I’ve been thinking about Newton Protocol’s use of BN254 certificates, and honestly I keep circling back to something a bit less exciting than the cryptography itself. What seems more interesting to me is how much verification can be packed into something so small.
Most systems, at least the ones I’m used to looking at, seem to assume that stronger verification comes with a higher cost. You replay more state, you check more signatures, you reconstruct more of the path that led to the result. Confidence and cost usually move together, or at least that is the assumption.
Newton seems to push against that a little.
When a destination chain like Base receives a BN254 certificate, it is not re-running the whole process that happened on the source chain. It is checking a compressed version of it. The operators, the stake, the signatures, the consensus behind the decision, all of that gets folded into something small enough to verify cheaply through Ethereum’s BN254 precompiles. That part is interesting, but what matters more to me is the shape of the trade-off.
Because verification wants evidence, but scalability wants shortcuts. And those two things do not always sit comfortably together. If you try to preserve every detail, verification gets expensive very quickly. If you compress too aggressively, then at some point you are no longer really checking the thing itself, you are checking a representation of it. Maybe that is acceptable. Maybe that is exactly the point. But it does create a gap.
Newton seems to live right inside that gap.
A destination chain accepts the certificate without having to relive the process that produced it. So the certificate is not the decision, not really. It is more like a compact claim that the decision was made under conditions the system is willing to trust.
And that keeps bringing me back to verification, not in the abstract sense, but in the practical one. How little information can move across chains before verification starts to feel less like verification and more like trust in compression?
#newt $NEWT @NewtonProtocol
Most systems, at least the ones I’m used to looking at, seem to assume that stronger verification comes with a higher cost. You replay more state, you check more signatures, you reconstruct more of the path that led to the result. Confidence and cost usually move together, or at least that is the assumption.
Newton seems to push against that a little.
When a destination chain like Base receives a BN254 certificate, it is not re-running the whole process that happened on the source chain. It is checking a compressed version of it. The operators, the stake, the signatures, the consensus behind the decision, all of that gets folded into something small enough to verify cheaply through Ethereum’s BN254 precompiles. That part is interesting, but what matters more to me is the shape of the trade-off.
Because verification wants evidence, but scalability wants shortcuts. And those two things do not always sit comfortably together. If you try to preserve every detail, verification gets expensive very quickly. If you compress too aggressively, then at some point you are no longer really checking the thing itself, you are checking a representation of it. Maybe that is acceptable. Maybe that is exactly the point. But it does create a gap.
Newton seems to live right inside that gap.
A destination chain accepts the certificate without having to relive the process that produced it. So the certificate is not the decision, not really. It is more like a compact claim that the decision was made under conditions the system is willing to trust.
And that keeps bringing me back to verification, not in the abstract sense, but in the practical one. How little information can move across chains before verification starts to feel less like verification and more like trust in compression?
#newt $NEWT @NewtonProtocol