1/ Governments from US, Canada, parts of Europe are now forcing social media platforms to verify your age using national IDs.
Reddit. Spotify. Discord. All asking for your ID & photo just to access features.
This is only going to get more common. And the way we verify identity today is a disaster waiting to happen.
@senatorshoshana@mullvadnet A snippet of how a private verification infrastructure can look like and ensure privacy for users.
*Passport is used as an example here; It can be implemented on a national level based on a specific jurisdictions' identity mechanisms.
This is exactly the difference between privacy as a wrapper and privacy at the asset layer.
Zama's cUSDC still depends on USDC, so Circle can freeze the underlying contract.
Zcash is structurally different: shielded ZEC is native to the protocol, not pooled inside an issuer-controlled ERC-20. No Circle-style blacklist exists for shielded notes.
That doesn't mean Zcash solves everything. Today, shielded ZEC is primarily used for private transfers and payments rather than powering large DeFi ecosystems, where users generally prefer stable-value assets for collateral and settlement.
The deeper lesson is that privacy and issuer control are separate problems. Even the strongest privacy layer cannot remove the dependency on a centralized stablecoin issuer.
Privacy built on top of an asset is ultimately constrained by the rules of that asset.
26/ Continuation to this: https://t.co/N0wGj9k41n
This is also where GPU finance starts looking less like DeFi and more like traditional infrastructure finance.
The focus shifts toward:
• Cash flows
• Collateral quality
• Monitoring
• Risk management
Not token prices.
1/ Everyone is talking about AI chips.
NVIDIA's rise.
Massive data centers.
OpenAI's compute needs.
Multi-billion-dollar infrastructure plans.
But almost nobody is talking about the layer that historically scaled every major industry: The credit market behind it.
A thread on why AI may need its own debt market 🧵
25/ In this model, data centers become surprisingly important. Not just as infrastructure providers. But as trust anchors.
They provide the physical custody layer that makes financing possible.
Once the hardware is installed, additional monitoring becomes possible:
• Utilization
• Workloads
• Revenue generation
• Operational status
The lender gains visibility into the asset supporting the loan.
50/ The bet is not that GPUs are valuable. The bet is that compute may become important enough to support its own category of debt.
And if that happens, the biggest opportunity may not be another AI model.
It may be the financial infrastructure that allows intelligence itself to scale.
26/ Continuation to this: https://t.co/N0wGj9k41n
This is also where GPU finance starts looking less like DeFi and more like traditional infrastructure finance.
The focus shifts toward:
• Cash flows
• Collateral quality
• Monitoring
• Risk management
Not token prices.
1/ Everyone is talking about AI chips.
NVIDIA's rise.
Massive data centers.
OpenAI's compute needs.
Multi-billion-dollar infrastructure plans.
But almost nobody is talking about the layer that historically scaled every major industry: The credit market behind it.
A thread on why AI may need its own debt market 🧵
49/ Do GPU clusters become the next infrastructure asset class?
Do compute operators eventually finance hardware the way airlines finance aircraft?
Do dedicated credit markets emerge around AI infrastructure?
Those questions remain unanswered. But that's what makes the thesis interesting.