
• Middle period: Applied analytics across domains (sports, business, entertainment)
• Recent years: Complex AI system analysis and strategic thinking about technology

Compare my 2017 post “AI Advancements” (a simple overview of AI beating humans at poker) to my 2025 garden entries on prompt engineering and agentic systems. The depth and nuance have clearly evolved.

The Digital Garden Revolution

2025 has been notable for more than just AI focus - I introduced a completely new content format. The 18 “garden” entries represent interconnected thinking on topics like:

• Agentic systems and AI workflows
• Knowledge management philosophy
• Bidirectional linking and content creation
• Personal productivity systems

This format allows for living documents that can evolve over time, which feels like a natural progression from the linear blog post format.

Methodology and Code

For the data nerds wondering about methodology, I used Claude Code with specialized subagents to:

1. Content Analysis: Parsed all 266 markdown files for themes, dates, and categories
2. Temporal Extraction: Pulled dates from frontmatter and filenames (many posts have ISO timestamps)
3. Theme Classification: Categorized content using both keyword analysis and full-text evaluation
4. Evolution Tracking: Mapped theme prominence over time

The Python scripts are in /scripts/ in my Github and include:

• extract_dates.py: Temporal data extraction with multiple date format handling
• thematic_evolution_analysis.py: Theme categorization and trend analysis
• ai_evolution_analysis.py: Specific focus on AI content progression
• blog_thematic_data.csv: The full dataset for further analysis

What This Reveals

This analysis reveals more than just posting patterns - it maps my intellectual journey from economics student to data analyst to tech enthusiast to AI practitioner. Each phase built upon previous knowledge while adapting to technological shifts.

The “quiet years” from 2021-2024 now make sense. I wasn’t writing less because I lost interest - I was absorbing and learning. The 2025 resurgence with sophisticated AI content shows that learning paid off.

Most importantly, the evolution from economics → data science → AI shows a consistent thread: using analytical thinking to understand and explain complex systems. The domains changed, but the approach remained consistent.

Looking Forward

If patterns hold, I should expect:

• Continued AI focus but with deeper technical sophistication
• More garden-style interconnected content
• Possible expansion into adjacent domains (maybe AI policy, AI business applications, or AI research methods)

The early adopter pattern suggests I’ll likely be writing about AI developments before they hit mainstream consciousness. And the analytical approach suggests I’ll continue applying data thinking to whatever domain captures my interest next.

After nine years of writing, the meta-analysis confirms what I suspected: this blog has been less about any single topic and more about applying analytical thinking to understand the world. The themes change, but the curiosity remains constant.