If the future is that every product is headless, and agents can interact with every API, and UIs can be custom generated and disposable every time, then it's possible to have every piece of software we use be in one design system and one set of UIs.
That would eliminate a real cognitive tax today. When you work at company X, all the software you use looks and feels and works the same. How nice.
Is anyone working on this today?
We’re building the agent-native alternative to AWS, but it's actually good.
- 33,000+ projects created (10% WoW growth)
- 2x weekly active projects (2000 → 4000+)
- 11,000+ GitHub stars (5x in the last 3 months) - faster than React, Supabase, and Linux
Introducing InsForge ↓
The hardest problem in fintech isn't moving money. It's making the movement disappear.
Card networks won because the swipe hid the rail.
Venmo won because "it just sent" hid the ACH.
Stripe won because the checkout hid the MCC coding, fraud scoring, and cross-border FX happening underneath.
Whoever makes stablecoins disappear from the user experience wins the next era. Not whoever runs the fastest blockchain.
Prediction: within 12 months, 3 of the top 10 gig platforms move payouts to stablecoins.
Not because crypto won -- because the math always wins.
-- Weekly ACH: 2-3 day wait, opportunity cost for workers living week-to-week
-- Instant Pay: $1.99 flat fee, regressive on small earnings days
-- Cross-border corridors: 3-7% off the top
-- Tempo: instant, near-zero, 24/7
DoorDash went first. They won't be last.
the PSP model most teams are using today was built for a simpler version of payments.
it was designed for a world where one provider could handle a payment’s full lifecycle. in practice, the original foundation was built to help e-com businesses accept card payments.
fast forward to today, and online payments are completely different. bank payments have scaled, real-time rails continue to gain traction, and stablecoin adoption is accelerating.
each of these trends contributes to the growing complexity of fully understanding what’s happening under the hood from initiation to final settlement as payments cross rails, currencies, and systems.
it’s an infrastructure gap because the market has outgrown legacy PSP architecture.
today, teams need a PSP that can handle fiat (across cards and bank rails) and stablecoins with compliance, accounts, and full lifecycle ledgering built in, so teams can manage any payin or payout AND understand where each dollar is, end-to-end in every single payment.
if you want a closer look at where PSPs stand in 2026, we put together a guide that breaks down how the model has evolved and what teams need to evaluate when selecting their next PSP.
link in thread
As someone who just moved from Seattle, arguably one of the best tech hubs in the country, I can confidently say that San Francisco is on a completely different playing field than the rest of the country.
The people here are much farther ahead in terms of AI adoption and development that it feels silly to say that it’s “equally likely” to find talent here as it is in the Carolinas. Sure there is talent out there, but as far as a talent POOL goes, nothing compares.
This is silly — Silicon Valley and the Bay Area remain the best place in the world to find operators. We’re the deepest pool of talent for people who know how to build and scale multi-billion dollar companies.
And if you’re building in the AI space — this is where your peers are going to be. You’re going to push harder and develop better ideas because you’re regularly seeing them for coffees, at events and around town.
Everything else is downstream of that.
In February 2026, stablecoins surpassed ACH in monthly volume for the first time.
Stablecoins - $7.2T
ACH - $6.8T
Visa - $1.2T
Stablecoins are quietly becoming the foundational infrastructure for global payments: no banks, no weekends, no borders.
Data used: @artemis