Attested Media

Last year, I met with three friends for a social catch-up. One is an engineer focused on content generation, another is an IT professional transitioning into programming, and the third, like me, works in communications. As is often the case with men, our conversation went off on tangents, and an interesting topic arose: the influence of Artificial Intelligence (AI) on content creation.

My friend in communications and I, perhaps a bit defensively, dismissed AI-generated content, believing it’s easy to identify. Our friends argued that most people wouldn’t notice AI-written content on social media. It felt unsettling like non-professionals were indifferent about the communication skills we had developed for years through sweat and blood.

This encounter sparked my curiosity about generative AI. So far, I’ve learned about its workings, how it’s impacting the communications field, some ethical concerns and potential future implications.

Generative AI: A Powerful Tool

My research revealed two key AI tools: supervised learning and generative AI. Supervised learning trains computers to perform tasks like spam filtering by analyzing vast amounts of data. This is why, after searching for leather shoes, your social media feeds suddenly overflow with shoe ads – from local stores in Kathozweni to international brands in Seattle 😊. You might even suspect your phone has secret voice recognition after discussing something and seeing related ads pop up.

Generative AI builds upon supervised learning. It involves labelling data within large language models (LLMs). Take an example of a prompt like “I love playing.” An LLM can complete this sentence with various possibilities – “with my friends,” “during holidays,” or “hide and seek.” Generative AI essentially uses supervised learning to repeatedly predict the next word.

Subsequently, if an AI system encounters the sentence “I love playing with my friends,” it breaks it down into data points. Each data point asks: “Given the phrase ‘I love playing,’ what is the next word?” The answer could be “with my friends,” “during holidays,” and so on.

The Art of the Craft: Balancing AI and Human Creativity

Those familiar with Kenya’s 8-4-4 education system might recall composition writing exercises. You’d be given an incomplete phrase to build a story around. Some students kept notebooks with “new words,” interesting terms they encountered, and collections of similes, idioms, and striking phrases. These notebooks became useful tools for crafting impactful compositions. Certain phrases became ubiquitous such as “he glared at me with dagger-like eyes,” or “He cried in excruciating pain.” Sometimes, students over used some of these phrases even out of context.

Interestingly, LLMs may have borrowed a similar concept. They can become monotonous, overusing certain phrases. Conversely, content written with specific vocabulary can be dismissed as AI-generated. This raises the question: can we no longer use words like “thrilled”?

The Future of Content Creation: Collaboration, Not Replacement

Like many professionals, I’m pondering AI’s role in content creation. Understanding generative AI’s capabilities and limitations is crucial. As the name suggests, “artificial” intelligence exists alongside “natural” intelligence – our own human minds. AI surpasses us in memory and information processing speed, but human intelligence excels in areas like conceptualization, decision-making, and creativity. Overreliance on AI for everything risks diminishing these vital human capabilities.

Therefore, I propose using AI to assist natural intelligence in information processing, rather than replacing content creation or idea generation. This stance may evolve as I delve deeper into the world of AI.

For now, feel free to use phrases like “I am thrilled to announce” or “I am happy to do this and that” – as long as you’re leveraging your natural intelligence alongside the power of AI. Those who stand out will be those who use their natural intelligence to identify the possibilities and limitations of AI.