With many media companies across the world just dipping into the world of Artificial Intelligence, it seems like a new and novel concept. But digging into the history of AI shows it’s actually been around for decades. 

During the recent INMA Latin American News Media Summit, participants heard from Diego Paramo Atalaya, an entrepreneur and founder of PulseUp, a company in Colombia that teaches other companies how to integrate AI into their business. Paramo Atalaya mentioned how Alan Turing began talking about AI back in the 1930s and how it could affect the outside world. Paramo Atalaya took it back to basics by defining AI as he sees it today.

“It's the manifestation of intelligence, very similar to human intelligence, but outside of a human,” he said.

A big difference, he mentioned, is comparing the capacity of a machine versus the capacity of the human brain and what that could mean when the two come together.

“We have the capacity to remember, we have memories,” Paramo Atalaya said. “And we have the capacity of memorising things. We have the capacity of having a perception of the world around us. So if I am in danger, I have the capacity of perceiving that. And we also have the capacity to anticipate. Machines have an enormous capacity of replicating and some even have the ability to anticipate future behaviours. They have an enormous capacity for remembering and memorising.”

4 ways this information applies to journalists and media companies

Paramo Atalaya walked through the ways this information is relevant to the media industry.

1. Interpreting human language, voice and feeling: The technology can analyse and summarise and understand voice. This part of AI is getting bette each day and really is transforming the media industry. 

2. GPTs: GPT, or generative pre-trained transformers, doesn’t just interpret but can also generate language. They can generate content for media industries in text and voice audio, as well as generate images.

3. Audience: AI is helping media companies understand audiences. It can use first-party data to cluster the information to include components that help capture and analyse data.

4. Advertising: Technology changes and evolution is giving media companies new and more exciting ways to entice their advertisers and they are able to use AI to help keep costs low.

Paramo Atalaya took a deeper dive into the use of AI when it comes to generating and analysing data.

“We have so much data, but do we actually use it correctly or is it even used in an efficient way in our organisations? This is something that AI data is already transforming and changing: media attention and the attention of our readers and our audience, everyone is competing for this.”

With competing priorities, platforms, and shortening attention spans, Paramo Atalaya acknowledges how much harder it is to retain the interest of audiences. His company helps other businesses by asking an organisation where they are getting their data. He asks how the audience is behaving and what they are looking at. Is there data coming from social media? Is the business holding events? This helps Paramo Atalaya decipher where the business wants to make an impact and if their database will allow them to achieve their goals.

There are also physical ways to capture data.

“How many QR codes are you using in your print media and in your digital media?” Paramo Atalaya asked. “How many captures are there? How many interactions are there based on QR codes? And where is this data? Are you using this data for something because the interaction happens here between the physical and the digital world? 

Another area Paramo Atalaya looks at when helping other businesses is efficiency of operation. CEOs of companies often ask him how their business can run more efficiently and what processes AI can take on, he said.

“We have to understand the users, our audience, how my audience is behaving, and where they're going. So how can I promote better engagement? How can I manage to have better clicking efficiency?”

Predictive analysis

One area identified for media companies is predictive analysis where a news event happens, the company covers the story, but after that, AI data can help predict what types of evergreen stories the company can generate that have high probabilities of the audience clicking on. An example would be an investigative piece related to a timely news story. AI can send the information to the editorial team to take a look at and see if there’s potential for a story in that area.

“This is based on content or the generation of content in an automatic way. So I don't defend that everything that is generated automatically be published,” Paramo Atalaya said. “I'm firmly convinced that human validation is very important, especially in the media because you have to deal with credibility.” 

Automatic content generation is suited for tasks like generating headlines or summaries. Transparency is also key, according to Paramo Atalaya. He encourages the use of disclaimers to tell the audience when pieces of news are generated by AI. 

When it comes to analysing data, Paramo Atalaya reminded media companies that the amount of data they use makes a big difference. With large amounts of relevant data, companies can find specific themes that can help companies improve their content. Again, Paramo Atalaya urges human validation and not just blindly following what AI is suggesting.

Having relevant data can also of course help with key metrics like how long a person stays on a Web page, bounce rate, what is being shared on social media, and what kind of engagement stories are getting. 

Personalisation is also key here.

“For each of the users that come to my page, I am supposed to be showing part of the content just for this person. Why? Because this will improve their experience,” Paramo Atalaya said. “There is going to be a higher probability of engagement. If you show your readers the articles that show a higher probability of clicking or reading so that you're going to click on it, you're going to spend more time there. So this generates better engagement.”

AI learning systems

What is also important and exciting about how AI plays a role in all of these metrics that are crucial to media companies is AIs are learning systems. They are consistently getting better, faster, and more efficient. 

“When we train in AI, we have to train in what's right and what’s wrong. And we have to take this into consideration because if I am good in only training in what is good and what is right, it's going to learn better,” Paramo Atalaya said. “This is a virtuous cycle in U.S. organisations. You’re going to become better and better in understanding the readers and also the commercial departments will become better too.”

When it comes to advertising optimisation and monetisation, Paramo Atalaya says like it or not, media companies are constantly being compared to metrics provided by giants like Google and Facebook. What he suggests is using AI to create metrics that make it easy for advertisers to compare apples to apples:

“So I’m not so far from Google or Facebook because look, I carried out a very deep analysis with AI and I am delivering now the same metrics.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .