Last week someone sent me a link to a $BABY co-founder account with one instruction: don't read the post first. Read the reply.
The reply was short, almost casual about its own news. He mentioned that research doesn't usually reach production this quickly, then moved on like it wasn't a major point.
That made me curious enough to track down the paper myself.
The paper, BABE, appeared on the IACR Cryptology ePrint Archive in January 2026. It focuses on one of the larger practical costs in BitVM3. In the baseline construction discussed by the paper, garbled circuits require around 42 GB of storage.
The authors propose a construction that reduces those storage requirements by roughly three orders of magnitude while preserving BitVM3's low on-chain transaction profile.
One detail also stood out to me. David Tse, a Babylon co-founder, is listed as one of the paper's authors.
A few months later, on June 2, 2026, Babylon announced the public testnet for Trustless Bitcoin Vaults with Aave v4.
I can't say the paper directly led to that launch, and Babylon doesn't explicitly make that claim. What caught my attention was the timeline. A research result addressing a significant BitVM3 bottleneck appeared only a few months before a product that relies on BitVM-based verification reached public testnet.
That sequence feels more interesting to me than the borrowing feature itself.
A lot of projects announce Bitcoin-backed lending. I pay closer attention to the infrastructure work that happens before those announcements, because improvements in proof efficiency can make trust-minimized systems more practical to deploy over time.
I almost didn't click that reply either.
@BabylonLabs_io #baby $BABY
The reply was short, almost casual about its own news. He mentioned that research doesn't usually reach production this quickly, then moved on like it wasn't a major point.
That made me curious enough to track down the paper myself.
The paper, BABE, appeared on the IACR Cryptology ePrint Archive in January 2026. It focuses on one of the larger practical costs in BitVM3. In the baseline construction discussed by the paper, garbled circuits require around 42 GB of storage.
The authors propose a construction that reduces those storage requirements by roughly three orders of magnitude while preserving BitVM3's low on-chain transaction profile.
One detail also stood out to me. David Tse, a Babylon co-founder, is listed as one of the paper's authors.
A few months later, on June 2, 2026, Babylon announced the public testnet for Trustless Bitcoin Vaults with Aave v4.
I can't say the paper directly led to that launch, and Babylon doesn't explicitly make that claim. What caught my attention was the timeline. A research result addressing a significant BitVM3 bottleneck appeared only a few months before a product that relies on BitVM-based verification reached public testnet.
That sequence feels more interesting to me than the borrowing feature itself.
A lot of projects announce Bitcoin-backed lending. I pay closer attention to the infrastructure work that happens before those announcements, because improvements in proof efficiency can make trust-minimized systems more practical to deploy over time.
I almost didn't click that reply either.
@BabylonLabs_io #baby $BABY