Lead author here. The mods have graciously given me permission to announce some computer science work as a Show HN. It concerns permissionless Byzantine consensus – the notoriously difficult problem of how to securely replicate a state machine in the absence of a reliable identity system, which is the underpinning of Bitcoin and other decentralized ledgers.
By copying the signaling techniques used by animals, my co-author and I have achieved a 40,000x improvement in security and performance over the prior state of the art. This vindicates a prediction made 10 years ago by a Chinese researcher, one of the world's rare dual-PhDs in biology and computer science, who believed that reverse-engineering animal-communication networks could produce a consensus-protocol breakthrough similar to the invention of public-key cryptography.
The parallel between asymmetric encryption and our discovery goes beyond the scale of the advancement. It actually concerns the mechanism that our protocol uses to protect itself from pseudo-spoofing or "Sybil" attacks, in which an entity uses sockpuppets to hijack consensus by casting extra votes. Existing technologies, like Proof-of-Work and Proof-of-Stake, are symmetric in the sense that they require correct agents to "outbid" the adversary by verifiably expending more money or computing power. If the adversary's budget for an attack is greater than the security budget of honest protocol participants, then the entire system collapses.
Our paper introduces the first asymmetric system, Proof-of-Balance. It guarantees that honest protocol participants will remain in control of the consensus protocol, even if their security budget is many times smaller than the adversary's budget for an attack. This verifiable asymmetry yields not only an exponential improvement in security, but also a corollary increase in performance: because the adversary's maximum fraction of total voting power is tightly constrained, transactions can be processed on the open internet using speeds that were previously possible only on permissioned networks.
Asymmetry is nothing new in access control – e.g., a lock increases the security of a house by more than its purchase price, so homeowners aren't forced to "outbid" burglars to keep their families safe. However, it has been ignored in resource-weighted consensus, because the field has been guided by the "handicap principle" – which claims that the reliability of a signal depends on its verifiable cost to the signaler. Bitcoin enthusiasts often expressly invoke this principle to justify the waste inherent in Proof-of-Work, claiming that it is a universal law of nature, which applies with equal force to biology and computer science.
Not so. That is close to what biologists believed in the 1990s, when formal game-theoretic modeling first substantiated the concept of handicap-authenticated signaling. However, subsequent work revealed that it is actually the verifiable cost of faking a signal that determines whether information can be transmitted reliably. If the cost for a dishonest entity to spoof a signal is sufficiently high, then honest agents can transmit reliable signals at zero cost. This is known as cue-authenticated signaling, and it is the key to our protocol, KRNC ("Key Retroactivity Network Consensus").
An intuitive example of the difference between handicap-authenticated signaling and cue-authenticated signaling is how male peacocks and tigers signal their fitness to potential mates. Male peacocks waste resources growing oversized tails, a handicap that proves their fitness based on the amount of self-inflicted punishment they can endure. Male tigers compete with one another to grow as large as possible to gain an edge in lethality, and their size happens to have the added bonus of providing a cue of their fitness.
We adapt the "cue principle" to obtain a novel solution to Goodhart's Law, the adage that a measure ceases to be a...
> An intuitive example of the difference between handicap-authenticated signaling and cue-authenticated signaling is how male peacocks and tigers signal their fitness to potential mates. Male peacocks waste resources growing oversized tails, a handicap that proves their fitness based on the amount of self-inflicted punishment they can endure. Male tigers compete with one another to grow as large as possible to gain an edge in lethality, and their size happens to have the added bonus of providing a cue of their fitness.
Sounds like r/K selection theory in evolutionary biology.
This looks fantastic, I can't wait to read this all the way through. It seems to solve a supreme fundamental issue of removing competition from the proofing.
> I want to use the money for effective altruism and existential-risk reduction.
Since the absolute most fundamental nature of any organization is risk reduction, I believe you mean the underpinnings of what the poli sci world is calling the 'network state' :)
However, with that said, to me the most promising goal of crypto will forever be the exchanging of computations for social mobility, ie PoW, PoS, PoST, etc. There needs to be some way to actually earn semi-passive income. Why work when a mining rig can do it for you? Banking the unbanked is cool or whatever, but unbossing the bossed is where the ultimate potential for humanity is at, imho. How does this consensus mechanism help with that, if there is no staking or mining?
In a sense, it's an alternate form of Proof-of-Stake. As Section 8.4 explains, the conventional wisdom is that Proof-of-Stake's flaw is that it's circular. We've proved that actually it's not circular enough, i.e., the stakes it assigns are different than the stakes in society's existing monetary game.
Proof-of-Balance allows "stakes" (what we call "weight") to be issued in proportion to monetary balances. Once those stakes are in users' hands, the protocol can run using the algorithms designed for PoS, including all of their reward and governance mechanisms.
