Defend Truth


Will artificial intelligence herald publishing’s winter of our discontent?


Riyaadh Ebrahim is a social investment specialist at Tshikululu Social Investments.

What is the future of the publishing industry as we race towards a time when artificial intelligence starts replacing many of the current human functions? And what does this mean for social investment, particularly in literacy?

“We are standing on the beach, watching a fast-approaching tsunami, and all that we have around us are a few pieces of driftwood. We can continue trying to build a shelter from the driftwood in the few seconds that we have remaining, or we can figure out how we are going to use the driftwood to ride the ineluctable wave.”

I thought of this analogy when I was at a recent meeting held between some of the key players in the book production industry in South Africa. I heard many complaints of how “various philanthropists are bypassing the official library chain in South Africa by establishing libraries”, “the number of books that were being registered on an official registry are lower than previous years” and so on.

In many ways, I could picture myself sitting in a meeting with key players in the music industry about 12 years ago, around the same time that the likes of Spotify were gaining traction, and they would have been having similar discussions focused on the future of music. But ultimately, all of those discussions would have been underscored by the fact that the business model of the music industry was about to be significantly disrupted at its very core.

In South Africa today, there is strong political will and public sector interest combined with a significant appetite from the private sector to invest in literacy. This is largely as a result of a much greater understanding of literacy as a systemic lever vital to addressing many educational challenges that the country faces. From an investment perspective, there are several issues that relate to this changing industry that need to be better understood as we dive head-first into the literacy waters.

Protected IP?

First, there is the issue of copyright and the changing face of how society engages with protected content. Many current business models in the literacy space rely heavily on protected intellectual property (IP) and this is becoming increasingly unsustainable. If the music and film industries are anything to go by, then subscription-based, shared-service models might be the way to go; having a tremendous number of users subscribed to content services and consuming content as and when they require. Ironically, this is similar to the traditional physical library model (libraries are increasingly being made redundant, but that is related to many other externalities and not necessarily to the usage model).

Looking into the future, there could be digital “Netflix” styled libraries that will host enormous amounts of content that can be accessed by consumers.

There are already a number of such platforms that are available, but they do not host much “premium” content as much IP is still being held and sold at a premium by various gatekeepers.

Technology and AI

Technology is developing at breakneck speed and without strategic visioning we may well find ourselves investing in outdated business models that are set to fail through disruption. Here, I am not referring to technologies of distribution and consumption such as devices (a very complex issue that continues to be a barrier – which will shrink over time – further entrenching inequalities). Instead, I am referring to the technologies of content production. The book production chain still consists of a mind-bogglingly large number of players. There are authors of the content, editors and proof-readers, various stylists and and and… usually all co-ordinated through well-resourced publishing houses and making up a very heavy supply chain.

Artificial intelligence (AI) poses the greatest threat to this chain; we’re rapidly approaching a world where AI is able to directly link the producer to the consumer. At present, AI can edit text and do stylistic edits as the author produces content, including basic graphics and cross-referencing, and so on. In the near future, AI will be able to instantaneously translate and localise content. Also, AI will be able to summarise content and decipher which distribution conduits would be best suited as well as develop basic marketing strategies.

In the slightly more distant future, AI will, combined with big data, be able to do everything that a publishing house does, including end-to-end marketing, comprehensive market analysis (including running focus groups and full market testing), integrated pricing structures and will essentially cost “next-to-nothing” thus drastically reducing the costing model of literature (AI might even be able to contract on behalf of authors with the AI bots of film production studios to secure exclusive movie rights for some books that test well through market analysis, all within a tenth of a second of the author tapping – or even thinking – “submit” as they finish typing the last word of their book). 

Writer development

One of the few jobs that should remain through this approaching tsunami is the work of the creatives. At least it would seem that way at present, where AI is capable of replicating what humans do but still largely incapable of conceptualising new ideas. In a century from now, we may look back and see that “innovation” as an activity destroyed all other forms of employment except its own; a certain humanistic “x-factor” that simply cannot be replicated.

Included in the preservation of creative jobs will be authors, but the way that they produce content may be very different from today. Already existing “speech-to-text” and grammar editing technologies combined with almost-ready thesaurus and writer assistant technologies (including potential fact-checking and cross-referencing) will drastically change the manner in which people are able to produce content. With a strong writer-assistant bot by one’s side, what will writer development look like in the future? There may be a strong shift away from traditional approaches toward a greater emphasis on the “human” elements in the creation of stories.

What does this all mean for social investors?

For social investors interested in literacy, it is important that we are cognisant of developments in the space – the real, the potential and even the wildly imaginative. On the one hand, investing heavily in any model that can be easily made redundant can be a very costly mistake and on the other hand, if we are looking to further develop the literacy space, we should aim to be the pioneers of the future.

For example, in the South African context when we are looking at disrupting a system of literature production that is controlled by certain vested interests, it is senseless (and very expensive) to simply replicate the same systems with different ownership. We need to re-engineer the way it is done to match the needs of the future. We need to climb into our time machines and travel through the next few decades so that we can help pioneer a future that is more prepared to tackle inequality and mitigate against some of its negatively reinforcing effects.

Unfortunately, we do not have time machines and so we need to settle for the next best thing: strong and diverse teams of critical thinkers who can engage with current trends, from multiple angles, to develop a future analysis that we can use in our planning. This is needed not only in the literacy sector but also more broadly in the entire developmental sector. DM


Please peer review 3 community comments before your comment can be posted