5 ms·

Say Gottfried and Isaac are disputing over the creation of a mathematical method to solve problems. Now if Isaac writes a cryptic paper that non one understand

by atmosx 12y ago

Say Gottfried and Isaac are disputing over the creation of a mathematical method to solve problems. Now if Isaac writes a cryptic paper that non one understands (but Gottfried) and doesn't publish these finds on a 'well known Journal' it's very hard for him to take credit for his work. If however, if publish his paper in the blockchain, then there is a proof that he actually FOUND at THAT point in time what he claims to. The good thing is that, it's there for all to see and cannot be tampered. That's all Isaac will need in court to win his case over Gottfried.

amelius 12y ago
So I guess we could also use this to build a collection of prior-art, which cannot be ignored by the patent office (?) Have an idea, just throw it into the block chain. And you can still patent it later on (if nobody beats you to it, in which case they cannot either unless you come to an agreement), because the idea has not been formally published yet. I just hope it is all formally (legally) correct.
- maxerickson 12y ago
  Be careful deciding that storing information in the blockchain would not be considered publication. See II here: http://www.uspto.gov/web/offices/pac/mpep/s2128.html http://www.uspto.gov/web/offices/pac/mpep/s2128.html An electronic publication, including an on-line database or Internet publication, is considered to be a “printed publication”... You are also basically trying to have your cake (establish prior art by public disclosure) and eat it too (maintain your ability to file for a patent). I think hash proofs don't end up establishing prior art, because they don't make the idea available (Which seems to be a key aspect of the prior art test).
  - URSpider94 12y ago
    You're probably correct that this isn't enough to establish prior art unless you actually publish the material as well. However, it could be a good way to document a trade secret. Even if another entity is granted a patent, you can continue to practice a trade secret internally if you can prove that you were using it before the patent was filed.
maxerickson 12y ago
So what advantage does this have over a centralized system that works to maintain a reasonable reputation? For example, arXiv.org accepts things liberally and is careful about managing the timestamps of submissions: http://arxiv.org/help/replace#minorchanges http://arxiv.org/help/replace#minorchanges
- mpyne 12y ago
  One advantage is that it results in tons more carbon into the atmosphere (Bitcoin alone is estimated to require the electrical input of Ireland). This is only really an advantage for the energy sector though...
tezza 12y ago
Could you not simply tweet a hash of a document that has the secret IP in it? Then should the case come to court, you link to the Twitter tweet to show the submitted date, and produce the document for the court. The court hashes the document, does a comparison... The court accepts your argument No need for anything more complex. There are hashes out there with few collisions
- Two9A 12y ago
  You could, sure, but that depends on Twitter still being around if/when the case comes to court. The blockchain, being independently held and cross-verified at every node of the network, is more reliable in this regard.
  - e12e 12y ago
    Maybe just publish a sha512 hash of the document in a newspaper classifieds section... If the goal was to have recorded evidence for posterity. I wonder what the odds are that the current block-chains will be active in 70 years? Probably lower than the likelihood that all public tweets are recorded in the library of congress or some similar institution. Does a 60 year old archive of a block-chain that "lived" for 15 years of any value? It could probably be "faked" with a cluster of machines and some time? At any rate for things like a paper, we already have the ability to go and get a document notarized somewhere -- and that has much more legal precedent than some clever mathematical hack. Would be interesting to see law firms offer "adding to and helping maintain" a block chain as a supplement to regular notary service though. [ed: Ah, contributing to the slashdotifization of hn. I see that the article is about a company that tries to do this. I still wonder about what happens if/when "everyone" moves away from (the current) bitcoin (chain) to a new protocol. I suppose it could be "easy" to log the current state of the old chain, in a new chain. Would make verification more complicated though?].
    - gus_massa 12y ago
      I Agree. But I guess that the Library of Congress (and archive.com and Google Blockchains) will also save a copy of the major blockchais, but not of every obscure altcoin. Some anthropologists and economists may find them interesting. But then you will have to thrust again a central authority, and the magic of the totally decentralized blockchain will be lost. Another possibility when it's clear that another coin is surpassing Bitcoin is to "register" the Bitcoin blockchain in the new blockchain, just store into the new blockchain a copy of the current sha1. (Or a harder hash of all the blockchain, to eliminate the possibility of collisions found with faster machines.)
abcd_f 12y ago
That's just trusted timestamping, which is a far more generic (and therefore useful) matter than what these guys are trying to do.

