
What do I mean by this? I ran a little experiment in January: continue to integrate the new and improved ChatGPT into my work life to see what it’s capable of and if the nature of my work changes.
ChatGPT has been in and around our agencies at WPP during its development for some time. Many creative teams have been using it as a ‘test and learn’ tool in the ideation process in a project’s initial stages.
Its output serves as a springboard for even more innovative and imaginative ideas crafted by human creativity.
No doubt, as we see more improvements in AI and we share techniques on how to use them, we will see AI enhancing the creative process but seeing it replace the power of the creative human is highly unlikely.
The advertising industry will need to continue to make strides in reducing bias in AI results, and continual efforts toward checks and balances will continue to promote ethical and empathetic creative work by human professionals.
The debate on what ChatGPT can and can’t do with the creative industry way of working will rage on, but the only way to find out is to put it to the test. And let me tell you, AI and machine learning are taking over the workplace faster than a robot on steroids.
The nature of work inside a creative transformation company is constantly changing and evolving, and the debate on what ChatGPT can and can’t do will be around for some time, but the only way to find out is to use it often for everything and see how it performs.
Artificial intelligence and machine learning technology have been growing with our agencies. They will play a more significant role in the workplace as the year role on as we integrate them into a broader range of systems and applications, and our clients demand help in their own evolution of AI-powered brands. It is difficult to predict precisely how these developments will play out and their specific impact on the nature of work this year but we are investing greatly that it will.
My experiment showed that conversational interfaces increase the demand for advanced language models like ChatGPT. I connected it to a text-to-speech engine, settling on Microsoft Azure Text-to-Speech with over 180 voices, and added delay to make the AI voice more like me.
ChatGPT shines in its real-time responsiveness, consistent responses, and multilingual language use. Although it has a higher error rate than my human approach, its knowledge base is constantly growing and updated.
The knowledge base of ChatGPT is still limited to the information and knowledge it has been trained on, whereas my human brain, after some fifty years plus on this planet, still has a broader range of knowledge and experiences. That is very reassuring, knowing that I still have the edge over my AI counterpart.
Our tech team put my bot version to test with a CAPTCHA. An Automated Public Turing test to tell Computers and Humans apart – and I passed with flying colours. But the reality is that AI is making it increasingly difficult to differentiate between bots and humans. This will play a huge role in agency communication in the future.
Thanks to ChatGPT, I’ve saved time and increased efficiency in my work. It’s a valuable tool for data analysis and allowed me to collaborate with my team by assisting with scheduling and communication.
But there are limitations, and content creation is one of them. ChatGPT can create basic content, like product descriptions or social media posts, but has yet it has not replaced the creativity of the amazingly talented people in our agencies. That said, the possible workstreams are exponential.
My next goal is to make ChatGPT a more engaging and interactive training tool, allowing learners to ask questions and receive real-time feedback. But in the end, the human touch is still essential today, but replication is possible and coming, and I expect considerable advancement over the coming months. It’s the brief of the year.