the cloud
Government
Facebook
Google
Data
Data Brokers
stored in a cloud and can never be forgotten. In addition, often this data is generated
and governed by private entities. For example, Facebook governs a lot of our social
interactions on their platform and keeps data about us, Google gathers everything
that people do with its search engine, and Twitter keeps score of all our interactions
via Tweets and others trace our love life (OkCupid, Tinder), our communication
(Gmail, Twitter, WhatsApp) and our entertainment (Netflix). This data is to some
extent owned by these companies, and whereas a long time ago interactions were
physical, and no trace was kept, these modern platforms are aimed at gathering as
much data as possible of all our interactions, and aimed at retrieval of that data
(of all users combined) for purposes such as profit and surveillance.
Despite the focus on data, it is only a consumable for the entities that really change
our world: algorithms. Algorithms are computer programs that autonomously utilize
data in order to do something. This can be sorting names in a database, computing
navigation instructions, or also organizing Facebook's news feed. The term
algorithm29 stands for any finite procedure/recipe, with well-defined instructions and
which is effective in solving a problem. Algorithmization is the phenomenon
where increasing numbers of tasks in society are carried out by intelligent
algorithms. The field studying and creating such algorithms is AI30 (McCarthy,
2007; Nilsson, 2010) which is seeing a recent explosion of advances, including
breakthrough technologies such as reinforcement learning (Wiering and van
Otterlo, 2012) and deep learning. AI's core is coming up with intelligent systems that
in some way exhibit observable behavior for which some form of intelligence is
required. Lately the focus is on adaptive AI, or machine learning (Domingos, 2015),
which ranges from baking cookies31 to driving autonomous cars by learning from
popular computer games.32 AI is rapidly becoming the driver for innovation (Stone
et al., 2016).
The transformation into a digital society can thus be characterized by the two
interrelating developments: digitalization, which turns once-physical interactions
into digital data, and algorithmization, which amounts to increasing analysis and
utilization of that data by algorithms. The transformation's impact on archives
(and libraries) is potentially huge. When it comes to digitalization, archives (and
libraries) are in transformation. Collections are constantly being digitalized to
provide wider public access to information, for example through the American
project Digital Public Library of America (DPLA33) and the European counterpart
Europeana.34 They unlock massive amounts of archival data such as books,
photographs and various documents. Initiatives such as the Google Books project
are similar in terms of technology, but have different goals. Google Books35 has a
long history of battles36 37 38 between a tech giant wanting to unlock all books
written by mankind, for everyone, and author organizations that think that Google
does not have the right to do that in this way. The ethical issues of access here are
severe, since Google may want to push the idea being a universal library but many
think this role should not be pursued by a commercial entity.
In general, libraries and archives (to some extent) have always struggled with their
exact role, especially in the transformation to our digital age, with the novel aspects
of born-digital records and books, and with the loss of being an information
providing monopolist (Licklider, 1965; Herring, 2014, see also Anderson, 2011,
p. 212) in the age of Google. Both Kallberg (2012) and Clement (2013) have
investigated how the archival profession changes in our digital age, and how
archivists think about that transformation. Paulus (2011) shows that the lifecycle
of information of archives and libraries changes, and that, for example, a
transformation is happening in which libraries may return to an ancient and
medieval model of the library or archive as a site of both production and
preservation. Cox (2007): "At last, archives have a real opportunity to abandon the
role of gatekeeper and invite user participation, interaction, and knowledge-
sharing." He continues: "What would happen if we could engage our users in
defining and describing archival content and in communicating it to others? Is it
possible that the analog archives tradition can learn from the movement of social
media and social design? Some of the opportunities include diminishing the role of
the archivist as gatekeeper, promoting participation and collaboration among users,
and enriching the archives itself by tapping into the specialized and diverse
knowledge of researchers".
The future of archives and libraries has many parallels with the development of
information technology such as the internet. Noh (2015) describes several stages
leading up to "library 4.0", which is where "technology will become one with users'
lives" and which also features 3D printing, big data, cloud computing and augmented
reality. One can also digitalize interactions that were purely physical until very
recently, for example using photocopiers39 and (personal) cameras (Cox, 2007).
archives in liquid times
272
martijn van otterlo from intended archivists to intentional algivists.
ethical codes for humans and machines in the archives
29 https://en.wikipedia.org/wiki/Algorithm
30 Science, special issue on how A.I. is transforming science http://science.sciencemag.org/content/3 57/6346/
31 http://www.wired.co.uk/article/google-vizier-black-box-optimisation-machine-learning-cookies
32 https://www.technologyreview.com/s/602317/self-driving-cars-can-learn-a-lot-by-playing-grand-theft-
auto/
33 http://dp.la
34 http://www.europeana.eu
35 http://books.google.com
36 https://www.wired.com/2017/04/how-google-book-search-got-lost/
37 https://www.theatlantic.com/technology/archive/2017/04/the-tragedy-of-google-books/523320/
38 See https://www.theatlantic.com/technology/archive/2015/10/fair-use-transformative-leval-google-
books/411058/ and https://www.wired.com/2017/04/how-google-book-search-got-lost/
39 This also connects back to Eco's "restrictions" described earlier on being able to photocopy in a hostile
library, but also to the ethical challenges concerning fairness when photocopying costs money.
273