It turns out that to fully unleash the power of those algorithms, you need a verifiably secure stake-distribution mechanism. That's what we've invented. (It's harder than it sounds, of course...)
13 comments
[ 2.7 ms ] story [ 42.6 ms ] threadLead author here. The mods have graciously given me permission to announce some computer science work as a Show HN. It concerns permissionless Byzantine consensus – the notoriously difficult problem of how to securely replicate a state machine in the absence of a reliable identity system, which is the underpinning of Bitcoin and other decentralized ledgers.
By copying the signaling techniques used by animals, my co-author and I have achieved a 40,000x improvement in security and performance over the prior state of the art. This vindicates a prediction made 10 years ago by a Chinese researcher, one of the world's rare dual-PhDs in biology and computer science, who believed that reverse-engineering animal-communication networks could produce a consensus-protocol breakthrough similar to the invention of public-key cryptography.
The parallel between asymmetric encryption and our discovery goes beyond the scale of the advancement. It actually concerns the mechanism that our protocol uses to protect itself from pseudo-spoofing or "Sybil" attacks, in which an entity uses sockpuppets to hijack consensus by casting extra votes. Existing technologies, like Proof-of-Work and Proof-of-Stake, are symmetric in the sense that they require correct agents to "outbid" the adversary by verifiably expending more money or computing power. If the adversary's budget for an attack is greater than the security budget of honest protocol participants, then the entire system collapses.
Our paper introduces the first asymmetric system, Proof-of-Balance. It guarantees that honest protocol participants will remain in control of the consensus protocol, even if their security budget is many times smaller than the adversary's budget for an attack. This verifiable asymmetry yields not only an exponential improvement in security, but also a corollary increase in performance: because the adversary's maximum fraction of total voting power is tightly constrained, transactions can be processed on the open internet using speeds that were previously possible only on permissioned networks.
Asymmetry is nothing new in access control – e.g., a lock increases the security of a house by more than its purchase price, so homeowners aren't forced to "outbid" burglars to keep their families safe. However, it has been ignored in resource-weighted consensus, because the field has been guided by the "handicap principle" – which claims that the reliability of a signal depends on its verifiable cost to the signaler. Bitcoin enthusiasts often expressly invoke this principle to justify the waste inherent in Proof-of-Work, claiming that it is a universal law of nature, which applies with equal force to biology and computer science.
Not so. That is close to what biologists believed in the 1990s, when formal game-theoretic modeling first substantiated the concept of handicap-authenticated signaling. However, subsequent work revealed that it is actually the verifiable cost of faking a signal that determines whether information can be transmitted reliably. If the cost for a dishonest entity to spoof a signal is sufficiently high, then honest agents can transmit reliable signals at zero cost. This is known as cue-authenticated signaling, and it is the key to our protocol, KRNC ("Key Retroactivity Network Consensus").
An intuitive example of the difference between handicap-authenticated signaling and cue-authenticated signaling is how male peacocks and tigers signal their fitness to potential mates. Male peacocks waste resources growing oversized tails, a handicap that proves their fitness based on the amount of self-inflicted punishment they can endure. Male tigers compete with one another to grow as large as possible to gain an edge in lethality, and their size happens to have the added bonus of providing a cue of their fitness.
We adapt the "cue principle" to obtain a novel solution to Goodhart's Law, the adage that a measure ceases to be a...
Didn't mean to imply any sort of endorsement beyond that.
I wrote about it for the Washington Post, as well: https://www.washingtonpost.com/opinions/how-jeff-sessions-ca...
Sounds like r/K selection theory in evolutionary biology.
> I want to use the money for effective altruism and existential-risk reduction.
Since the absolute most fundamental nature of any organization is risk reduction, I believe you mean the underpinnings of what the poli sci world is calling the 'network state' :)
However, with that said, to me the most promising goal of crypto will forever be the exchanging of computations for social mobility, ie PoW, PoS, PoST, etc. There needs to be some way to actually earn semi-passive income. Why work when a mining rig can do it for you? Banking the unbanked is cool or whatever, but unbossing the bossed is where the ultimate potential for humanity is at, imho. How does this consensus mechanism help with that, if there is no staking or mining?
Thanks again, I can't wait to read it!
In a sense, it's an alternate form of Proof-of-Stake. As Section 8.4 explains, the conventional wisdom is that Proof-of-Stake's flaw is that it's circular. We've proved that actually it's not circular enough, i.e., the stakes it assigns are different than the stakes in society's existing monetary game.
Proof-of-Balance allows "stakes" (what we call "weight") to be issued in proportion to monetary balances. Once those stakes are in users' hands, the protocol can run using the algorithms designed for PoS, including all of their reward and governance mechanisms.
It turns out that to fully unleash the power of those algorithms, you need a verifiably secure stake-distribution mechanism. That's what we've invented. (It's harder than it sounds, of course...)