Finding our bearings in a mass of data



Ach, who doesn’t wonder from time to time what will remain?  Not just of our individual selves, but rather of what we value, not just of an individual life, but of an entire age with its struggles & beliefs.

Anyone with some interest in (art) history, knows how sheer material contingencies play a crucial role in the survival of cultural artefacts and texts. (1)  The choice of supporting material (stone or organic), a wet or a dry climate, quality of varnish, …  And on the immaterial side – how transient fame can be, how difficult to predict the continued consideration felt for a tradition, or the continued belief in a metaphysically sanctioned message. 

Now, what will still be understood of our digital age, say, 1000 years hence? What posterity will there be for our massive collective,  distributed effort to digitize everything, to put “everything” on line – from fleeting private expressions to commercial banalities  - from a garish dark web to edifying massive online courses,  to scientific networks … 

What will happen with the ever swelling mass of digital data stored on numerous servers? (2) Will all data that haven’t been accessed for some time, at some point be systematically deleted?  (just as even so-called perpetual cemetery plots are finite, if no one renews the lease). Will there be an artificial intelligence able to make sense of the gigantic mass of data? Will “history writing algorithms” assess historical relevance of data based on numbers of views and followers?  So that a make-up video tutorial on YouTube or a contagious 15 sec goofy dance on TikTok will be considered as emblematic for our age?   

In the early days of internet much was made of its encyclopedic scope  - it was compared to a universal library accessible to all, hyperlinking everything  and easily searchable through search machines.  But since no single human is able to digest it all, on the internet the individual mind is  easily outdone by "big data" crunching algorithms. Will artificial intelligence create some sort of new collective mind, a "mind of the hive"?  (Not a coincidence that there’s an AI company called Hivemind).  

How different the 21st C Internet of everything & everyone is from a library purposefully built by erudite and sensitive individuals.  How different the Internet is from the 20th C archetypical library: the Warburg Library;  made up of pictures  & books obsessively brought together by an individual reflecting mind,  continuously seeking meaning, establishing affinities & correspondences between historical images.

 venir en aide à l’historien d’art qui a perdu ses repères dans une masse inerte de données  […] leur donner un sens dans le contexte de l’histoire de la culture […] Warburg n’avait pas de méthode, mais il avait un message » (3)

«  que l’histoire de l’art importe encore non pas en tant qu’accumulation de faits, mais en tant que témoignage des souffrances et des triomphes de l’humanité » (4)

In order to find their bearings in the massively accumulated data, future (art) historians will need the help of algorithms with superior data-digestion capabilities. 
But how could we expect from algorithms to understand the meaning - the aesthetic, moral and emotional meaning -  of the data they crunch?  
How will they be able to convert these big data again into testimonials that at some point in the future might still appeal to organic specimens of humankind?



Artificial Notes


  1) some figures  & thoughts on book production and survival, from “CrossRoads – travelling through the Middle Ages” by Marco Mostert :  

“from the sixth through the eighth century some 67.000  manuscripts were produced in the Latin West […] From the entire eighth century less than 2.000 manuscripts (or fragments) have survived”
 The cost of making books was high so  the early-medieval manuscripts that have come down to us all represent considered decisions to make a copy of a text or to write a text […] That is why every manuscript book from this period is worth studying."

(picture from Wikipedia page with list of key works of Carolingian illumination


2) some figures & thoughts  on 21st C “content production” : “As of May 2019, more than 500 hours of video were uploaded to YouTube every minute. This equates to approximately 30,000 hours of newly uploaded content per hourwww.statista.com. “More than 95 million photos are uploaded to Instagram every day.” “More than 1 billion videos viewed on TikTok every day”  . 
The trend may be versus more ephemeral social media (such as snapchat, instagram stories)  with eg  programmed deletion of content after it has been viewed – that would help stop the exponential  growth in saved data. 

  3) quoted from the  French translation of Gombrich’s book “Aby Warburg – Une Biographie intellectuelle”.
Translated again into English by an algorithm – which doesn’t understand the meaning of language but is uncannily competent all the same : “to help the art historian who has lost his bearings in an inert mass of data […] give [the data] meaning in the context of the history of culture […]  Warburg had no method, but he had a message “ 

  4) “ that art history still matters not as an accumulation of facts, but as a testimony to the sufferings and triumphs of humanity (also translated by ab algorithm)


No comments: