Exe0.1 Audrey Samson

Memory execution / immortal data

Disclaimer the following is the early stages of a draft. Please critically comment on the argument structure, does it hold water, does it follow, refs missing, etc., rather than the syntax or spelling. Thank you!!

Abstract: The following text considers the issues that arise from the static storage model. By looking at mechanisms of death which occur in cells and taking human memory as an example of labile storage, static memory is posited to be a problematic model. A link is traced between cellular death mechanisms and memory transmission (mutations), and the new developments in synthetic DNA storage. As the near future nature of truly embodied data stored in DNA is thought of within the static storage model, I use this text to sketch the implications of this anachronism, and hope to inspire thoughts on possible mutable memory models.

Static storage is understood here as a model, in the first instance, because it ranges across several disciplines as a framework. In literature, static storage is a model in which memory is not altered and representational. This notion has been challenged namely by Thomas Wägenbaur (1998) and Kyle Pivetti (2015), accounting for amongst other factors, recent neuroscientific memory research. In the sciences, neuroplasticity has replaced the notion of the immutable memory. Memory changes every time it is recalled, transmission between neurons is highly mutable (Nader et al. 2000). In computer science, static may refer to storage, variable allocation, or objects (to name but a few). Static storage is defined as fixed, for example each object in the storage is located at a fixed address, or it is allocated at compile time, that is, before the program runs, or executes (Laplante 2001). Static usually refers to consistent data, data remains throughout runtime, or without refreshing, as in the case of static random-access memory (SRAM) (Laplante 2001). In computer science, the intact retention of data is a measure of success. Reliability is the carbon copy, "directly representational" (Pivetti 2015, 19). Hierarchy of memory was calculated with speed of access relative to the cost to the CPU, a negotiation of dynamism and stability. These dichotomies are slowly becoming deprecated. The development of the MeRAM chip for example, which can both compute and remember, blurs the CPU/storage boundaries, placing memory at the site of execution (Khalili; Wang 2015). That said, in all these cases, information or data changing due to outside factors (outside the program) would be considered a failure. If an archive is corrupted, or missing files, it normally causes great distress to its owner, as has too often been the case with data loss. Data recovery is big business. Notably, in domains such as bacterial growth, viral propagation or organism survival, adaptation is the saving factor. To mutate could mean survival in the face of crisis whereas stasis is certain death.

The concept of static memory, in comparison to human memory and cell 'mechanisms', is not labile and mutable. As mentioned previously, human memory is increasingly understood as dynamic. Forgetting, or erasure, is integral to its functioning (Hadziselimovic et al. 2014). On the cellular level, we find inbuilt death mechanisms such as apoptosis or programmed cell death (PCD), and necroptosis. Apoptosis is a form of organism regulation, cells which are no longer needed activate an "intracellular death program" (Alberts et al. 2002). The cell suicide program leads to "fragmentation of the DNA, shrinkage of the cytoplasm, membrane changes and cell death without lysis or damage to neighbouring cells." (Alberts et al. 2002). This phenomenon is not only normal, it is integral to organism survival. It is instrumental during organism metamorphosis (such as from tadpole to frog) or in the regulation of cell numbers. The unregulated reproduction of cells, or cell division, causes cancer, often resulting in death. In this case, the unmitigated propagation of data is the anomaly, the disease. Necroptosis is another programmed form of cell death characterised by a cellular explosion. It is crucial in combating viral infections for example. Research into this form of cell death has been stated to be a promising target for drug development, which could have applications such as cancer treatment (Vandenabeele et al. 2010). The recently increasing study of programmed cell death, like neuroscience memory research, shows that regulatory functions include death and erasure.[1]

Static storage can also 'die' through hardware destruction, or more usually through deprecation of soft/hardware. A hard drive's shelf life does not normally surpass 10 years.[2] Though, as previously mentioned, this is seen as a problem in need of solving. The importance of linking cellular memory and propagation with so called digital storage becomes pertinent with the advent of data storage in DNA. Research into writing data into DNA has such applications as DNA computation, and DNA data storage. One of the major impetuses for this research is the gargantuan spatial demands and high energy consumption that the vast amounts of information amassing in servers engender. The later is mainly fuelled by big data analytics and state surveillance. As the ensuing storage necessity increases rapidly, where will all this data be stored? Meanwhile, DNA data storage offers high-density information encoding, a non-deprecating format, and low maintenance properties. As much as 2.2 petabytes can be stored in one gram of DNA (normally the size of a football field), it can be stored at room temperature (for centuries), and as this 'format' has been around since life itself, the hardware will not foreseeably become obsolete (Goldman et al. 2013). The excitement about the potential for eternal data storage is feeding headlines about the preservation of knowledge. Immutable data longevity (human preservation?) is the stake (see figure 1).

Figure 1 – Longevity of current storage methods (Kazansky Lab)

As scientists and entrepreneurs ask how could we store data for millennia, even millions of years (Grass et al. 2015), there is little thought towards the implications of such hyperthymesia. Some heed has been paid to formulating the social, ethical, and legal issues relating to scientific data storage and DNA banking practices for biomedical research (Godard et al. 2003). Whistleblowers and hackers have warned against the institutional misuse of data analysis and surveillance. The 'right to forget' has also emerged in reaction to socio-political issues arising from persistent personal data. The bandied round solution is bigger security, better cryptography, and law enforcement vigilance towards leaky data security abusers (i.e. hackers, not Facebook, Google or NSA). Could this Sisyphean security breach repair cycle be flawed in itself? Are we asking the wrong questions? Neuroplasticity is accepted in the scientific community. Forgetting is integral to the proper functioning of human memory. Cells contain several modes programmed degradation. From the floppy disc to the hard drive, media storage inventions have systematically failed to eternally survive. What is it about static storage that sells? I propose that we should critically examine the model of static storage as an ideology to help problematise security/privacy issues.

