Slowly but surely over the last decade, artificial intelligence (AI) and machine learning (ML) has become commonplace. From virtual assistants like Siri and Alexa, to creating code and poems with ChatGPT, AI is now everywhere around us and this trend is only going to accelerate as the technology develops.

AI and ML underpin the progress towards self-driving cars, more efficient national grids, safer security, and much more. We are still in the very early stages of the technology, but we have already reached the stage where people are questioning how we define creativity in this new world and whether we need to limit the technology’s use to avoid possible issues down the road.

It was not long ago that AI and ML were mostly being used for relatively trivial tasks such as helping your roomba vacuum find its way around your home or helping you with your online casino games selection. or recommending you the next book to read on Amazon. In 2023, however, we are already seeing AI that can successfully complete law school and medical school exams and create new “art” with a few simple prompts.

Here are two of the major trends in AI we expect to make headlines over the next 12 months.

Generative artificial intelligence (GAI)

AI is most widely understood as a way to automate routine and repetitive tasks, but generative AI is altogether different. These tools and algorithms, such as ChatGPT and Dall-E, can be used to create new and unique content, including text, images, videos, and code, and they have the potential to change how we create content and perceive creativity itself.

Traditionally, if you needed to create an article on a topic, you would need to employ a copywriter to research and create the content, which may take days or weeks. Generative AI means that such tasks can be completed in seconds with the touch of a button. And it is no different with images or audio – generative AI can create content that might take humans hours or days to produce in a matter of seconds.

These tools are trained on huge data sets, often scouring vast swathes of the internet and hoovering up all before them so that they can learn from what humans have produced in the past, and then use these elements to craft things that are entirely new. Or partly new at least, as generative AI can create new content, but it is still not capable of being “creative” and making huge leaps forward and create a new style or art form – it can only rehash what has come before.

Is this creativity or is this plagiarism and copyright infringement? It is likely we will see a series of court cases in the coming years that might answer that exact question and the creative industries take aim at the bots.

Explainable artificial intelligence (XAI)

As artificial intelligence progresses, the methods these algorithms use to deliver their output has become increasingly opaque, which has resulted in people finding it difficult to trust the results. Just because people can ask a chatbot a question and receive an answer does not mean that they trust that answer, and unless we can understand how these answers are derived it can be difficult to trust the results.

Explainable AI or “XAI” is a rapidly developing field, where the AI model and its output can be understood with with reference to both its impact and biases. This transparency can help developers know that a system is working correctly, help regulators know that a system meets their requirements, and most importantly allows those affected by the AI’s decision to be able to challenge or change the outcome. If people can trust how an algorithm has come to a conclusion, they are more likely to trust the results.


Comments are closed.