not really, not quite.
but you could play with this machine. It is auditable - the only such on the market - and has little blinking lights for power on/off
http://nosuchlabs.com/
Topic: provable destruction of information (Read 182 times)
Is this possible?
an algorithm or process of some sort that can take a piece of entropy or information (e.g. a random number), use it (e.g. as a seed for something), then erase/destroy the source information, provably, so that it's gone and can't be retrieved.
Obviously no human can set eyes on the original number.
Edit: this could be used for secret sharing, for example. Pieces of the original number could be doled out to several parties who would then be able to reconstruct it. And you could use combinatorics to create an m of n scheme. Simple example: divide the info into three pieces, N1 N2 and N3. Give Alice N2 and N3. Give Bob N1 and N3. Give Charlie N1 and N2. Then you have a two-of-three sharing scheme where two people out of the three are necessary and sufficient to reconstruct the source entropy. This could be done with an arbitrarily large m of n sharing set. You could set it to have four of seven people, or as many as you wish, with the right combinatorics.
The challenge here is how do you prove that the machine that divvied up the original number didn't keep it in memory.
an algorithm or process of some sort that can take a piece of entropy or information (e.g. a random number), use it (e.g. as a seed for something), then erase/destroy the source information, provably, so that it's gone and can't be retrieved.
Obviously no human can set eyes on the original number.
Edit: this could be used for secret sharing, for example. Pieces of the original number could be doled out to several parties who would then be able to reconstruct it. And you could use combinatorics to create an m of n scheme. Simple example: divide the info into three pieces, N1 N2 and N3. Give Alice N2 and N3. Give Bob N1 and N3. Give Charlie N1 and N2. Then you have a two-of-three sharing scheme where two people out of the three are necessary and sufficient to reconstruct the source entropy. This could be done with an arbitrarily large m of n sharing set. You could set it to have four of seven people, or as many as you wish, with the right combinatorics.
The challenge here is how do you prove that the machine that divvied up the original number didn't keep it in memory.
Jump to: