The first master plan I wrote four months ago is now well into execution. Build the financial layer for trading and spending. Grow an ecosystem of developers on top of it. Turn trading into culture. Bridge on-chain and off-chain finance. Give people a reason to keep coming back.
The reason we started with a trading app is simple. That's where the users were, that's where we could prove the product worked, and that's where we could generate enough revenue to fund what comes next.
Now we're ready for the next part.
I should explain something that has shifted materially since we wrote that first piece.
AI is getting dramatically better, faster than most people expected. Eight months ago our engineers were using it because it typed faster than humans. That sounds almost quaint now. Today, frontier models produce work that is genuinely close to what strong developers produce themselves, and the iteration loops are getting tighter every few months.
The implication of this is straightforward. If any team can build any software product quickly and cheaply, then software is not a defensible business anymore. The value has to come from somewhere else. It comes from the users you have, the liquidity you sit on top of, and the rails that actual commerce flows through.
That is what Based has been building. And it is why the next phase looks the way it does.
Here is what we believe will happen over the next 6 months, and what we are positioning for:
Everyone will have at least one agent running alongside them handling tasks, making decisions, executing transactions. Companies will run fleets of agents working concurrently. The number of "users" on any financial platform will multiply, but most of them won't have names.
Right now, most AI inference usage is concentrated around chat apps and, more recently, coding workflows. That will change. We are going to start seeing collaborative agents automate more parts of work, consumption, finance, and personal decision-making. The amount of AI-inference spending required to support that world will rise materially.
Programming was one of the first domains to get reshaped in a visible way. Spending is next. We believe AI agents will increasingly compute, compare, aggregate, and execute commerce decisions on behalf of users. Over time, AI will compress consumer decision-making itself; finding better prices, better routes, better timing, and more efficient actions.
The question worth asking is: where does all of that activity settle? Who owns the rails?
The goal is for Based to be the financial execution layer for both humans and agents. Not a wallet. Not a trading app. The system where economic intent, whether it comes from a person or an agent acting for that person, gets turned into action.
Here is how we get there.
If agents are going to operate at scale, they need reliable access to intelligence. That means inference has to become easy to consume, easy to pay for, and easy to embed into workflows (it needs to flow like liquidity).
Based has already built AI inference infrastructure (ai.based.one) that can be natively used by both agents and humans. More importantly, payments for this infrastructure have been designed to settle through Hyperliquid and Based Pay, our TradFi payment system.
We have already done this.
Infrastructure without usage is just overhead. We are releasing consumer products that sit on top of the inference layer and generate compounding demand. Product usage drives inference usage. Inference usage strengthens the infrastructure. That loop is what turns an infrastructure bet into a real business.
The long-term opportunity is not merely selling access to intelligence. It is owning the interface where intelligence meets action.
If AI becomes responsible for more decision-making, then the most valuable layer will not just be the model, or the app, or the wallet in isolation. It will be the system that can take an intent and translate it into execution across trading, payments, purchasing, and financial actions.
When we came up with that phrase, it described a product. Now it describes something larger.
Trading is not just perps and spot. It is any decision with an economic outcome that can be turned into an executable action. Spending is not just cards and checkout. It is any transfer of value, regardless of whether the entity initiating it is a human or an agent. Intelligence is not a chatbot bolted onto a wallet. It is the layer that sits inside the commerce stack and optimizes how capital moves through it.
When commerce becomes programmable, the companies that win will be the ones that already understand liquidity, settlement, incentives, and distribution as a single system. That is what we have been building.
We build around one of the most powerful ideas in crypto: that open financial rails enable faster experimentation and stronger alignment than closed systems do.
We are making a clear bet on two things.
First, that Hyperliquid becomes the dominant chain for AI commerce. The team is building toward housing all of finance on a single, fast, neutral chain. If they get there, it becomes the natural settlement layer for a world running on programmatic transactions. We have built most of our infrastructure there already. There is no comparable alternative.
Second, that agentic commerce is real and coming faster than most people think. OpenClaw gave an early glimpse of what it looks like when agents start replacing humans as the primary users of financial systems. That future is not theoretical for us. We have been building with agentic tooling for longer than it has been fashionable to talk about it.
We were early on Hyperliquid. We were early on agentic infrastructure. Being early is the only edge that consistently compounds.
Software is getting cheaper to build. Intelligence is getting cheaper to access. The window to build a defensible position around liquidity, users, and commerce rails is not permanent.
We have been in the right places early before. We intend to do it again.
The TGE is not just a financial event. It is the moment when the people who built this community with us get a real, structural stake in what comes next.
We do not just want to build products for our users. We want to build the future with them.
The future is Based.