1HowardWu

It's very surprising how far we've gotten as an industry without private smart contracts.

We've found product-market fit in trading, stablecoins, and basic DeFi all despite lacking fundamental privacy. Imagine what's possible once we enable Web2-like UX on crypto rails.

We have come a long way.

When we started, developers had to handwrite ZK circuits.

Now with Leo, we've created a programming language that looks and feels like JavaScript but compiles to ZK circuits.

The goal: letting web developers build private applications without reasoning about cryptography.

Some Aleo lore (when we restarted the whole project):

Building a general-purpose ZK VM meant rethinking circuit architecture from first principles.

We moved from R1CS gadget libraries to an opcode-based design where every operation maps to a single circuit. This enables program-level upgradability and formal verification.

Two years into development, we hit a wall.

Our formal verification team discovered bugs in our circuits, but with the gadget-lib approach, we couldn't isolate which applications used the vulnerable components. It was impossible to trace and identify which parts needed fixing.

We made the painful decision to scrap our entire design and rebuild from scratch.

Instead of a monolithic circuit structure, we created a clean separation: every operation (add, mul, div) became its own distinct circuit with clear boundaries.

This architectural shift parallels traditional computing.

Just as bytecode creates an abstraction layer in the EVM, Aleo Instructions create a boundary between the language (Leo) and the underlying circuits. The language becomes a thin compiler targeting these instructions.

The benefits are enormous: if a vulnerability is discovered, we can precisely identify affected programs, upgrade specific instructions, and maintain backward compatibility. This is critical for a production system handling real assets.

This architecture also unlocks formal verification. Each instruction can be individually verified, creating mathematical certainty about system behavior. Without this architecture, upgrades would require replacing everything, creating massive security risks.

Sometimes the hardest decisions lead to the best outcomes. Rebuilding our VM from scratch cost us nearly a year, but created a foundation that can evolve securely for decades to come.

This *was* Hilton's website.

It made me think of how privacy in blockchain is like HTTPS for the web.

In the 90s, Hilton dot com was just a static page with a phone number.

No online bookings until HTTPS arrived.

Similarly, real-world blockchain adoption awaits proper privacy infrastructure. We're building the HTTPS moment for crypto.

People have no idea how big this is.

This is the biggest opportunity I can think of:

Did you know that less than 0.01% of stablecoin payments in 2024 were encrypted?

The overwhelming majority left the sender, receiver, and amount completely vulnerable.

BTC brought decentralization. ETH brought programmability. ZEC brought privacy.

$ALEO brings all three.

It's incredible how far ZK has come.

So many people are talking about it these days.

That wasn't the case 5 years ago.

And we're still early.

I love to see it.

Thanks for having me @consensus2025!

I recently had a great conversation with @DLNewsInfo about @AleoHQ and ZK.