Pinned Tweet
Most DePIN teams forget this:
You can’t trust the data until you trust the device.
Without cryptographic validation at the edge, the entire reward system is vulnerable.
Privacy is good but provable contributor integrity is better.
And that’s the future of infrastructure.
2
577
Joe Joe retweeted
almost every financial product has been API-ified
we’ve already seen
> card issuance as an API
> checking accounts as an API
> yield as an API
> brokerage as an API
and now we’re seeing
> mortgages as an API
> auto loans as an API
> insurance as an API
> private credit as an API
> wealth management as an API
what other financial products will be boiled down to an API?
Lmao. While their on-chain risk model is smart, they already leaving a massive blind spot verifying the off-chain income streams actually feeding those wallets..
Your crypto portfolio is invisible to banks.
The credit system was built to read salary slips, bank statements, and bureau scores. That works fine if you keep your money in a bank.
A lot of crypto users function differently from that:
→ Freelancers in Argentina hold USDC because of peso inflation
→ Developers in Nigeria are likely receiving their salary onchain
→ Remote workers from the Philippines move money through crypto because it's faster than their country's banking rails
Their income and history are real, but they're not in the banking system.
These are the exact user groups that @surfcashx built loans for.
Here's how it works ↓
1. Wallet history
SurfCash reads your onchain transaction history and uses it to set a credit limit.
It doesn't pull a bureau score or ask for a bank statement, and it doesn't need a payslip from an employer your bank would recognize.
That makes a lot of sense because an onchain wallet already shows a lot of the signals lenders care about: money going in and out, spending patterns, repayment behavior, and consistency over time.
I'm just surprised no one has thought to read that information as credit thus far.
2. Unsecured credit
Most onchain lending still makes you lock up more than you borrow.
That’s collateral, not credit.
@surfcashx takes a different route by lending against wallet history. You borrow USDC based on your onchain reputation, without locking up your own capital first.
This is a crucial part.
Locking up collateral only works if you have spare capital, and a lot of people who earn and spend onchain don't want their funds tied up just to borrow against them.
This is exactly what opens credit up to the people who should've had it years ago. Wallet activity is more than enough to qualify a user, and you don't need extra capital lying around to prove you can be trusted with credit.
I know people who've been earning onchain for years, who would even be overqualified for these credits.
3. Local spending
Borrowing USDC onchain is only useful if you're able to spend it.
SurfCash covers borrowing in a few taps:
• One-tap signup (KYC pre-filled)
• Choose your amount and category
• USDC lands in your Solana wallet
From there, it can be spent through local payment rails across Vietnam, the Philippines, Brazil, Mexico, Nigeria, Argentina, and Peru (with more countries coming next).
Then you pay it back in USDC, onchain, in scheduled installments. The full loop is simple:
Hold → borrow → spend locally → repay onchain.
4. My take
Crypto has been promising to bank the unbanked for years, but most products still need you to bring capital to the table first, ready to lock, bridge, or stake.
The way I see it, if you already earn, save, and spend onchain, credit might be the only missing piece in that flow.
@surfcashx is legit bringing a novelty to the table.
Loans are currently live as well. Right now, 9,400 people already use it for daily spending, with 31,700 transfers processed so far.
If this is something you need, you can check it out at app.getsurf.cash.
1
69
Joe Joe retweeted
Most crypto neobanks promise the same trifecta:
> Yield.
> Custody.
> Availability.
But here’s the truth: you can only pick two.
Not three.
Want high on-chain yield + self-custody?
Try @Plasma One. But fiat ramps are limited.
Want a global card access and high APYs?
Use @cryptocom. But you give up your keys.
Want to hold your own assets + swipe Visa anywhere?
There’s @gnosisdao Pay. But don’t expect DeFi returns.
So far, no one nails all three.
(Mantle @UR_global might come closest, but it’s still early days.)
Every neobank lives on one edge.
→ Builders must decide what not to optimize.
→ Users must know what they’re sacrificing.
Self-custody, stablecoin yield, and spending power can coexist, but until then, not at full strength.
I'm cautiously optimistic.
The triangle may shrink.
But for now: the trilemma holds.
Accept your tradeoff.
Choose wisely.
