In Part 5 of our Convergence Or Collision? AI and Content Marketing series, we’ll answer the question: what’s compelling platform adoption?

Why are we using AI for content creation?

It’s pretty simple, really. Using AI helps save time and money, two factors any organization values!

This is particularly true when it comes to analyzing and reporting on large volumes of data: AI is faster than humans. Technological advances have now brought us to the point where some AI content is virtually indistinguishable from human-produced content.

Additionally, incorporating AI allows publications to expand their scope and reach.

So, how did we get here?

Speed Racer: AI zooms ahead

90% of the world’s data in 2017 had been created over the previous two years, according to independent research firm SINTEF. That’s a lot of data.

As Automated Insights points out, this data explosion means we need more efficient ways to analyze increasingly large sets of data. Content automation can do this far better and faster than humans can.

This is especially true for a large volume of data. We’re talking hundreds, thousands, or even millions of data points, like:

To this extent, it’s important to note that the goal of AI content creation is not to replace journalists or content creators. Instead, the future of AI content creation is about making the roles of content creators easier.

One of the best examples to see this new tango between human creators and AI content creation programs is in sports coverage.

Sports and AI content creation

Let’s take a closer look at sports content creation and AI. In the past, the boxscore would be derived from a few dozen data points, which provided the best way to cover the game.

Today’s a whole new ball game (pardon the pun) in terms of data analysis. Live camera and sensory data is captured in nearly all professional-level sports. This commonly clocks in at 100,000s or even over a million data points per game. Having a team of analysts or journalists digest this much data in near-time or real time? Forget about it.

Instead, sports data analysis is the perfect arena to leverage automated content. AI can produce the insights, generate an automated report, and send to the journalist. Meanwhile, you can grab your favorite jersey and enjoy the game.

Changing the game: Automated Insights

In 2013, Automated Insights produced 300 million pieces of content. That’s more than all other major media companies combined. Jump to 2016, and the same company generated over 1.5 billion pieces of content.

That’s a pretty big leap, and the credit goes to Wordsmith, a natural language generation system. Wordsmith uses scads of data and quantitative analysis, couples that with writing rules, and produces hundreds of millions of narratives each year.

The question then arises: how does Wordsmith manage to produce that volume of content without sounding repetitive? Essentially, the same way a human writer would. Wordsmith varies story structure, phraseology and incorporates historical anecdotes to generate diverse stories.

In fact, a professor conducted a study comparing an NFL recap generated by Wordsmith to a recap written by a human. The result? Half of subjects found the pieces very difficult to tell apart.

Despite that surprising stat, let’s not forget that content generated by AI software is often described as boring and overly descriptive. For reference, check out the study Enter the Robot Journalist,

However, for content pieces that require the consistent analysis and reporting of large quantities of data, content automation can be a supremely useful tool.

Two, four, six, eight, it’s high time you automate!

The case to automate is strong. Think of financial advisors: would you want your accountant to add up numbers by hand, using a paper and pencil?

Content automation is like a calculator: it can eliminate or reduce the time spent on tedious tasks that don’t require human intellect or creativity.

This is especially true if the data fits neatly into the NLG framework (i.e. if we input certain data into the framework, it creates the prose). Think of content like weather reports or police reports. There’s really no reason we shouldn’t take such tasks off human hands.

AI makes more time for creativity

“The AP [Associated Press] estimates that the automated stories have freed up 20% of the time its journalists spent on earnings reports as well as allowed it to cover additional companies that it didn’t have the capacity to report on before.” – Nieman Lab,

Beyond reducing the need for human writers to churn out rote pieces of content, automation also saves writers time. This gives writers the chance to focus on more complex stories. To that end, incorporating AI changes the type of work and content that journalists pursue.

“Once you have a template, all you have to do is update the data to generate new articles…The work, then, becomes less about writing individual reports as it is maintaining certain templates and narratives, updating for sake of creative variety and editorial standards,” explains Ross Miller in The Verge.

Spread your wings (and your reach)

Finally, AI content creation offers one more big benefit writers. Automated content can help writers produce more pieces of content more affordably.

For example, automation can allow journalists to cover more events. In the case of sports journalism, AI means hundreds of games can be covered without the expense of sending human reporters to different locations to cover them. Automation can expand our reach greatly, while also being cost efficient.

In a similar vein, stocks and quarterly earnings by the Associated Press have ballooned to cover over 5,000 companies in a quarter. Before they partnered with Automated Insights, this number was roughly 400 companies. This increase, according to a study by Stanford University and the University of Washington, actually magnified trading volume and liquidity over time.

Bottom line? AI content creation can free up human writers to do what they do best: create well-researched, nuanced, and highly creative content.

In our next post, we’ll explore AI developer profiles.

Spread the word: