by Nicole Borg, PR & Content-Writing Client Partner
While the term Artificial Intelligence (AI) was coined by John McCarthy in 1956 and has since then been researched and developed year upon year, in the past few months it has gained more and more popularity. AI and its latest advancements started to capture the public’s eye again when in July 2022, a set of visuals generated by an AI called DALL-E 2 predicted what the last selfies to be taken on planet Earth will look like. The AI generated burnt and hurt human like individuals with a backdrop of fire or balls of fire falling from the sky that got people talking.
A few months later, we witnessed another AI craze on Facebook with people using various AI art generators that turned their photos into art. And after that, the craze moved to TikTok following the launch of an AI filter that turned individuals and their surroundings into manga art.
In the recent weeks, there was a shift from visual AI generators to text AI generators with the main one being ChatGPT. These advancements started an interesting discussion on whether this was the start to the end of several roles within the Marketing Industry. However, before we go into the effects of such a platform on the Industry particularly when it comes to content generation, let’s start by understanding what this platform is all about.
What is ChatGPT?
Ever since the invention of Google, we have become accustomed to access the search engine and type in the most varied of questions expecting to get a list of links that could help us get to the answer we are looking for. ChatGPT aims to help us answer our questions too, however, instead of linking us to other platforms that can give us the answer we need, it uses an AI-powered language model developed by OpenAI to generate human-like responses to natural language queries, making it capable of having conversation with people on various topics, using advanced machine learning algorithms.
This language model is based on the Transformer architecture, which is a type of neural network designed for processing sequential data such as text. The platform is trained on a large dataset of text, such as books and articles, which allows it to learn the patterns and structures of human language. When given a prompt, such as a question or a sentence, it then uses the patterns it has learned from the training data to generate a response that is contextually relevant and grammatically correct. The technique used by ChatGPT is called “masked language modelling” where some tokens in the input are replaced with a special token called “mask” and the model predicts the missing token based on the input context. The model is then fine-tuned on specific tasks by training it on task-specific data, such as customer service conversations or product descriptions. This means that ChatGPT can also be used in various applications such as chatbots, text summarisation, text generation and more.
Should Content Writers be worried about ChatGPT?
When accessing ChatGPT you are presented with several disclaimers, the main one being that the answers it provides may need to be fact checked as they might not be 100% accurate. The reason being that the platform learns more and more depending on the frequency with which it is asked questions. Additionally, users are also told that the platform may occasionally produce harmful instructions or biased content, while it is also limited with knowledge regarding world events after 2021. Meaning that there needs to be a certain level of human detail and effort put into the content produced by ChatGPT to ensure that what would then be released to the public is factful, true and not offensive.
Users around the world have been turning to ChatGPT to easily find answers to questions that might be a little bit more difficult to research. Others have used it to support them with tasks such as the debugging a programming code or support when writing a blog post. If one would like their content to thrive in the online world and be picked up by search engine giants such as Google, a platform like ChatGPT should only be used as a support mechanism for content generation and not as the sole method of content generation. Taking Google as an example, the search engine platform tends to penalise websites and blogs that publish and promote AI-generated content and marks it as spam.
Using ChatGPT as a support system for and not a replacement to Content Writers
All of what we said so far brings us to the main realisation that while ChatGPT is an interesting tool, it is far from perfect, and in order to remain relevant within the online world and as a business, we need to take such platforms as tools to aid human processes and not to replace them.
When writing a blog, ChatGPT can help you in the following ways:
- Generating text – you can provide ChatGPT with a prompt or topic and it can generate brief text that is grammatically correct and that can help you get started in the content writing process.
- Summarizing information – if you’re over your word count but you’re not understanding what content can be removed, you can supply ChatGPT with long-form text or a website link and it summarizes the information for you by extracting the main points. This could also be used to do quick research when tight on time.
- Proofreading and editing – After drafting your content, you can request ChatGPT to assist with proofreading and editing.
- Generating headlines – if you feel like the headline, you’ve given your blog is not attractive enough, you can use ChatGPT to help you come up with a compelling headline and meta descriptions that will help you drive traffic to your content.
- Generating ideas – if you’re going through a writer’s block and have no idea from where to start your next blog post content, you can use ChatGPT to help you identify key points you could tackle or a potential list of topics and angles for your articles and blogs.
It is important to keep in mind that while the text generated by ChatGPT is grammatically correct, it might not make sense semantically. Therefore, it is always recommended to thoroughly review the content generated by any AI before publishing.