DNA storage is bridging the gap between previously separate domains of technology and biology/health. Though some thought is being put into DNA contamination, complex implications or possibilities which could arise from synthetic DNA propagation are still unclear. Also, DNA is itself part of a cycle of life that involves duplication, mutation and suicide. What happens if we try to impose a static model upon a mutable format? Researchers at ETH have developed a sheath of silica glass to encapsulate stored DNA data, which could allegedly survive a million years (Grass et al. 2015). This literal example of impeded mutation begs the question, who does this contrived approach serve?

Data 'donation' is already being exchanged for medical services (ADD REF - cant find atm). The new hype is crowdsourcing the genome, exemplified by the DNA.LAND project.[3] Viral promotion based on gaming campaigns is promoted as the key to success. David Mittelman, geneticist and entrepreneur, explains that "people can be convinced to share their data if it is easy and fun".[4] The lucrative promise of DNA storage and synthetic biology has lured Silicon Valley investors.[5] While the bio-data merger is being negotiated under the auspices of health and knowledge development, it becomes crucial to ask who can profit from it? And what kind of control do the funding bodies have on the developments?

One can easily envision a possible future where the hottest commodity is erasure. Pay a modest fee to have your search history wiped. Pay a larger fee to have your nudies deleted. High security becomes the privilege to erase, governments and corporations operate under constantly erasing surveillance. Control data erasure and you control the future? I have previously posited that the read/write process could place memory at the site of execution, rather than storage (Samson, 2015). In this sense, DNA data storage is the embodiment of memory execution and encompasses all socio-political implications inherent to DNA synthesis. As mutation and death is a condition of life, I propose to consider possible mutable memory models. A potentially dynamic system built upon the materiality of DNA, that is to say, situated (influenced by surrounding agents), insecure, transparent, unpredictable and subject to untimely death.

NOTES

[1] Alexei Degterev and Junying Yuan identified necrostatins (compounds that block necrotic cell death) in 2005. See http://www.nature.com/nchembio/journal/v1/n2/full/nchembio711.html

[2] A study by Blackblaze states "22% of drives fail in their first four years". See https://www.backblaze.com/blog/how-long-do-disk-drives-last/

[3] See http://www.nature.com/news/scientists-hope-to-attract-millions-to-dna-land-1.18514

[4] Ibid.

[5] See http://www.nature.com/news/synthetic-biology-lures-silicon-valley-investors-1.18715

REFERENCES

Alberts B, Johnson A, Lewis J, et al. Programmed Cell Death (Apoptosis). Molecular Biology of the Cell. 4th edition. New York: Garland Science; 2002. [1]

Godard, B., Schmidtke, J., Cassiman, J-J., Aymé, S. 2003. 'Data storage and DNA banking for biomedical research: informed consent, confidentiality, quality issues, ownership, return of benefits. A professional perspective'. European journal of human genetics : EJHG 11 (Suppl 2): S88-122.

Goldman, Nick, Paul Bertone, Siyuan Chen, Christophe Dessimoz, Emily M. LeProust, Botond Sipos, and Ewan Birney. 2013. 'Towards Practical, High-Capacity, Low-Maintenance Information Storage In Synthesized DNA'. Nature 494 (7435): 77-80.

Grass R.N., Heckel R., Puddu M., Paunescu D., Stark W.J. 2015.'Robust Chemical Preservation of Digital Information on DNA in Silica with Error-Correcting Codes'. Angewandte Chemie International Edition, 54, 8, 2552,-2555, DOI: 10.1002/anie.201411378.

Hadziselimovic, N., Vukojevic, V., Peter, F., Milnik, A., Fastenrath, M., Fenyves, B. G.,… Stetak, A. (2014). “A plastic nervous system requires the ability not only to acquire and store but also to forget”. Cell 156.6 (2014): 1153–1166. Print.

Khalili, P., Wang K. L. The Computer Chip That Never Forgets. IEEE Spectrum (2015) [2]

Laplante, P. A. (2001). Dictionary of Computer Science, Engineering and Technology. Boca Raton: CRC Press LLC.

Nader, Karim; Schafe, Glenn E.; LeDoux, Joseph E. “Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval”. Nature 406 (2000): 722–726. Print.

Pivetti, K. (2015) Of Memory and Literary Form: Making the Early Modern English Nation. Lanham: University of Delaware Press & Rowman & Littlefield Publishing Group.

Samson, A. “Erasure, an attempt to surpass datafication”. APRJA journals/newspapers, Datafied Research Issue 4.1(2015).

Vandenabeele, P., Galluzzi, L., Vanden Berghe, T., Kroemer, G. (2010) 'Molecular mechanisms of necroptosis: an ordered cellular explosion'. Nature Reviews Molecular Cell Biology 11(10):700-714.

Wägenbaur, T. (1998). The Poetics of Memory. Tübingen: Stauffenburg (Stauffenburg Colloquium) (http://www.stauffenburg.de/asp/books.asp?id=140)