AGA 57 Put Its Finger on a Caribbean Media Problem: Getting Our Stories Out

There is no shortage of Caribbean stories. Spend a morning looking through the region’s newspapers, radio sites and independent publications and the problem becomes fairly obvious. Plenty is being produced. Much less of it travels.

A political story published in Guyana may be important in Trinidad. A Dominican tourism story could interest readers in Jamaica or Barbados. Reporting about migration, energy, climate or crime frequently crosses borders in substance even when the article itself doesn’t.

That old problem found its way into the Caribbean Broadcasting Union’s 57th Annual General Assembly in Guyana this week. Artificial intelligence was inevitably part of the conversation, but underneath the AI discussion was a rather more practical question: how does Caribbean media get more of its own material in front of people outside its immediate market?

Guyana’s President Irfaan Ali put the matter quite plainly. Caribbean media, he argued, already have content. Technology should be used to distribute it farther and compete for audiences beyond the region.

There is an appealing simplicity to that argument, although distribution has never been quite as easy as the technology industry likes to suggest. Putting something on the internet technically gives it a global audience. So does leaving a book on a park bench. The difficult part is getting the right people to find it, trust it and want to read it.

Anthony Greene, President of the Caribbean Broadcasting Union, was noticeably cautious about treating technology as the answer to everything. He spoke about AI improving efficiency, audience reach and potentially revenue, while keeping editorial judgment, accuracy and trust where they belong—with people.

That caution is useful. Media have been promised technological salvation before.

AI certainly changes what is possible. Whether everything that is possible is also desirable is another question.

Erwin Jay Saunders of the Turks and Caicos Islands raised one application at AGA 57 that deserves more attention: translation. AI can now help Caribbean content cross language barriers at a speed and cost that would have been difficult for smaller publishers to contemplate only a few years ago.

For this region, that is no minor matter.

The Caribbean is geographically compact and linguistically fragmented. English, Spanish, French, Dutch and several Creole languages and varieties coexist across markets that are already small. An article can be highly relevant on the next island and still never get there in a form that people will read.

Translation helps, but anyone who has worked between Caribbean markets knows that translation isn’t enough. Readers in Kingston are not readers in Santo Domingo with the language switched. Neither are readers in Georgetown interchangeable with those in Willemstad or Miami.

References need explaining. Measurements, currencies and terminology differ. Headlines that work in one place fall flat somewhere else. Sometimes the whole emphasis of a story needs adjusting.

Editors have always done this sort of thing. AI merely makes some of the mechanical work faster.

The interesting question is what happens if that technology is placed between publications rather than used simply to manufacture more content inside each newsroom.

Could a publication in one Caribbean country make a story available to another, allow the receiving editor to adapt it for local readers and still preserve who wrote it and where it came from?

We should declare an interest here. That is very close to what we are trying to do with CaribPubsPro.

CaribPubsPro was being developed before AGA 57 and neither the CBU nor any of the speakers at the assembly has endorsed it. We are building a publisher-to-publisher system through which a publication can make a story available to another publication. The receiving side decides whether it wants it. AI can assist with translation, localization and adaptation, but an editor decides what, if anything, gets published.

At least, that is the theory.

The more useful test is what happens when real editors start using it. Perhaps they will find parts of the process unnecessarily complicated. Perhaps publishers will be reluctant to make good stories available to potential competitors. Perhaps attribution alone will not provide enough incentive. Those are precisely the things that cannot be discovered from a diagram or a business plan.

We are beginning those tests now.

There is a financial reason why the experiment is worth attempting. Caribbean publications generally operate in small advertising markets while competing online against organizations with vastly larger audiences and resources. The internet removed geographical limits on distribution, but it did not remove the economics of scale.

Regional cooperation might help. It might also prove harder in practice than it sounds from a conference podium.

News organizations compete. Editors are protective of their material. Publications have different standards and political cultures. Some will happily receive stories while being considerably less enthusiastic about sharing their own. Technology doesn’t make any of those tensions disappear.

What it can do is remove some of the friction.

That may ultimately be AI’s more useful role in Caribbean media—not writing thousands more articles, but helping existing stories move between languages, publications and audiences without stripping away their origin in the process.

AGA 57 did not produce a grand solution to the distribution problem. Probably just as well. Caribbean media have seen enough grand solutions.

But several speakers appear to agree on something more basic: the region produces stories that ought to travel farther than they do.

Now somebody has to work out how to make that happen.