Joe Joe retweeted
dollar denominated neobanks are cool but there's a reason why wise and revolut has so many users GLOBALLY
being able to hold multi currency in a neobank makes sense so users don't have to always pays fx fees
if a crypto neobank could hold multiple currency stablecoins that would be pretty dope and be a huge differentiator in the current market
Joe Joe retweeted
You always Hear, "The World is evolving and if you dont stay in Vogue you Might Miss it"
And that's True!!
Everyday, there is a new trend, a new way in which live is made easier and today @EvolveTKN is making owning RWA Easier and here is how🔻
Joe Joe retweeted
who is active?
I’ll be hosting an AMA session for @Coinkove by 6pm tomorrow ~ Friday
$50 Giveaway included
$10 to 5 people each
To qualify:
-follow: @Coinkove
-like and Rt
-Set reminders
-tag 2 frens
Set a reminders for the Space! xitter.cf/i/spaces/1yxBeMkvEYrJN
Joe Joe retweeted
Nigeria is ranked #1 in global USDT and USDC ownership.
Not the US. Not the UK. Not Singapore.
Nigeria.
59% of Nigerian crypto users hold USDT. 48% hold USDC. More than any other country surveyed. India is third. Brazil close behind.
The reason is obvious once you see it: in countries where local currency loses 20–40% of value annually, stablecoins aren't a crypto product. They're a savings account. A dollar-denominated store of value that doesn't require a US bank account.
Here's what the data doesn't show: most of those stablecoin holders can't use their USDT to buy anything.
No merchant acceptance. No subscription billing. No automatic payment. No way to pay a supplier. They hold the dollars. They can't spend the dollars. They convert back to fiat every time they need to transact.
The gap between stablecoin adoption and stablecoin utility is most visible not in San Francisco or Singapore. It's in Lagos, Jakarta, São Paulo.
$308B in circulation. The people who need stablecoin commerce most have the fewest tools to access it.
That's not a distribution problem. That's an infrastructure problem. The rails exist everywhere. The billing layer doesn't exist anywhere.
That's who we're building for.
𝗠𝗼𝘀𝘁 web 3 𝘁𝗲𝗮𝗺𝘀 𝗱𝗼𝗻’𝘁 𝗳𝗮𝗶𝗹 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝗼𝗳 𝗯𝗮𝗱 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻.
They fail because they execute inside the wrong constraint.
You can hire world-class engineers. Hit every sprint.
And still lose if you’re solving a problem the market no longer prices
55
Calling yourself a builder because you vibecode prompt without understanding execution, constraints, or failure modes
is like calling yourself an engineer because you assembled IKEA furniture
Tools don’t grant authorship
When something breaks, the skill gap shows immediately.
1
54
Vibe coding is where the next $500B company will start.
But it won’t be built by people chasing vibes.
It’ll be built by teams using new tools to ship real systems fast.
Revenue-first. No VCs. No angels.
Target: $100k ARR.
Constraint: ship, break, fix, repeat.
Build. Ship.
1
42
The winning DeFAI Teams of 2026 will be those making their systems defensible under scrutiny.
1. Verifiable execution,
2. provable decisions and
3. clear responsibility boundaries.
Anything else is engineering effort spent in the wrong layer.
1
29
Most teams are misapplying zkML by aiming it at the wrong layer.
zkML is not for LLMs.
zkML is for the computations that actually move value.
LLMs only assist agents, models decide "outcome," and only "outcome" decisions need "proof."
1
25
Classical ML only are deterministic and provable.
A sane DeFAI stack looks like this:
LLMs → planning, reasoning, tool selection
Models → execution, allocation, capital movement
zkML → proof over the model outputs.
If your agents final decisions come directly from an LLM,
1
22
If you want to truly win in 2026 -
Hard work doesn’t move you up.
It only just keeps you from sinking.
What actually moves you is:
→ Leverage (distribution, capital, network, tech edge)
→ Positioning (right arena, right timing, right narrative)
→ Sharp signal, not effort
1
34
You don’t win because you build, ship, or grind hardest.
You win because you’re plugged into accelerating systems right before they compound.
People underestimate this:
You can be a monster of discipline and still be strategically invisible.
1
27