by atmosx 12y ago

amelius 12y ago

So I guess we could also use this to build a collection of prior-art, which cannot be ignored by the patent office (?) Have an idea, just throw it into the block chain. And you can still patent it later on (if nobody beats you to it, in which case they cannot either unless you come to an agreement), because the idea has not been formally published yet. I just hope it is all formally (legally) correct.

maxerickson 12y ago
Be careful deciding that storing information in the blockchain would not be considered publication. See II here: http://www.uspto.gov/web/offices/pac/mpep/s2128.html http://www.uspto.gov/web/offices/pac/mpep/s2128.html An electronic publication, including an on-line database or Internet publication, is considered to be a “printed publication”... You are also basically trying to have your cake (establish prior art by public disclosure) and eat it too (maintain your ability to file for a patent). I think hash proofs don't end up establishing prior art, because they don't make the idea available (Which seems to be a key aspect of the prior art test).
- URSpider94 12y ago
  You're probably correct that this isn't enough to establish prior art unless you actually publish the material as well. However, it could be a good way to document a trade secret. Even if another entity is granted a patent, you can continue to practice a trade secret internally if you can prove that you were using it before the patent was filed.

maxerickson 12y ago

So what advantage does this have over a centralized system that works to maintain a reasonable reputation? For example, arXiv.org accepts things liberally and is careful about managing the timestamps of submissions: http://arxiv.org/help/replace#minorchanges http://arxiv.org/help/replace#minorchanges

mpyne 12y ago
One advantage is that it results in tons more carbon into the atmosphere (Bitcoin alone is estimated to require the electrical input of Ireland). This is only really an advantage for the energy sector though...

tezza 12y ago

Could you not simply tweet a hash of a document that has the secret IP in it? Then should the case come to court, you link to the Twitter tweet to show the submitted date, and produce the document for the court. The court hashes the document, does a comparison... The court accepts your argument No need for anything more complex. There are hashes out there with few collisions

Two9A 12y ago
You could, sure, but that depends on Twitter still being around if/when the case comes to court. The blockchain, being independently held and cross-verified at every node of the network, is more reliable in this regard.
- e12e 12y ago
  Maybe just publish a sha512 hash of the document in a newspaper classifieds section... If the goal was to have recorded evidence for posterity. I wonder what the odds are that the current block-chains will be active in 70 years? Probably lower than the likelihood that all public tweets are recorded in the library of congress or some similar institution. Does a 60 year old archive of a block-chain that "lived" for 15 years of any value? It could probably be "faked" with a cluster of machines and some time? At any rate for things like a paper, we already have the ability to go and get a document notarized somewhere -- and that has much more legal precedent than some clever mathematical hack. Would be interesting to see law firms offer "adding to and helping maintain" a block chain as a supplement to regular notary service though. [ed: Ah, contributing to the slashdotifization of hn. I see that the article is about a company that tries to do this. I still wonder about what happens if/when "everyone" moves away from (the current) bitcoin (chain) to a new protocol. I suppose it could be "easy" to log the current state of the old chain, in a new chain. Would make verification more complicated though?].
  - gus_massa 12y ago
    I Agree. But I guess that the Library of Congress (and archive.com and Google Blockchains) will also save a copy of the major blockchais, but not of every obscure altcoin. Some anthropologists and economists may find them interesting. But then you will have to thrust again a central authority, and the magic of the totally decentralized blockchain will be lost. Another possibility when it's clear that another coin is surpassing Bitcoin is to "register" the Bitcoin blockchain in the new blockchain, just store into the new blockchain a copy of the current sha1. (Or a harder hash of all the blockchain, to eliminate the possibility of collisions found with faster machines.)

abcd_f 12y ago

That's just trusted timestamping, which is a far more generic (and therefore useful) matter than what these guys are trying to do.