Will artificial intelligence assist content editing?

The beauty of editing is the opportunity it offers to be at the center of the relationship between author and audience.

Editors taught me to understand the minds of authors—how they think, how they write, what they know, and how much of that translates to their computer screens.for software as a service The technology industry in the past, I can say, To some extent, I understand the writing; to some extent, I am the audience.

The core of the editor’s role is to strike a balance between what the author thinks and what the target audience wants to read, while ensuring writing quality.It is also vital to contribute to team goals and achieve our goals Service Level Agreement (SLA).

In order to ensure quality and complete tasks on time, it is crucial to establish efficient processes. Ask any editor, and they’ll likely say that an article is never truly “done”—the deadline determines its fate.

While artificial intelligence is getting all the hype for its promise of efficiency, I can’t stop thinking about its potential for editors. The nature of new technology is scary at first, but that’s how I thought of Facebook (um…Meta?) years ago, and look where it is now. So I decided to try AI-assisted editing.

I could say the rest is history, but I’d be lying to you. But it does have great promise, and I don’t say that on a whim, but after several rounds of trying to incorporate artificial intelligence into G2’s editing process.

And answer the biggest question ever – Will AI take my job? let’s see.

Artificial intelligence promises to transform content editing workflows

I manage G2’s editorial team. Under normal circumstances, my global editorial team writes and publishes a variety of articles written by in-house writers, freelancers, and external thought leaders. The odds were never in our favor in terms of efficiency, but we were a strong, talented team, so we persevered.

when Chat GPT After becoming everyone’s favorite new content creator, G2 launched a project to use artificial intelligence to “do 10% more” – a project to use artificial intelligence to improve daily productivity. This prompted me to challenge the team and explore whether we could use artificial intelligence to solve the fundamental problem of achieving efficiency goals.

Setting up our test run

My expectation at the time was to use AI to give my team some time so they could start thinking about the writer’s voice and the needs of our audience, rather than just doing copy editing and customary line edits to fit our brand voice.

To do this, we first performed a test run.

Experiment first round

No agenda-driven experiment can run smoothly without proper methodology. Taking into account the above efficiency challenges, we established a framework for the first round of experiments.

Each editor on the team is assigned a specific AI tool and a number of articles in the editorial process. Ask them to use AI tools to edit these articles and document their learning.

The process is subjective. It depends on how editors want to use the AI ​​tools and for which use cases.

Their only goal is to answer a questionnaire and support their answers with screenshots of the tool, as well as their opinion or understanding of how the tool works.

The questionnaire is shown below:

So, what did we learn in the first round?

good aspect:

• Artificial intelligence helps improve the structural flow and logical progression of content, effectively contextualizing scattered ideas.
• It’s proven to be great at filling content gaps and expanding existing content, identifying ideas that may have been missed that can be incorporated into articles, and serving as a good source of inspiration when building sentences or when you’re stuck creatively.
• Artificial intelligence helps improve grammar and spelling accuracy.
• Artificial intelligence has the potential to save time during the content editing process, but this depends on the type and complexity of the content.
• Artificial intelligence generates eye-catching alternative title options and polished titles.
• Artificial intelligence efficiently edits content to comply with specific brand guidelines or editorial standards.

Bad place:

• Artificial intelligence sometimes misinterprets context, leading to incorrect recommendations.
• In some cases, AI-generated content is robotic or unnatural.
• Integrating artificial intelligence into existing workflows can be challenging.
• Reviewing and validating changes made by AI tools is time-consuming.
• Artificial intelligence is not well suited for contextual editing or long-form articles.
• Artificial intelligence seems more suited to writers than editors.
• Sometimes, artificial intelligence makes unnecessary changes to the context or destroys the originality of the writing.
• Even after specific prompts are entered, the AI ​​may rewrite the article rather than adjust the grammar or sentence structure.
• Certain editing tasks appear to be beyond the current capabilities of artificial intelligence.

Yes, these results are mixed!

While one editor said the AI ​​tool was helpful in establishing contextual gaps, another team member said it was not suitable for contextual editing.

But all in all, the results were definitely better than we expected.

Second round of experiment

For round two, we moved from a subjective process to a more objective process.

Editors are asked to use artificial intelligence at every step of the editing process—from the first round of style edits to developmental edits to final proofreading. As they did so, they were asked to note how long it took to complete a round of editing.

Here is an example of it:

We follow the same technique to analyze the working of artificial intelligence tools. This includes tips related to improving the introduction of the article, suggesting changes to sentence structure, and proofreading.

What did we learn in round two?

