Sun, May 23, 2010
I think you misunderstood it all. the string "immortal since XXXX" refers to the fact that, after becoming a SC member, you cannot die anymore. and that I know for sure, as after joining this website I was hit by a Bus and got shot 3 times, with no effect : )
Sun, May 23, 2010
Thats quite good to know, but I don't expect to make it to the singularity, ( and Kurzweil sure isn't going to make it... but don't let him catch on ).
Mon, May 31, 2010
shot 3 times?
Fri, Jun 4, 2010
I suspect in another 100 years the amount of data taken up by the present day internet will fit into somebodies mobile phone. Someone will probably back it up, even if it's just by accident.
Sat, Jun 5, 2010
— Assuming 4x10^19 bits of data on the internet
Assuming a hopelessly optimistic storage strategy of a silicon lattice where the presence of an atom indicates a 1, absence 0, and assuming each silicon atom is 271*10^-12 meters across, and our data is stored in a cube, I get : ( (4*10^19)^(1/3) )*(271*10^-12) * 1000 = ~ 1mm^2
This should be taken as a reasonable good lower bound for the volume of a molecular storage mechanism for the internet. In practice, atomic scale data storage will be error prone and therefore require redundancy. Read hardware will add additional spatial overhead, even more if you want to be able to write to it. Furthermore, all current fabrication methods are 2D which severely limits the density of readable storage in a third dimension.
how about some more practical order of magnitude guesses :
Using the wiki figure of 1Tbit/in^2 as thh maximum theoretical density of perpendicular magnetic storage, (4 * (10^19)) / (10^12) = 40000000 in^2 = 6.37 acres of magnetic storage. Assuming you could stack this at 1 mm we're still looking at 26 m^3 for the storage volume ( 6812 US gallons, for more pointless unit conversion ). At what I suspect is an unrealistically optimistic stacking distance of 1µm, this still takes a cube 30cm on each edge. Thats small enough to keep in your office.
IBM has some recording tech that can do 10x10nm bumps. So, lets call that 100nm^2/bit. ( (4*10^19) bits ) * ( 100 nm^2 / bit ) = 4000m^2 = about 1 acre. This packs to 4m^3 at 1mm spacing. If you can pack layers at 1um ( optimistic ), I get a cube ~15cm on each edge. I could carry that in a backpack.
For DNA, using 2.6x2.6x0.33nm^2 per base ( two bits ) I get about 10mm^3. Now, THAT looks promising. Maybe High density DNA storage media are the future, though I'd hate to see the data retrieval times. Even something at 100, 1000, 10000 times bigger could still fit in your pocket.
I wonder if anyone is researching polymer data storage tech ( that is, the data is contained in the functional groups on a polymer backbone ) ? This would seem to be "where it is at" in terms of raw information density.
Sat, Jun 5, 2010
Actually, I think polymer storage can achieve a much higher density than my silicon-lattice lower bound guess ( so much for todays thought experiments ). I guess a more appropriate lower bound on molecular storage would consider that you can use more than one molecule, maybe form a metallic lattice with 8+ elements per lattice points so you can store multiple bits at each lattice point.
Consider the 21 amino acids at ~4.4 bits per residue in a protein polymer. I don't really know the volume of a single amino acid, lets just say its 1nm^3, so that gives you about a 2x2x2 mm cube. I bet there is some interesting way, using redundancy, and an amazing pile of bio-nanotech, to store and retrieve data in dissolved polymers. Perhaps this will achieve your "all the internet in your cell-phone" information density.
Perhaps we could use a polymer with even greater diversity of functional groups, perhaps storing up to 7 or 8 bits per nm^3. Then the entire internet could fit, several times over, in your cell-phone.
so, in conclusion : more funding for bio-nanotech !
Wed, Jun 9, 2010
a super high-density storage medium is proposed
Thu, Jun 17, 2010
Cheers for the info michaelerule.