Digital Legacies on the SAFE Network


This really interesting book, which I admit I haven’t even properly dug into, has got me thinking about digital legacies and how they will manifest themselves on the SAFE Network:

The beautiful thing about regaining control over your personal data, is that you no longer have to trust hundreds (>!) of third parties with your personal data after death; and how they may treat your wishes and the needs of your family and estate after you’re gone.

You’d have the opportunity to be really specific if you wished.

So rather than surviving friends, enemies, and corporations interpreting vague laws in a multitude of jurisdictions, and getting access to your whole digital life (or none, or anything in between) perhaps you could have a safe deposit box system, with the keys given to a trusted party (say your solicitor, to be included in a will), and anything private data tagged as part of your legacy is accessible via those keys.

Likewise, it’s worth considering what happens to all your private, unshared data, after your death. Is it a responsible thing to allow the network to store and maintain all that data, that will never be accessed again? Is it a good use of resources?

Perhaps there could be a system where (optionally) all private files, could drop off the network, at a specific time, say 100 years after your last login :grinning:


No need to do that, in 100 years from now 1TB will be like 2KB today. :slight_smile:


It’s not just storage though is it? There’s bandwidth and processing too, and all for data which can never be accessed again.

And if you consider there will eventually be more dead SAFE Network users than living, it’s perhaps responsible to think about the options.


The same story here, transfer of 1TB of data in 100 years will cost similar to the cost of transferring 2KB today. Basically nothing.


I would not worry. Computer stuff grows exponentially. There will be this little bit of dead weight but that will be a drop in the bucket when the next generation is adding the same resources from an external hard drive sized thing as todays data centers.

In any case I am excited a lot of my personal info would have the ability to perhaps survive for 1000 years and then some future archeologist could check me out to see how people in our time lived. So I would like some kind of way to make everything public after 200 years. Like I don’t see any harm to myself or my friends to tell everything at that point haha. I think it could turn out to be this huge reservoir of historical knowledge about humans. To just be like never tell anyone would have to be an option for sure. I think most people would think it was cool/desirable to be part of this project to give future historians so much data to analyze. In the short term I hear you are pitching your project to universities for funding. I would mention this potential future anthropological academic interest to them in addition to the immediate computer science academic interests.


I suspect in 200 years computer science will do that for you :wink: maybe a bit longer, but anyway I believe what we have as secure will not be in hundreds of years. Most encryption is a function that uses a key to encrypt. The key is smaller than the data and needs to be found. RSA it was factoring large primes (easy now up to 512 bit keys), EC is solving discreet log problems. These are all likely to be solved in hundreds of years, quantum computing will certainly have a big impact. Therefore it is likely that in all that time today’s secrets all over the internet will be public documents, probably even secure email between spy agencies etc.


ok so assume your data will become public one day unless you have a self destruct timer set to wipe it totally at some point. in that case might as well build it right in how long until you would like people to wait to open you time capsule then it just automatically becomes public without hacking it. If someone digs it up earlier can’t really help that other then to have already self destructed. I would make it a clear option though. Like when you make a new ID you have to select either a wipe date or a public date since in between is not really an option in the very very long run so its best people make a choice on that.

1 Like

Think more like this, your data even encrypted is probably already copied, how many times, where etc. is all up for grabs. Even stored locally in your own hard drive you never show to anyone and destroy in a furnace after not looking at it while connected to the Internet , etc. etc. even then and even in such a paranoid state where you have apparently prevented access then you need to trust.

  1. Trust the OS is bug free and does not leak
  2. Trust the hardware is not compromised or buggy (Intel ???)
  3. Trust no camera see you using it etc.

So even manically paranoid use of data can still leak. However we live in a more real world. I think it is safe to just imagine all data we currently have is probably stored in some “storage” by various agencies and users. It will one day be decoded and analysed (probably by AI) in the future.

I am not being pessimistic here, but probably I do think we need to face a reality that current computer / crypto intelligence is likely infantile in a future. We don’t know how far in the future that will be, but exponential increases in AI capability is hard for us mortals to understand.

tl;dr, We can protect for the near future, for the longer term then it is best to assume if you put any info in a digital medium then it will be readable, regardless of today’s encryption. One time pad can help, but it is hard to do (our self-encrypt comes close for sure) as you need truly random data, but truly random is a math notion like 1+1=2 and it likely does not exist.


I was thinking very long timescales (e.g. 100 years after last login) as a proxy for time of death, but it may well be desirable to have far shorter time horizons than that; either to say “let’s make all this public” like @andyypants’ cool suggestion, or to say, “I no longer need the network to maintain these data”.

Another consideration would be bequeathing Safecoin, and how that might be handled.

Interesting paper here on the predicted number of deceased Facebook users by 2100:

Spoiler: Between 1.4 and 4.9 BN!

If SAFE follows as similar growth trend, that’s a lot of locked away Safecoin, does that matter to the ecosystem?

1 Like

Unless your sealed will contains the keys to get it.


Precisely (20 char) .

1 Like

When the inactive digital footprints of past humans outnumber those of living people, I wonder how much effort will be placed in cracking whatever encryption we implement today? In a sense, it would be sad if the only things a future civilization could know about ours are the things we chose to make public.


I think by then decrypting or just reading will be similar and probably you would not know the difference in terms of how long it took to open such files. Of course we prevent those intrusions by emitting CO2 and Methane :slight_smile: The ultimate future proof encryption saviour :smiley: :smiley:

1 Like

–> Cracking the cryptographic puzzle from 1999 below took 3,5 years.


Yes I do see your point that eventually AI could figure out the non random pattern and crack the encryption. I think one solution to this would be a physically triggered seed. Like if a radioactive molecule has a 50% chance of decaying in 10 minutes, if it decays when we check it in 10 minutes we seed with 0 and if not seed with 1. Basically like a large sequence of Schrodinger cats. I think this would work unless AI cracks some kind of non random pattern at the subatomic level where things do indeed appear to be truly random based on our present understanding. I mean if it does find a pattern ok I think that’s a good trade: I need to find a new way to generate randomness but on the bright side we have discovered a deep secrete of the universe :slight_smile: Of course you can still decode by brute force, but any “pattern” would just be an illusion that would disappear the larger the data set becomes. So I can’t deny for this to be feasible I need a way to measure the decay of many separate atoms in a predefined sequence and translate that into a large enough seed brute force can’t find it easily… but hey we are talking about 100 years in the future I thought, so maybe these devices will be easy to come by then. In the meanwhile we could use “almost random” physical events; like let the location of lightning strikes in the next hour generate a key. Many wallets use the “draw something” approach. And I take it that is something like what you have right now: some kind of mathematical way to be “almost random” but still at some level there are deterministic events leading to the outcome. Like some kind of thing with prime numbers or pi seems like it could produce an “almost randomness”

So you see the long term solution is not to pump out more methane and CO2… its to make radioactive waste lol. I would say if you are going to pitch this to a university it is interesting to anthropology and computer science, as mentioned before, but also PHILOSOPHY

While that might mater to individuals (in which case they should have it in their will that an attorney or someone trustworthy hands over the keys) I don’t think it really matters to the larger ecosystem. If some safecoin gets burned it just means what is left buys more. I assume safecoin will go to at least like 8 decimal places. It only really messes things up when we start to hit the limits of dividing it IMO