Company Video
Data Is Not Oil. It Is Alive.
A conversation with Aleks Jakulin about business directories, AI's data bottleneck, and why data infrastructure should be cultivated rather than mined.
- Featuring Aleks Jakulin
- Topic: business maps, AI, data infrastructure
- Source transcript: edited from automatic transcription
Why Listen
This is the more philosophical of the two company videos. Jakulin starts from early internet standards work, including PNG and JPEG-LS, and moves toward the question that defines data.Flowers: what would it mean to build an internet where data is maintained, updated, and economically supported instead of extracted until it decays?
The answer starts with a humble data type: business directories. Lists of companies look mundane, but they determine how customers find suppliers, how investors understand markets, how event attendees navigate trade shows, and how AI systems decide which organizations exist and deserve attention.
Data is not oil to be mined. Data is alive. It needs to keep getting updated.
From Lists to Maps
The conversation draws a sharp line between lists, search, and maps. A list tells you what someone already knows. Search helps when you know what you are looking for. A map shows what you did not know you did not know. That is why the first data.Flowers products focus on living ecosystem maps for associations, events, investors, and market researchers.
A map, in this framing, is not a static graphic. It is a living database presented visually. It changes as the underlying ecosystem changes, and it gives both people and AI systems a more reliable way to navigate reality.
The AI Infrastructure Argument
Jakulin argues that the next major AI advantage will not come only from bigger models, faster chips, or more electricity. It will come from better data infrastructure: reliable, stewarded, fresh, auditable data that AI tools can use without swallowing the web whole.
He also connects data infrastructure to AI safety. As AI agents begin acting on the world through interfaces built for humans, society needs ways to track, authorize, and rewind changes. That is not just a model problem. It is an infrastructure problem.
Edited Transcript Highlights
Damir First: How do we build data infrastructure properly?
Aleks Jakulin: We cannot do everything at once. It goes step by step, data type by data type. The most important data type for us to focus on initially is business directories. They are how customers and businesses find each other, but right now, in the absence of infrastructure, everyone is shouting and not communicating.
Damir First: Why is the company called data.Flowers?
Aleks Jakulin: Data is not oil to be mined. Data is alive. It needs to keep getting updated. If you just take it and mine it, you have killed the system that maintains it, invests in it, and benefits from it. We need to think less about extraction and more about ecosystems, nurturing, and sustainable growth.
Damir First: Why maps instead of lists?
Aleks Jakulin: Lists are hard to use. You scan down them, and there is usually not enough information for meaningful search. Search tells you what you knew you did not know. Mapping is the next level: it tells you what you did not know you did not know.
Damir First: What is a map in this context?
Aleks Jakulin: A map is a living database presented visually. Our visual cortex is like our GPU. Moving information from letters and numbers into visual maps lets us understand ecosystems more directly. A map is also infrastructure because it allows you to navigate.
Damir First: What should investors understand about AI right now?
Aleks Jakulin: The opportunity is not only making the next AI tool. It is making the greatest tool for AI tools to use. Better data infrastructure is one of the most important ways to amplify the value we get from AI.
Damir First: How should we think about AI regulation?
Aleks Jakulin: AI is like an engine or like fire. You do not regulate cars only by writing laws about driving. You regulate them by building roads, traffic lights, fences, and safety systems. We need infrastructure for building with AI, and we need infrastructure around how data is stored and changed.