You walk into a store, tap your card for coffee — but the money comes straight from your crypto wallet. No bank. No approval. No waiting.
This isn’t sci-fi. It’s already happening in 2026.
But most crypto cards still suck with fees, UX, and limited tokens.
Here’s what actually works — and what doesn’t. 👇
All of this is Leone Parise, co-founder and CTO of @tracefinance , $10B+ moved across borders, clients like Amazon and HBO.
Full conversation is live now:
🎥 youtu.be/bAxArI38sYY
And his most contrarian take:
The AI layoff narrative is theater.
Say "we cut headcount because we're not innovating" → your stock drops.
Say "we cut headcount because AI" → your stock goes up.
Same layoff. He's watching those same CEOs quietly reopen engineering roles.
Meanwhile the thing that keeps him up at night isn't competition. It's this:
Attackers only need to break your system in one way you didn't predict.
Now they can do that 24/7, with a million agents running in parallel, testing inputs you never imagined.
So where's the actual money?
Treasury management.
Every fintech you admire makes money the same way: receive from one party, pay out to another, and always have a balance parked somewhere earning yield.
The fees are the story you tell. The float is the business.
Now the harder truth.
He thinks most blockchains launched in the last few years add zero value. They compete on TPS and fees, then get funded by VCs as if infrastructure were a growth business.
"Blockchain is a horrible business model." His words. He's built on it for a decade.
Revenue? Fees for listing tokens.
Endless pull requests from shitcoins paying to get listed. Thousands of dollars a month.
Plus revenue share from swap partners.
No ads. No promoted tokens. A wallet that pushes coins to its users isn't a wallet, it's a casino.
They moved all the heavy lifting into the app itself.
Your phone connects to every blockchain, pulls balances and transactions, stores it in a local database, syncs in the background.
The backend only sent push notifications.
That's how you scale to 15M users with no budget.
First lesson: wallets are a terrible business.
Trust Wallet's parent company paid salaries. Not marketing. Not servers.
The backend ran on Heroku with a single Postgres instance and two engineers.
So they did something clever.
A guy joined a crypto wallet when it had 3 engineers and 50,000 users.
He left 2.5 years later. The wallet had 15 million users.
Here's what he learned about why almost none of it was a real business đź§µ
Dr. Debra Clary breaks all of this down — plus what changes when AI makes every answer instantly available — on episode 92 of Colors of Tech & Entrepreneurship.
Out now on YouTube and Spotify 👇
youtube.com/watch?v=bgiq7AGx…
His team's response: "we come to you with ideas and you shut us down."
Revenue had slowed. Margin was slipping. Competitors were shipping faster.
The problem was sitting at the head of the table.
The uncomfortable part is that this is measurable.
One founder took a curiosity assessment with his 13 direct reports. Scores were shown anonymously. Everyone scored high on openness to new ideas — except one person.
The CEO raised his hand and said: that's me.
The kicker: these are frontline employees.
They have the most information about what's actually happening — and they're consulted the least.
Gallup has been measuring employee engagement for decades. It's never been lower.
When researchers asked disengaged employees why, the answer wasn't about pay or perks:
"My leader doesn't know what I can contribute. They just tell me what to do."
Blockbuster ran the same playbook.
Netflix and Redbox were emerging. Leadership said nobody would ever go for that.
One of those companies still exists.
Here's the trap:
Experience builds confidence.
Confidence builds certainty.
Certainty narrows curiosity.
And the moment curiosity narrows, engagement, satisfaction and performance start sliding — usually before anyone notices.
A researcher who spent 30 years inside Frito-Lay, Coca-Cola, Jack Daniel's and Humana has a name for what happened.
Certainty.
BlackBerry saw the iPhone coming.
The CEO looked at it and told his team not to pay attention.
That single decision cost the company everything — and it wasn't a technology failure.
Colors of Tech and Entrepreneurship retweeted
The recent @stripe's acquisition of @OpenRouter is more interesting than “payments company buys AI gateway.”
AI apps will increasingly choose models the way payments businesses choose payment routes: not by loyalty to one provider, but by a policy such as, “Keep response time under 2 seconds, quality above this threshold, and gross margin above 60%.”
The router then decides which model gets each request.
“Model agnostic” stops being a developer preference. It becomes dynamic margin and risk management. The winning AI infrastructure may be the control plane, not the model.
9/ We discussed this with Matthew Schneider @mattschneid3r, CEO and founder of @BuildingInc, on episode 91 of Colors of Tech & Entrepreneurship.
youtu.be/WUn4qlPssuc



