Comments on: Shakespeare’s Sonnets and MLK’s Speech Stored in DNA Speck

By: Telanis

Telanis — Fri, 22 Feb 2013 22:48:47 +0000

@Thomas, DNA is not a “read-once medium”. They may have used a read-once technique for this experiment for all I know, but that says nothing about DNA itself.

By: Michael Carroll

Michael Carroll — Mon, 04 Feb 2013 04:37:57 +0000

But what about the copyright? Did they perhaps open themselves up for law suits:
http://nucambiguous.wordpress.com/2013/01/29/is-that-fair-use/

By: James Rowlinson

James Rowlinson — Wed, 30 Jan 2013 13:24:59 +0000

@Jeremy Excellent point with regards to the Cartesian circle of obsoleteness. However I am sure it is possible to also encode on the DNA how to read these formats or how they work, to enable someone to read them in the future.

My knowledge of computers is not the best but is the JPEG and PDF not converted into binary first before this process is started. So they are universally readable, never going obsolete?

By: Robert Elliott

Robert Elliott — Tue, 29 Jan 2013 09:21:24 +0000

@Thomas Weigel – further to @Jeremy ‘s point about copying and reading DNA, you make the assumption that DNA sequencing will always be a destructive process. Technological improvements may infinitely improve the way DNA is read and, indeed, written.

By: Jeremy

Jeremy — Sun, 27 Jan 2013 08:00:49 +0000

@Doug Jones, DNA storage does nothing to combat the data format obsoleteness problem you had with your word processor files. Indeed, the sample mentioned in the article encoded an MP3 file and that format could easily become obsolete over the next few centuries (if not the next few decades).

@Thomas Weigel, your characterization of DNA as read-once is way off the mark. To begin with, the initial “write” of DNA created about 12 million copies. After that, one of the key evolutionary features of DNA is that it is super-easy to copy. Copying the existing DNA (referred to as “amplifying”) is generally the first step in reading (“sequencing”) it, so in practice every time you read it you end up with more copies than you started with.

As for the economic arguments, I agree they are nearly worthless. The future cost of data storage and retrieval as well as the future cost of DNA synthesis and sequencing are only going to go down, but how fast and by how much is impossible to know and it seems unreasonable to assume the ratio between the cost of writing DNA and the cost of transferring tape archives is going to be even close to constant over the next few decades.

The main technology-proofing and cost saving argument comes from the fact that if we store DNA in a cave we can be pretty sure that we will be able to read it in 10,000 years, whereas to be sure of the future readability of man-made digital storage formats we have to regularly convert them as technology changes.

By: Thomas Weigel

Thomas Weigel — Thu, 24 Jan 2013 18:57:02 +0000

Doug Jones: “the problem with your figures is your assumption of the cost of new tech.”

Longterm data storage today uses tape. Tape is proven and cheap. It is not “new tech.” It’s several decades old!

One does not use “new tech” for anything one wants to keep around for 5,000 years.

DVDs are just barely tolerable today. They’re still a bit too new – as late as 2008, studies indicated that they had worse longevity and data integrity than CDs, which were themselves never a good fit.

I picked DVDs because they are (a) more expensive, (b) barely tolerable, and (c) not quite as innately superior in every possible way to DNA as a storage medium. Tape is cheaper, faster, better for reads, easier to upgrade from, and proven by decades of use.

I was making a point: the worst barely tolerable choice I could think of was still better than DNA.

Using DVDs a decade ago would have been data murder. But even with that $505 figure? DVDs are cheaper than DNA storage over a 5,000 year span. Even if you spend $505 per year, instead of per decade.

Doug Jones: “What the DNA technology seems to be most useful for is long-term archival copies, rather than “read all the time” materials.”

You make this statement, but I cannot understand why. It isn’t cost-effective or trustworthy at any human timescale. And it’s read-once, which is practically data murder.

By: Berta

Berta — Thu, 24 Jan 2013 17:10:43 +0000

Very cool. Sounds like something one would read about in cyberpunk novel.

By: Niko

Niko — Thu, 24 Jan 2013 10:10:29 +0000

The initial cost of writing to DNA is probably more expensive than other established technologies for the foreseeable future but maybe the process of making copies with PCR is more efficient and cheaper?

By: Doug Jones

Doug Jones — Thu, 24 Jan 2013 04:28:54 +0000

- Thomas Weigel, the problem with your figures is your assumption of the cost of new tech. A little over 10 years ago, the cost of a (then) new DVD recordable drive was $500. A recordable blank DVD was $5. So migrating from the CD’s I’d put something on not quite 10 years earlier – also at a cost about the same to acquire at the time – would have been around $505 for 4 GB. That assumes that the format for the CD’s didn’t change (as they did), that they were still readable, and that I could read the DVD’s when their time was up to write them to something else. That’s also assuming that the files are in a format that’s still readable. I have old procedural word processing files which, if I hadn’t translated formats, would be “unreadable” now, in less than 20 years. What the DNA technology seems to be most useful for is long-term archival copies, rather than “read all the time” materials.

By: Tom Jay

Tom Jay — Wed, 23 Jan 2013 23:13:50 +0000

I do not think that money comes matters at all. Seems to me as to trying to compare costs of communication in Marconi’s and Tesla’s time with today’s TV or fibre optics. That the DNA may be encoded. Remember, do not worry about money, worry about success, money will take care about itself.