At a glance
WHAT IT’S REALLY ABOUT
Meridian aims to make Excel work 20x faster using AI
- Meridian is building “AI for spreadsheets,” treating Excel as the world’s most distributed programming language and targeting massive productivity gains for spreadsheet-heavy roles.
- John Ling’s operating style—curiosity-driven learning, first-principles thinking, and taking on ambiguous problems—was sharpened at ScaleAI through research-heavy work on data quality, benchmarks, and internal LLM efficiency.
- The key product bet is that unlike coding (where toolbuilders are also power users), spreadsheet/finance workflows lack sustained experimentation with AI, leaving large unrealized gains from decomposing tasks and automating sub-workflows.
- Ling argues that personal advantage in the AI era comes from sustained hands-on time with models (e.g., “10,000 hours”), building intuition about what’s possible now and soon, and improving “prompting” as a thinking tool.
- The cultural prescription is to build teams with psychological safety to experiment, fail, and iterate—because discovery work is required to unlock AI-driven workflow transformation.
IDEAS WORTH REMEMBERING
5 ideasTreat spreadsheets like a programming platform, not just a UI.
Meridian’s framing—Excel as the most distributed programming language—implies the biggest wins come from adding an AI layer that can generate, transform, and validate spreadsheet “programs” at scale.
The spreadsheet AI opportunity is underexplored compared with coding assistants.
Coding tools improved fast because builders are daily users who can precisely define failures; in finance/spreadsheet work, fewer people have put in the experimentation time to map what AI can reliably do.
Decompose complex spreadsheet workflows into AI-solvable modules.
Instead of expecting a model to produce a perfect end-to-end LBO, break the workflow into parts (assumptions, data ingestion, formulas, checks, formatting, scenario runs) where models can perform strongly and be verified.
Build an “unfair advantage” by logging sustained hours with AI.
Ling’s recommendation—spend thousands of hours doing your domain with AI—builds intuition about current capabilities and near-term progress, helping you choose feasible products and workflows ahead of others.
Prompting is both an execution skill and a thinking tool.
Explaining tasks precisely to an LLM forces you to clarify objectives, constraints, and definitions, similar to how writing a YC application clarifies a business.
WORDS WORTH SAVING
5 quotesWe think about, like, Microsoft Excel as the most distributed programming language in the world, and our goal really is to say, "Hey, how can we help all of the people that spend a lot of time in spreadsheet software today just move 20 times faster?"
— John Ling
I don't believe any person on the planet spent 1,000 hours trying to build financial models with AI.
— John Ling
Everyone on this team must vibe code, and if you don't know how to vibe code, I feel like you're just gonna be lost or you're gonna be left behind.
— John Ling
It's really easy to fall into this narrative that like, oh, these things are like impossible, but you actually don't know.
— John Ling
I think it will be advantageous to be one of the people that have spent... Let's say you're interested in finance, right? That have spent, you know, like 10,000 hours trying to do finance with AI.
— John Ling
High quality AI-generated summary created from speaker-labeled transcript.
