21 Days Away from AI: What I Missed—and Why It Matters
Reflections on AI, climate tech, and the tools shaping modern investment research
What I’ve Been Learning: AI, Sustainability, and Investment Research
After stepping away for a few weeks to recharge, I returned to an AI landscape that had already shifted. In just 21 days, OpenAI and Google released new models, Agent-to-Agent systems were showcased at Google Cloud Next, and generative AI-created Ghibli-style images filled my feed. Thoughtful discussions around GenAI evaluation frameworks are also gaining traction.
It was both exciting and overwhelming.
This experience reminded me that it’s impossible to follow every new development in AI. I’ve realized the most effective approach is to focus on specific use cases and filter out the noise. The goal isn't to chase the newest tool—it’s to identify what actually works for your team, your product, and your problem set. That might mean balancing cost and performance, accuracy and explainability, or simply making integration more seamless.
Takeaways from SF Climate Week
During SF Climate Week, I attended several events that offered valuable insight into how AI intersects with climate innovation and sustainability. Two in particular stood out:
1. AI Innovations for Business Sustainability (hosted by Unravel Carbon)
Grace Sai, Co-founder and CEO, demoed three AI-powered agents that help companies accelerate their sustainability efforts:
A data transformation agent
A disclosure and reporting gap analyzer
A climate risk identification agent
The demos were concrete, well thought out, and already integrated into real-world business workflows. It was encouraging to see the audience's engagement—and clear that companies are hungry for tools that make sustainability work more actionable.
2. From Insight to Action: Breakthrough Climate Research (hosted by Swissnex)
This session introduced me to the concept of Open Science. As someone who’s long embraced open-source tools and public data, I found this framing refreshing. Open Science creates broader access to academic research and unlocks new possibilities for independent researchers and technologists.
The event also featured a keynote by Dr. Zia Mehrabi, who presented research on how increasing biodiversity in farmland can lead to more resilient food systems. The combination of ecological insight and data-driven modeling made for a compelling discussion.
AI Tools I’m Exploring for Investment Research
I also came across a couple of new tools I’m excited to try:
DeepWiki (by Cognition AI): This tool generates architecture diagrams, documentation, and source links for public GitHub repositories. It can be a fast and effective way to ramp up on unfamiliar codebases.
Paper2Code: An emerging tool that converts scientific research papers into runnable code. For someone like me—who lives in the space between research and applied data science—tools like this can shorten the distance between insight and implementation.
Final Thoughts
Every time I unplug, I come back reminded of how fast this field is moving. AI’s role in finance and sustainability is evolving rapidly, and it’s clear that many of the most meaningful innovations are coming from the intersection of these domains.