What is generative AI?
Generative AI refers to algorithms, such as ChatGPT, that have the ability to generate diverse forms of content, ranging from audio and code to images, text, simulations, and videos. Recent breakthroughs in this field have the potential to revolutionise how we approach content creation, opening exciting possibilities for innovation, creativity, and efficiency.
How I Started
When I embarked on my journey of using generative AI to create an AI marketing campaign, I began experimenting with two different AI models. Among them, the Chat GPT Bing plug-in stood out as a more user-friendly and intuitive option. Despite lacking some customisation options, it exhibited a more effective learning capability and demonstrated a more "human" memory. For example, when I requested both AI models to retain specific parts of the text and rework the rest, the plug-in offered a smoother experience compared to the built-in Chat GPT in Bing, which struggled with this task.
What I Did
One important aspect to remember, especially in the beginning, is that you’re not interacting with a human when using generative AI - it’s crucial to be explicit and precise in your instructions. I discovered that the AI models had difficulty grasping more nuanced requests, such as writing in a specific tone or reworking text in a particular manner. They were not programmed to comprehend instructions like "pull out key features and benefits," leading to challenges in achieving the desired outcome. When used for crafting concise marketing emails, the AI models were also not the most effective tool. They often failed to adhere to specified word limits, and when provided with industry-specific blogs they assumed the reader belonged to that industry, resulting in misplaced focus. Despite my efforts, it was challenging to make them understand and deliver my desired copy.
While the built-in Chat GPT in Bing offered additional content generation options, such as choosing the format, tone, and length of the generated text, it was not an intuitive tool to use. Initially, I attempted to be polite in my prompts, but I soon realised that being direct and clear was more effective, even if it felt odd compared to polite human interactions. Engaging in a considerable amount of back-and-forth interaction with the AI proved to be time-consuming and frustrating. Generating larger, meaningful pieces of text was a struggle, although the AI excelled in crafting subject lines for emails and generating effective call-to-action statements. I leveraged these strengths to enhance certain aspects of my writing.
Amidst the challenges, there were notable positives in using generative AI for an AI marketing campaign. The Chat GPT Bing plug-in, with its intuitive, user-friendly nature, provided a valuable foundation for content creation. Its effective learning capability and ‘human’ memory allowed for a smoother experience when retaining specific parts of text and reworking the rest. Additionally, the AI models excelled in crafting subject lines for emails and generating compelling call-to-action statements. These strengths were instrumental in enhancing the overall impact of the marketing campaign.
Throughout the process, I encountered difficulties in instructing the AI systems - even after repeating instructions multiple times. Basic tasks such as requesting the use of a single comma or converting text to British spelling often led to perplexing responses and unrequested rewriting of the text. It became evident that the need to phrase instructions precisely and in detail is crucial to unlocking the full potential of generative AI.
Using generative AI to create an AI marketing campaign is a complex endeavour – it certainly isn’t as simple as inputting some text and expecting the AI to do your job for you! If you’re a novice using AI to streamline your work, you will be disappointed. From my experience I could have written the email content just as quickly, if not quicker, than the AI.
While certain AI models may offer user-friendliness and intuitive experiences, challenges arise in conveying nuanced instructions and generating larger pieces of text. It’s important to balance the strengths and limitations of generative AI, understanding that it is a tool that requires precise communication and human expertise to achieve optimal results.
Nonetheless, the potential of generative AI is massive, offering a wellspring of creative inspiration, a reduction of content creation time, and fresh perspectives.
To use these tools effectively, you need to understand how to properly prompt these AIs to get the results you need. One of the main ways to do this now is to work with consultancies such as BJSS that offer comprehensive understanding of these technologies and can provide solutions that are more tailored to your organisation.
BJSS has the skills, knowledge and innovative outlook to understand your business objectives to implement the best MLOps and AI solution to meet your needs, all using industry-leading technology supported by our partners.
Visit our AI page to learn more.