My journey with AI started in 2013. When my co-founders and I were building FiscalNote, we focused on natural language processing (NLP) and machine learning (ML) to predict whether legislation would pass or fail with high degrees of accuracy (94%+). Back then, AI wasn't as widely accepted as the next big thing the way it is today, and the 100+ VC firm rejections we received were probably a reflection of that. But to us, it was clear that these technologies had the potential to transform data collection, analysis, and decision-making at scale.

I took Andrew Ng's machine learning course around that time, though I never completed it due to the demands of building a startup. My exposure to AI remained tied to product applications. Over the years, I stayed engaged with AI's evolution, but my focus was primarily on FiscalNote's customers and expansion, not becoming technical.

The Generative AI Shift

In 2023, AI reached an inflection point. The rise of generative models wasn't just another incremental improvement—it was a fundamental shift. As I worked on FiscalNote's AI strategy, I focused on understanding:

  • How generative AI models work at a fundamental level.
  • What AI would get better at versus what humans would still be uniquely suited for.
  • The rapid acceleration of AI capabilities due to competition and funding.

I assumed that AI would improve exponentially while becoming cheaper, making it a skill worth mastering. That led me to 2024, when I finally decided to learn AI coding seriously.

From Strategy to Implementation

Rather than prioritizing 'vibe coding' (I did try it for a few days to see how far the technology could go as a non-technical person—short answer: not far enough), I went back to the fundamentals. I needed to deeply understand the mechanics of generative AI, how models are trained, and what best practices exist for integrating AI into applications.

I started with structured learning, enrolling in courses like Nat Eliason's Build Your Own Apps and Takeoff. These courses weren't just about writing code—they forced me to think about product design, system architecture, and how AI-powered applications can actually be useful. Learning in a structured way made all the difference.

Applying what I was learning, I built small AI-powered projects to reinforce concepts. I wasn't just coding for the sake of it—I was solving real problems, testing assumptions, and seeing firsthand how AI could enhance workflows. I leveraged AI tools themselves to accelerate debugging, explore new approaches, and iterate quickly.

Why This Matters

I fundamentally believe that this newest wave of AI represents a unique technological inflection point that will reshape industries, workflows, and human decision-making. The trajectory is headed toward rapid improvements, lower costs, and widespread adoption in the coming years. The original vision of the internet's capability to disrupt, even after going through the dot-com bubble, ultimately came true. AI is no different—whether or not it experiences a bubble burst, the vision will come true at some point.

For me, the best way to navigate this shift isn't to passively observe—it's to engage, experiment, and build. Staying at the cutting edge means understanding the technology at a deep level, and the only way to do that is by building.

This is the topic of blog post #1 because I'm literally coding this website with AI tools. I have other topics I'm curious about or would like to write about, and those can go in this blog too. The theme of posts will really be me writing whatever I feel like sharing with the world—likely to share perspectives, help others, and ask questions in public.

Updated: This post was edited to ensure the full title appears correctly throughout the site.