Covering Disruptive Technology Powering Business in The Digital Age

Rewriting the Rules of Creativity With Generative AI

Written by: Izzat Najmi, Journalist, AOPG.


Generative Artificial Intelligence (AI) has been one of the most significant breakthroughs in the field of AI in recent years. Generative AI refers to the class of AI that involves training machines to generate new content, such as text, images, music, and videos, among others. These AI models have the ability to generate realistic and coherent content that can be used in various fields, ranging from entertainment to healthcare, education, and finance, among others.

From GAN to GPT

Generative AI has a rich and diverse history that spans several decades. It all began in the 1950s when the first AI models were developed to generate content. However, these models were primitive and limited in their capabilities. It was not until the advent of deep learning techniques, such as deep neural networks, that generative AI models became more feasible and effective. In recent years, the development of generative AI has been accelerated by the availability of large-scale datasets and computing power. This has enabled the training of large-scale generative models, such as GPT-3 and BERT, which have been used in various applications, including natural language processing and content generation.

One of the most significant milestones in the history of generative AI was the development of the Generative Adversarial Networks (GAN) algorithm in 2014 by Ian Goodfellow and his team. GANs are a class of AI models that are composed of two networks: a generator and a discriminator. The generator is responsible for generating new content, while the discriminator is responsible for determining whether the generated content is real or fake. The two networks are trained together, with the generator trying to generate content that can fool the discriminator, and the discriminator trying to distinguish between real and fake content. This process results in the generation of highly realistic and coherent content, such as images, videos, and text.

Another significant milestone in the history of generative AI was the development of the Variational Autoencoder (VAE) algorithm in 2013 by Diederik P. Kingma and Max Welling. VAEs are a class of AI models that are used to generate new content by learning a latent representation of the input data. The model is trained to encode the input data into a compressed latent representation, which can then be used to generate new content by decoding the latent representation back into the original input space.

The development of deep learning techniques, such as GANs and VAEs, has led to the creation of a new generation of generative AI models that are capable of generating highly realistic and coherent content like what we are seeing now with ChatGPT and Bard. These models are now being used in various fields, ranging from entertainment to healthcare, education, and finance, among others.

As generative AI continues to evolve, we can expect to see more applications and use cases in various fields. It is important for researchers and developers to continue exploring the potential of generative AI and to ensure that its development is guided by ethical and responsible principles. By doing so, we can maximise the benefits of generative AI while minimising its potential risks.

Benefitting the Beneficial

Generative AI has opened up new avenues for innovation and has impacted various domains of human society. With its ability to create new and innovative content that is relevant to specific domains, generative AI is set to revolutionise various industries. The following are some of the fields that have greatly benefitted from generative AI:

  • Entertainment: Generative AI has already made its mark in the entertainment industry. AI models have been developed that can generate music, art, and videos, among other things. For example, Amper Music, a company that specialises in generating custom music, has created an AI model that can generate music tracks for users based on their preferred genre, tempo, and mood. This technology can be utilised to create customised soundtracks for movies, TV shows, and advertisements, among other things.
  • Healthcare: Generative AI has also made a significant impact in the healthcare industry. It has been used to generate synthetic medical data, such as medical images and patient records. Researchers have used this data to develop more accurate and effective models for diagnosing and treating various medical conditions. For example, the company Insilico Medicine has developed AI models that can generate synthetic patient data to speed up drug discovery and development.
  • Education: Generative AI has the potential to revolutionise education by generating personalised and engaging educational content. AI models have been developed that can generate textbooks, quizzes, and interactive simulations, among other things. This technology has enabled educators to provide a more personalised and engaging learning experience for students.
  • Finance: Generative AI has also made significant strides in the finance industry. It has been used to generate financial reports, investment advice, and risk assessments, among other things. This technology has enabled investors to make more informed investment decisions based on accurate and timely information. For example, the company Ayasdi has developed AI models that can analyse large amounts of financial data to identify investment opportunities and minimise risk.

Generative AI has opened up new horizons for innovation in various industries. Simon Dale, who is the Managing Director of Southeast Asia & Korea at Adobe, has expressed his excitement over the potential of advanced AI. He believes that this technology could unlock a whole new world of possibilities for businesses and individuals alike. Just imagine, with the help of AI, we can delegate the more repetitive aspects of our work, freeing up valuable time to focus on the truly innovative and strategic aspects of our job.

For Simon, the allure of AI lies in its ability to provide an intelligent assistant for creative and marketing tasks. This would enable us to shift our focus towards more strategic aspects of our work, thereby enhancing our productivity and creativity.

However, it is important to note that generative AI is not meant to replace human creativity and expertise, but rather to complement them. The true potential of generative AI lies in its ability to collaborate with humans to achieve remarkable results.

Here To Stay but Not To Replace

The emergence of generative AI has opened up new possibilities for human creativity, allowing us to generate content that was once unimaginable. From music to art to writing, the potential applications of generative AI are vast and ever-expanding. With the help of machine-learning algorithms, we can create entirely new pieces of content that are tailored to specific domains, unlocking new opportunities for innovation and discovery.

Nevertheless, generative AI is by no means a “magic box”, so to speak. Some thought has to go into how we leverage it. For example, according to CK Tan, Qlik’s Senior Director of Solutions & Value Engineering, most generative AI chatbots provide valuable and relevant responses to questions asked, but developing the right questions is essential to using these tools effectively, as the quality of the responses is directly tied to the quality of the questions presented. Hence, at this stage in the development of this technology at least, it is crucial for users to ask well-formed and precise questions in order to maximise the usefulness and accuracy of the chatbot’s responses. Otherwise, garbage in > garbage out.

Moreover, it is important to remember that generative AI is not a replacement for human creativity and expertise. While AI models like GPT-3 and BERT have made significant strides in generating human-like text, they still lack the ability to understand the context and nuances of human language. This is where humans come in, bringing their own unique perspectives and expertise to the table to help improve the quality of the generated content.

By combining the strengths of generative AI with human creativity and expertise, we can achieve remarkable results. For instance, AI models can be trained on large datasets to identify patterns and generate new content, while humans can provide the necessary domain knowledge and context to ensure that the generated content is both accurate and relevant.

Ultimately, the future of generative AI lies in the collaboration between humans and machines. As we continue to develop and refine these technologies, we will discover new and exciting ways to apply them in different fields, from entertainment to education to healthcare. With the power of generative AI at our fingertips, the possibilities are endless.