The Participation Problem
Most people are inside the digital economy but outside the value loop.
They use payment apps, subscribe to services, scroll feeds, make purchases. They move money. But they don't shape it. They participate as consumers, not as economic actors.
This wasn't always a problem. For decades, the path was clear: get educated, get employed, save consistently, let compound interest do its work over 30 years. That model assumed continuity: stable wages, predictable asset prices, a ladder that stayed in place long enough to climb.
That model is breaking.
Asset prices have decoupled from wages. Housing is financialized. Inflation outpaces savings. Careers decay faster than they mature. The old advice (be conservative, wait it out) isn't caution anymore. It's slow liquidation.
Meanwhile, AI is accelerating a restructuring that was already underway. The productivity gains are real, but they're accruing to a shrinking group: those with capital, infrastructure, and the technical fluency to direct these systems. Everyone else watches from the outside.
We're not heading toward a world with slightly more inequality. We're heading toward a bifurcation: a small class of AI-augmented capital allocators, and a much larger class locked out of the value creation entirely. The window to change this is narrow. Maybe a few years.
The Interface Gap
Here's what's strange: access isn't the bottleneck anymore.
Anyone can open a brokerage account. Anyone can download a crypto wallet. Prediction markets, tokenized assets, DeFi protocols: they're all technically available. The tools exist. The markets are open.
But participation hasn't followed. Why?
Because access without understanding is useless. And the interfaces we've built assume expertise that most people don't have.
The financial landscape has fragmented into dozens of disconnected systems: traditional brokerages, crypto exchanges (centralized and decentralized), prediction markets, DeFi protocols, tokenized securities. Each speaks a different language. Each carries different risks. Each requires different mental models.
A normal person trying to make sense of this faces a maze. Too many platforms, too many narratives, too much noise. The information is abundant, but what's missing is a coherent way to reason about it.
We gave people a jet engine and kept the cockpit from 1998. That's the real gap. Not access. Interface. Not information. Interpretation.
What Changes With AI
AI doesn't automatically fix this. Most "AI in finance" is a chatbot layered on a legacy terminal. A better search bar, not a new paradigm.
But the capability is there for something different.
Pattern recognition is now cheap. Simulation is cheap. Explanation is cheap. The things that used to require years of training or expensive professionals (technical analysis, risk modeling, scenario planning) can be done computationally, interactively, in real time.
What's missing is the integration layer. A system that:
- Perceives markets as dynamic structures, not static dashboards
- Reasons about risk and uncertainty over time
- Remembers past decisions and learns from outcomes
- Translates human intent into safe, executable action
- Makes the complex legible without dumbing it down
This is what we mean by General Financial Intelligence. Not a trading bot. Not a signal generator. An operating system for financial reasoning.
What We're Building
General Liquidity is an applied research and product lab building the software layer for this shift.
Our first product is Gordon, an AI-native agent that lives where real work happens. Right now, that's a command-line interface for traders who understand markets but are bottlenecked by tooling. People who have ideas but don't want to spend weeks writing Python to test them. Who want the power to turn intuition into a risk-sized plan.
Gordon connects to exchanges and data sources through a standardized protocol. It runs locally. Your keys never leave your machine. It reasons about markets, generates trade plans, manages risk, and executes with discipline. But it doesn't move without you. Every action requires human approval. Safety is structural, not optional.
This is phase one: proving the agent works in live markets. Think of it as Claude Code for trading, a conversational CLI that reasons about positions instead of code. Phase two is a generative interface: when you ask for analysis, you don't get text, you get a chart, a visualization, an interactive widget. Phase three is the full environment: an IDE for directing financial agents. Cursor for vibe trading.
The Broader Vision
Gordon is where we start. It's not where we end.
The destination is a world where more people have agency over capital, not just access to consumption. Where participating in markets doesn't require a finance degree or a quant background. Where the tools that used to be reserved for institutions are available to anyone willing to learn.
This doesn't mean turning everyone into day traders. Most people shouldn't trade actively. What it means is giving people the ability to:
- Understand how value moves through the economy
- Express beliefs about the future and test them against reality
- Take calculated risks with clear understanding of downside
- Learn from outcomes and improve over time
Economic intuition, not button-pushing. Participation, not speculation.
Principles
What We Believe
We believe financial market intelligence should feel continuous, context-aware, and disciplined. Not reactive, noisy, or opaque. The experience of working with an intelligent financial system should be one of clarity and control, not anxiety and confusion.
We believe the next generation of market participants will work alongside agents that genuinely remember, reason, and improve over time. Specialist systems that learn from their own mistakes. That respect the constraints their human operators define. That extend human capability rather than trying to replace human judgment entirely.
We believe this kind of system requires building carefully and deliberately. It demands deep respect for risk, obsessive attention to user control, and the understanding that markets are unforgiving environments that punish sloppiness and reward only systems that hold up under genuine pressure.
What We Refuse
A tool like this could easily become another vector for harm. Gamified speculation. Dopamine-driven engagement. False promises dressed up in AI language.
We refuse that.
General Liquidity will not optimize for engagement over outcomes. We will not profit from user mistakes. We will not hide risk behind complexity or speed.
Gordon is designed to slow you down when needed. To show you the downside before the upside. To say "don't act" when that's the right call. Discipline is encoded into the system, not left to willpower.
We measure success by whether users learn, whether they avoid catastrophic mistakes, and whether, over time, they become more capable economic actors. Not by how often they trade or how much revenue they generate. If the product can't meet that bar, we'd rather not build it.
Why Now
The restructuring is already underway. AI systems are being deployed across finance, mostly by institutions, mostly in ways that concentrate advantage further.
The interfaces that let ordinary people participate intelligently are not inevitable. They have to be built deliberately, by people who understand the stakes.
That's what General Liquidity exists to do.
We're not promising to fix structural inequality or turn everyone into an investor. We're building an on-ramp for people who feel the pressure, who know the old playbook is broken, and who are willing to engage seriously with a complex world.
The ladder is getting pulled up. We're trying to build another way to climb.