How Buffer’s Content Team Uses AI

In this post in this article, in this article, the Buffer Content team discusses precisely how and why we employ AI for our projects.

There’s been a lot of talk concerning AI recently, especially since ChatGPT was made availableto the general public on November 20, 2022. Since then, we’ve witnessed an increase in ideas and tools related to the subject and the application of AI.

We loveAI within Buffer and have even launched our own Artificial Intelligence Assistant to assist our customers produce content faster and better. We are aware of its limitations, particularly in the area of creation superior to the majority of. We also recognize the (valid) concerns that a large number of marketers and creatives have regarding the effects AI’s impact on AI in the creation of their content.

Even though we employ AI for our work within the Buffer Content team (there only three of us and AI allows us to perform our work more efficiently) We wanted to know the ways and whatwe make use of AI to help us in the work we do.

Our past with AI

We’ve been experimenting with various AI tools for some time and it’s evident in our archive of content of the last year.

The very first article from Buffer on AI was published in May 2022. It addressed the question ” Should you be using AI for content creation?” Looking back about this piece is intriguing because our position back at the time and today is the same: let AI handle the smaller things, and you’ll be able to better manage the bigger creative task.

Our curiosity about AI has only increased since then. As ChatGPT became public, we wrote various articles on AI on our blog and we also launched our own AI Assistant.

The time spent with the various tools based on AI has helped us determine our boundaries more precisely. This means that there are no written guidelines regarding AI in our content. There was a pattern that emerged during two events, which prompted us to consider how we can communicate our position regarding AI?

The second was our most recent (now closed) recruitment drive for Content Writers. The article provides more details the subject, but it is interesting to note that a surprising amount of applications were developed by using AI. Understanding the standard AI-generated content was helpful in identifying the applications, but it was widespread enough to warrant highlighting.

The second event is currently running The third event is ongoing – our recently inaugurated Guest Posting Program. The purpose of it is to open our Buffer blog to stories of knowledge by creators and small-business owners. We’re looking for writing that is personal and makes our readers feel as if they’ve learned something new. However, a lot of pitches we receive are created by AI with little or no editing. They all follow the same pattern we’ve seen in AI-generated responses, particularly those generated by ChatGPT.

Despite the rumours that AI is an imminent threat to people working in creative industries We disagree. AI is not able to replace the human mind or its ability for creating art which is deeply connected. It’s important to keep in mind that AI is simply a broad model of language learning. It draws inferences based on what you say to it, and uses its small database of information to execute what many have referred to as “spicy autocomplete”. It’s a device – but it’s only as effective depending on the user who is using it.

We needed to clarify our views since, while it’s great it’s not going to replace anyone of us anytime in the near future.

Our strategy for AI in content

These are the four assertions which cover 90% of our thinking about the use of AI for content.

• AI aids, but it doesn’t replace
• AI never completes drafts or paragraphs unless they are checked by humans
• AI content is constantly reviewed to ensure its accuracy.
• We don’t believe that AI will replace creativity in any way

Our Content team’s experiences working with AI

Buffer’s Content team comprises three people who manage the team’s blogs, social media, PR, and other content and communication-related needs. The size of our team in relation to our roles often mean that we are forced to decline certain projects or opportunities. We must prioritize and prioritize, but ultimately social and content are the top priorities of our team, which is why they should not be pushed aside.

Implementing AI tools has enabled us to gain more time to investigate these possibilities and also collaborate with other teams in a more efficient manner. For instance, the initial edits are made using Grammarly instead of one of us in order to detect the most obvious mistakes in punctuation and grammar. This lets us concentrate on more complex structure such as voice, tone, and editing. I utilize Otter frequently in recording my interviews. It connects to the calls I make on Zoom calls and takes rough notes, so I don’t have to begin recording and transcribe from scratch. It also provides the summary of every call.

Other than AI-heavy tools Some of our other tools contain AI integrated into them, such as

• Threads are used to communicate internally
• Notion that we employ to manage knowledge
• Canva is a tool to design blogs and social media posts graphics
• or even Buffer.

AI is a necessary element of our work And although we canchoose not to utilize AI tools, we’d be less efficient as a result.

What we use AI for

We employ AI to complete critical, time-consuming jobs that are not human-like.

• To proofread the writing and look for grammatical mistakes
• For inspiration to be able to unlock before you get into the research process
• For help with outline
• Templates for creating templates
• To come to interview topics
• To suggest title ideas
• The draft for the first draft of the social posts
• Repurpose existing content to be used on multiple channels
• In order to make our content more relevant to an audience that is hyper-specific

We’re not making use of AI for is

We don’t and will not ever employ AI for anything.

• Research is flawed because it’s incorrect most of the time.
• Draft your article fully or create full paragraphs within our drafts. We use AI only as a tool in writing and brainstorming.
• Make final content for your social media
• Create your own artwork completely from scratch. We currently only utilize real images from small-scale companies and creators who submit to themselves or on Unsplash. If AI is used in our images, it will only be a part of our design process.

We’ve seen real ways to use AI

The Marketing and Content teams see AI as a partner and we’ve had projects as evidence. For more information on the practical aspects of how we work with AI here are some examples of how we’ve used AI.

The social media posts experiment

Mitra, our social media director Mitra has shared with us the results of her research into posting social media posts using AI.

Mitra pointed out the fact that AI was not a factor in the quality of the video and even used words she would never ever use. But, Mitra gave the idea an A+ for the captions she created when she was not inspired.

Lunch and take a class

I spoke at an afternoon and learned session on AI to the entire marketing team at Buffer in which I shared my experiences with used AI in my work.

Social media glossary

We have recently launched an Social Media Glossary that was a result of many months of work and cross-functional cooperation on marketing team.

 

Try Buffer for no cost

140,000+ small-scale businesses similar to yours rely on Buffer to help build their online presence each month

It could have taken us longer complete, but the use of AI reduced the amount of time spent thinking about what we should write about under each word and the best way to write it down in half. We created a prompt that assisted us in filling out the outline of every word in our Glossary, which we used as the base for every word.

We also considered the effect of content that is AI-only on the performance of the page as well as its value to the user. We took a long time making edits and adding sources to boost the value users receive from visiting every page.

The entire Buffer website

Our team is working on a plan to make Buffer’s website more local to other languages, beginning with Spanish and AI has assisted us in achieving the majority of this.

Of course that we’re using a real translator to verify our work however, we were able do the majority in the first translation using AI.

What’s the next step for AI on Buffer’s Content team?

What do we need to do from this point? In truth the truth is that nothing else is going to alter for us. Most of the time, we’ll be experimenting with new tools and implement them when they are logical. We’ve used AI tools to help us in our work for several months and they’ve helped us complete our job better and more efficiently.

It’s important to remember that these safeguards (apart from the kind of things we’ll not do using AI) aren’t necessarily strict. We don’t know what’s to come over the next few months because of the advances in AI and we’ll remain open to making adjustments in the event of need.