3 ms·

DNA is probably your best bet for bottom up approach. Its surprisingly difficult data to find and the rate probably varies a lot, but wikipedia claims: "The h

by VLM 9y ago

DNA is probably your best bet for bottom up approach. Its surprisingly difficult data to find and the rate probably varies a lot, but wikipedia claims: "The human germline mutation rate is approximately 0.5×10−9 per basepair per year" My RAID array does better, polaroid pictures somewhat worse, etc. A bit error rate of 1e-9/yr is something that can be worked around over a lifespan but probably not over a million years. https://en.wikipedia.org/wiki/Mutation_rate https://en.wikipedia.org/wiki/Mutation_rate

dmurray 9y ago
Why not over a million years? That would still only be 0.1% of bits flipped, and it should be possible to design a code with enough redundancy or error correction to take care of that. Is there some compounding effect I'm not thinking of? Or simply that it's hard to keep a species alive for a million years?
- VLM 9y ago
  Ouch I did some research and found "Mammals, for instance, have an average species "lifespan" from origination to extinction of about 1 million years, although some species persist for as long as 10 million years." Evolution: Library: The Current Mass Extinction - PBS www.pbs.org/wgbh/evolution/library/03/2/l_032_04.html So yeah, it would be a shame if space aliens visited and encoded a message to us in saber toothed tigers or trilobites. Also I have no religious hangups, I'm a direct descendant of the ancestor of amoebas or whatever exactly, and if the message were encoded into whatever makes an amoeba an amoeba, well, its definitely gone now. I think it would be hard to "infect" a species with stored data. They'd have to burn more protein to replicate that extra DNA material, which would provide a competitive advantage to any individual not carrying the data, so its not going to last long. Even thinking outside the box with a virus, wouldn't smaller ones have better odds of survival?
  - Razengan 9y ago
    > I think it would be hard to "infect" a species with stored data. ... Yeah, but what about encoding a message/knowledge into the mating rituals of an species (or artificially-modified subspecies) from the start? [0] > if the message were encoded into whatever makes an amoeba an amoeba, well, its definitely gone now. Are we sure humans don't have any "useless" DNA? What about seemingly superfluous organs like the appendix? [0] https://news.ycombinator.com/item?id=14548854 https://news.ycombinator.com/item?id=14548854
Chaebixi 9y ago
But if you're just storing random data, like text, it probably offers little to no fitness benefit to the organism. So it could all be lost with a random deletion and no one but you would care. https://en.wikipedia.org/wiki/Deletion_(genetics) https://en.wikipedia.org/wiki/Deletion_(genetics)
- Razengan 9y ago
  > if you're just storing random data, like text, it probably offers little to no fitness benefit to the organism. Wow, this makes me even more interested in the idea of using mating rituals as a mechanism for very-long-term knowledge storage! Things like birdsongs, dances, and the aforementioned courtship art... If an individual can only reproduce if it correctly retransmits the "data", then the data will survive as long as the species does..no?

by VLM 9y ago

dmurray 9y ago

Why not over a million years? That would still only be 0.1% of bits flipped, and it should be possible to design a code with enough redundancy or error correction to take care of that. Is there some compounding effect I'm not thinking of? Or simply that it's hard to keep a species alive for a million years?

VLM 9y ago
Ouch I did some research and found "Mammals, for instance, have an average species "lifespan" from origination to extinction of about 1 million years, although some species persist for as long as 10 million years." Evolution: Library: The Current Mass Extinction - PBS www.pbs.org/wgbh/evolution/library/03/2/l_032_04.html So yeah, it would be a shame if space aliens visited and encoded a message to us in saber toothed tigers or trilobites. Also I have no religious hangups, I'm a direct descendant of the ancestor of amoebas or whatever exactly, and if the message were encoded into whatever makes an amoeba an amoeba, well, its definitely gone now. I think it would be hard to "infect" a species with stored data. They'd have to burn more protein to replicate that extra DNA material, which would provide a competitive advantage to any individual not carrying the data, so its not going to last long. Even thinking outside the box with a virus, wouldn't smaller ones have better odds of survival?
- Razengan 9y ago
  > I think it would be hard to "infect" a species with stored data. ... Yeah, but what about encoding a message/knowledge into the mating rituals of an species (or artificially-modified subspecies) from the start? [0] > if the message were encoded into whatever makes an amoeba an amoeba, well, its definitely gone now. Are we sure humans don't have any "useless" DNA? What about seemingly superfluous organs like the appendix? [0] https://news.ycombinator.com/item?id=14548854 https://news.ycombinator.com/item?id=14548854

Chaebixi 9y ago

But if you're just storing random data, like text, it probably offers little to no fitness benefit to the organism. So it could all be lost with a random deletion and no one but you would care. https://en.wikipedia.org/wiki/Deletion_(genetics) https://en.wikipedia.org/wiki/Deletion_(genetics)

Razengan 9y ago
> if you're just storing random data, like text, it probably offers little to no fitness benefit to the organism. Wow, this makes me even more interested in the idea of using mating rituals as a mechanism for very-long-term knowledge storage! Things like birdsongs, dances, and the aforementioned courtship art... If an individual can only reproduce if it correctly retransmits the "data", then the data will survive as long as the species does..no?