Benedict Evans’ recent newsletter on LLMs and data analysis got me thinking about the vast potential of public data—and its limitations. https://www.ben-evans.com/benedictevans/2025/2/17/the-deep-research-problem

On one hand, we have an incredible alliance of public data sources, fueling AI-driven insights. But on the other, we lack truly valuable data. 

Take water usage or snowpack levels in local mountains—both essential for climate planning, yet often fragmented or inaccessible.

Meanwhile, federal administrative data is set to shrink in the coming years, raising concerns about how we sustain high-quality analysis without it.

As AI continues to evolve, the question isn’t just what models can do, but what data they can access.

Would love to hear others’ thoughts: Where do you see the biggest gaps in public data today?