Several Indian companies from creative industries are beginning to experiment and deploy AI technologies. This is problematic for many reasons.
Richard Linklater’s 2014 film Boyhood took over a decade to make because the celebrated director wanted to capture the coming-of-age story in real time. Reflecting on the challenges of filming from 2002 to 2013, Linklater remarked that through the production, “fashion, cars — it all looks the same. The only thing that changes is technology. Laptops and smartphones. Everything else is stasis”. Linklater’s “anthropological takeaway” on technology echoes Amitav Ghosh’s 1992 coinage of “petrofiction”, which sought to direct attention towards the presence of oil in our stories and storytelling.
Much like oil, digital devices, online infrastructure, and virtual spaces are ubiquitous in contemporary storytelling. Today’s young audiences may find the drama involved in driving to the airport to declare one’s love a little bit “extra”, when a video call can do the job just as well, albeit less poignantly. Like fuel-injected automobiles, networked digital devices keep the stories of our times moving. Could a similar approach be used to explore how data, the so-called “new oil”, is shaping our fiction and cultures?
Digital devices and networks play a straightforward role in most narratives. They are represented as they are used in everyday life. However, scratch the surface, and more serious implications of datafication in transforming our stories and cultures become visible. One way to do this is to engage with the current hype around the promise of Artificial Intelligence (AI) in creative industries.
The knowledge/data that AI systems seek to “train” models already exists. It is a result of thousands of years of human beings telling stories, asking and answering questions that have helped us understand the world. Sometimes this data is obtained through consent and compensation, but most of the time it is simply appropriated. This amounts to centuries of collective knowledge and heritage being enclaved by a few for profit. It is more significant for those of us in the Global South as most of the capital and resource-intensive research in AI is being carried out by a handful of large global corporations.
On the other hand, low-paid workers in India, among other places, are cleaning and labelling the data to feed the machines. The anxieties of the non-alignment era around media imperialism have come alive again in the current structures of data imperialism.
Whether it is the fables of Panchatantra, the Hamzanama or the corpus of over a century of scripts of Indian cinemas — everything is grist for the AI mills. Much like the oil explorers drilling through the earth to extract oil and turn it into various marketable commodities, the commercial AI industry is drilling through our collective cultural heritage to crunch up and serve a bevy of services and products. Both industries are built on similar models — exploitation of resources that belong to all. To stretch the analogy further, the oil and the AI industries are also polluting. Other than their negative ecological impact, the latter has also begun to contribute significantly to what is being called the Internet of Garbage in the form of spam and spam-spewing bots. The surfeit of trolling and harassment, manipulated media and fake news, and rampant plagiarism is, ironically, facilitated by supposedly intelligent systems.
Several Indian companies from creative industries are beginning to experiment and deploy AI technologies. This is problematic for many reasons. First, as mentioned earlier, is the appropriation of collective cultural heritage without consent or compensation. The Writers Guild of America (WGA) went on a long strike last summer to demand, among other things, regulation of the use of AI in the film, television, and streaming industries. In the absence of specific laws, writers and artists seek to regulate it democratically. With the Screenwriters Association (SWA) endorsing the globally coordinated initiative for regulating the use of AI in screenwriting, the conversation in India is getting stronger.
Regulating AI technology is not the burden of storytellers alone. It is also inaccurate to see it as a tug-of-war between storytellers and producers. It should be a collective endeavour of everyone in the creative industries to push against its large-scale capture by the tech industry. There may be some short-term gains, but the long-term economic and cultural damage will be difficult to repair. Indian cinemas, unlike many modern storytelling forms, took root in the country simultaneously with the rest of the world and created their own unique identities and enduring appeal. Our heritage must not be allowed to be stripped for parts.
Second, there are strong reasons to limit the use of AI in creative industries as it has implications for sustainability. Economic, social, and cultural well-being must be prioritised in the face of massive investments and the breakneck speed of development related to AI. It needs democratic regulation — a particular challenge for India, which has a more informal creative economy. AI in creative industries can be useful as long as it does not seek to supplant human labour and creativity significantly. Even then, acknowledgement and fair compensation for input data is non-negotiable.
Third, such use of AI in creative sectors is culturally limiting. Serving up pieces of past stories in various permutations and combinations may lead to a stale and impoverished culture. It is not future-oriented; worse, it may perpetuate the biases of the past.
New ideas replenish and enrich cultures. Human creativity will forever push the boundaries of our shared cultural heritage through new stories that are sensitive and responsive to time and context. It is also freakishly renewable.
The writer is a Mumbai-based media practitioner