• Artificial intelligence doesn’t help reduce the time required for editing, but it helps reduce the workload associated with ideation.
• Artificial intelligence can help tailor audience personas and generate Ideal Customer Profile (ICP).
• Artificial intelligence has taken over in proofreading.
• Artificial intelligence helps identify unnecessary metaphors, wrong jokes, awkward and redundant uses in articles and helps change the language to fit the desired brand voice.
• Artificial intelligence can scan for sarcasm and help make text more constructive.
• Artificial intelligence can identify absolute words in text, such as “never”, “always”, etc., and suggest modifications to the sentence.

Compared to the first round, this round gave us more promising results, allowing us to conclude that artificial intelligence can help us in many ways.

But there’s still a long way to go.

Create use cases

After two rounds of experiments with two different methods, we now have a fairly clear idea of ​​what artificial intelligence can and cannot do. Despite its limitations, it’s becoming increasingly clear to me that artificial intelligence could be the savior to our efficiency challenges.

That means it’s time to dig deeper and conduct at least another round of experiments. But this time, we need a sharper methodological structure in order to identify specific use cases of artificial intelligence in editing.

Experiment round 3

The third round consists of creating a template with prompts for each use case that can be used in each sub-step of the editing process. Every tip we used in this phase was tried and tested in the first two rounds of experiments to provide the desired results.

The template also comes with a set of guidelines for editors, which are based on our experience. This guide includes which AI tools to use in certain use cases. For example, while Google’s Gemini can be used to find contextual gaps, ChatGPT will help craft better article introductions. This document will serve as a single set of guidelines (or source of truth) for the team.

Once the template is ready, editors review and execute the prompts using suggested AI tools. The final step is to share their thoughts on the above issues as a team and address any obstacles.

Here are some examples of it:

Editors can now measure gains or losses in efficiency simply by copying and pasting a standard set of prompts into an AI tool.

Moment of truth: Experimental results

Now it’s time for the big reveal.

However, it is important to realize that this level of efficiency improvement is unlikely to be achieved after one or two iterations. It may not work in all situations. We have encountered many situations where editors spent more time on artificial intelligence than manual editing.

But for now, let’s celebrate this victory.

focus

Artificial intelligence does a great job of finding gaps in writing. But every now and then it might find a gap that doesn’t exist. Often, artificial intelligence alters a writer’s voice so that the human voice sounds like a Skynet-era robot. Artificial intelligence will almost certainly start your article with “in an evolving business environment,” or worse, turn a phrase like “employee happiness” into “worker prosperity.”

The trick is to find what works for you. After working through several bad tips, we found the best one. The best use cases for artificial intelligence are the things you can do.

My final conclusion about using artificial intelligence in editing is that it does solve our efficiency challenges. But it may not always offer us the best path forward. Editors are required to always use their best judgment.

However, it does become a tool we didn’t know we needed.

With the help of AI, we can now confidently meet strict deadlines and adhere to our SLAs. Artificial intelligence can also help us “get out of trouble” faster. When editing, we often think about whether there are gaps in the article that we didn’t see, or whether there is a better way to structure the sentences.

Artificial intelligence helps speed up ideation, thereby increasing efficiency.

Based on our experience, the following AI content guide highlights the only use cases where we have found AI helpful for editing. But keep in mind that this depends on the type and complexity of the article.

Download your guide to AI content editing!

So, will artificial intelligence take my job?

Short answer: not yet. AI has no personality However.

While AI may be great at grammar checking and filling in structural gaps, it has no original ideas, which is the strength, the best, and the most important thing that we as humans can bring to the table.

Artificial intelligence cannot replace the human element of recognizing context, nuance, sarcasm, voice, and intent.

The good, the bad and the ugly of AI-assisted editing

When I first asked my team about the AI ​​experiment, I began to wonder whether I would make the use of AI mandatory or optional if the experiment was successful. Now that I’m here and we’re successful, I choose to leave this optional. In the creative process, freedom is crucial.

While we see the value that AI brings, integrating it into existing workflows is not without its challenges.

There is a huge learning curve with any new technology. Every editor has their own unique editing style. Incorporating tools into this workflow means destroying their creativity. So it will take some time to get used to the change. But once your experimentation comes to fruition, it’s hard to go back to a completely traditional editing process.

More importantly, what is needed is perseverance.

Even after round after round of failed experiments, continuing to persevere is not an easy task. Using AI as a tool to augment rather than replace their existing skills and a desire to do ten percent more, editorial teams are sure to find victory in AI in the end.

Looking for more information on how artificial intelligence is changing the world of content?Learn more about the future artificial intelligence writing.