As our team works with Wikipedia to help it identify its vulnerabilities and opportunities in a new time of digital transition, its contributors and extensive research flag Wikipedia’s positioning in our information systems. Simultaneously, these contributors’ concerns over AI tools are entirely comprehensible – extant media and research has regarded the sudden and rapid adoption of Large Language Model and other AI-adjacent technologies with more apprehension than optimism.
This apprehension is already feeding our creative engines (even as AI tools are being positioned to replace them). Adding to a fairly screaming time for cinema, rave reviews were reserved for a tense film highlighting a terrifying technological innovation that frays and reshapes our entire global order. No, not Barbie; spoiler alert, I’m talking about Mission Impossible: Dead Reckoning. Despite being more firmly science fiction than Tim Burton’s Mars Attacks, its central bogeyman will be depressingly familiar: the growing spectre of information disorder, entrenched and enshrined in algorithmically structured media ecosystems. A recent NYT piece delves into the real-life, and sometimes comically clunky, version of the big baddie in the digital person of ChatGPT, in the process demonstrating that the problems are more firmly rooted in the social sciences rather than the information sciences.
While nailing the zeitgeist of our AI fears, what the film – and its many analogues – misses is the extent to which the problems are rooted not (entirely) in our algorithmically structured digital media systems but rather in the social processes we use to identify, prioritise, and utilise salient and high-quality information. Where the public were once more inclined to look to cultural intermediaries like journalists and academics to separate a world of infinite data into reliable bite size chunks, our burgeoning social media sphere capitalises on stoking rage and sowing distrust, leaving many feeling no source of information can be trusted. NYT’s Jon Gertner’s investigation reinforces the importance of this role in his findings that the big data trawl on the internet has yielded poorer results than a smaller data set curated by Wikipedia’s expansive policies and army of volunteer editors, which has become a cultural intermediary in its own right. Alongside its essential role in Google’s knowledge graphs, the importance of Wikipedia in powering our fact machines underscores our rapidly growing need for cultural intermediaries that can intervene in the internet’s information disorder, with the better decisively triumphing over the internet’s ever-expansive more.
Rather than once again trying to re-create the Torment Nexus, research suggests that we should be focusing on those boring and banal systems of information curation as well as the means to hold both the curators and distributors accountable. There is no magic key to this machine.
Tim Koskie, CMT researcher
This is an article from our fortnightly Enewsletter of 28 July, read it in full here or subscribe!