Stablecoins are having a real moment, and it makes sense. This is infrastructure maturing, not some magic overnight reset.
Roy at Breez put it plainly, “They’ve grown from practically nothing at the turn of the decade to a market cap in the mid-12 digits and monthly transaction volumes in excess of $1 trillion. Citigroup expects the aggregate stablecoin market cap to hit around $2 trillion by the end of the decade.”
That’s the macro story.
And if you’re building in AI right now, you’ve felt the other side of that story too.
The capital moats have already formed around the most crowded AI sectors.
As a founder, trying to break into a brigade of “good enough” clones can feel like running through the walls of Troy in the dead of night.
The gates are unguarded, the courtyard is packed, and every path looks like it’s already been staked a claim.
The concepts are new, but the expectations are brutal.
You’re asked to ship, differentiate, and find product fast inside a market that moves quicker than it can monetize.
And as capital moves downstream from pure infrastructure into 2026, the pressure shifts from “can you build it” to “can you turn it into real revenue with real customers” on terrain that still feels artificial and unsettled.
That can sound like bad news.
But at PlebLab we believe that it’s usually the opposite.
Because the main gate is where everyone is fighting, and the edges are where builders actually win.
Real innovation doesn’t announce itself from the center of the city.
It shows up in the blind spots, always shows itself around the edges, the places that are unclaimed.
Because here’s the shift a lot of people still haven’t internalized.
Your customer might not be human
A huge portion of what’s getting built on top of AI is not a classic app. It’s a chain of actions.
An agent searches.
An agent calls a tool.
An agent hits an API.
A model calls another model.
That means the thing consuming your product is often another system. Not a human with a credit card and patience for a billing dashboard.
And we’ve spent decades designing monetization for humans.
Logins, pricing pages, subscriptions, usage dashboards, credits, seat counts.
Machines don’t want any of that.
They want a clean loop:
Make the call. Pay. Get the result.
Why Lightning keeps showing up
This is where we believe at PlebLab, the path gets clearer.
Simply Lightning.👌
Not because it’s trendy in Austin, Texas.
But, because it collapses complexity.
When you’re building in a world of agents and tool calls, you need payments that behave like the internet.
Fast. Programmatic. Reliable. Minimal ceremony.
Lightning gives you that.
It works with user experience on top when you need it, but it also works without the fluff.
It’s simple enough for machine to machine commerce, and that simplicity is the feature.
Speed isn’t a nice to have, it’s the product
If an agent is waiting on a response, slow settlement is not a minor inconvenience. It breaks the workflow.
When computation is happening in seconds, payments can’t take minutes. They can’t take “later.” They can’t require a human to intervene.
So the architecture pressure becomes obvious.
Either you introduce a middleman who keeps score, or you use a payment layer that settles fast enough to live inside an automated loop.
Lightning fits the loop that leads to revenue.
Stablecoins will be there, and Lightning can still be the clarity layer
Stablecoins are moving dollars around the world. They’re easy to reason about, and they’ve become a default for a lot of teams.
But for machine native commerce, the denomination needs to be frictionless.
Machines don’t care about the narrative. They care about completing a job.
They need to be able to pay for a tool call, instantly, and continue the workflow. Anything else is overhead.
That’s why this isn’t an either/or.
Stablecoins can keep growing as a dollar transport layer.
Lightning sats can be the clean execution layer that turns “work” into “paid work” without slowing the machine down.
What it looks like when builders actually ship it
Take Runstr.
Through Startup School, the team integrated Lightning into a powered fitness app, while also leveraging Claude Code. That combo is the shape of what we’re seeing more and more.
AI expands what the product can do.
Lightning tightens the incentive loop.
The builder and community of users get paid without turning their product into a fintech company.
That’s clarity but its also just the beginning of how to increase revenue over the long term.
The path forward
LA-based startup Quilter has outlined Project Speedrun, which marks a milestone in computer design by AI. The headlining claims are that Quilter’s AI facilitated the design of a new Linux SBC, using 843 parts and dual-PCBs, taking just one week to finish, then successfully booting Debian the first time it was powered up. The Quilter team expressed that the AI-enhanced process it demonstrated could unlock a new generation of computer hardware makers.
Honestly looking back PlebLab and Oshi were early in 2021 with Lightning, in 2025, Lightning started getting treated like a real payments option for existing tech companies.
By the end of 2026, we think AI and Bitcoin stop being separate categories. They just become part of the same modern stack.
Most serious Bitcoin builders I talk to are already living in a multi-model world. The game isn’t “who has the biggest model anymore.”
It’s who ships the most focused product with the strongest distribution and the cleanest loop from value to revenue.
Lightning is that loop for machines.
It’s the AI layer that brings clarity to complexity.
Interested in the next round of PlebLab Startup School? Jump on the newsletter so you’re first to know when applications open.
If you need help now, grab an Online Builder Pass and get your project in front of people who have been in the trenches. PlebLab’s mentors and network can help you work through roadblocks, sharpen the idea, and move faster toward real sats flow and real customers.
If you’ve got something specific you want to jam on, you can also schedule a one off call for a focused project or startup consultation.
