The hype about generative AI is huge, and it continues to grow. According to a Gartner report, generative AI is one of the most impactful and rapidly evolving technologies that brings productivity revolution. 

By 2025, generative AI is expected to produce 10 per cent of all data (now less than 1 per cent) with 20 per cent of all test data for consumer-facing use cases. It will be sided by 50 per cent of drug discovery and development initiatives. Thirty per cent of manufacturers will use it to enhance their product development effectiveness. 

What is generative AI? 

Generative AI refers to unsupervised and semi-supervised machine learning algorithms that enable computers to use existing text, audio and video files, images, and even code to create new possible content. Generative AI allows computers to abstract the underlying patterns related to the input data so that the model can generate or output new content.  

Generative AI can already do a lot. It can produce text and images, spanning blog posts, program code, poetry and artwork. The software uses complex ML models to predict the next word based on previous word sequences or the next image based on words describing previous images.  

Presently, there are two most popular generative AI models. Generative Adversarial Networks can create visual and multimedia artefacts from imagery and textual input data. In contrast, transformer-based models such as Generative Pre-Trained (GPT) language models can use information gathered on the Internet to create textual content from website articles to press releases and whitepapers. 

Human involvement 

To use generative AI effectively, human involvement is necessary at the beginning and the end of the process. For example, a human must enter a promo into a generative AI model to have it create content. “Prompt engineer” is likely to become an established profession, at least until the next generation of even smarter AI emerges.  

Once a model generates content, it will need to be evaluated and edited carefully by a human. Alternative prompt outputs may be combined into a single document. Image generation may require substantial manipulation.  

Applications of Generative AI 

Generative AI in location services involves converting satellite images to map views. This can be a huge step towards venturing into unexplored geographic locations. In the motion picture industry, with generative AI, one wouldn’t need to wait for hours or days to capture a frame in perfect lighting or weather conditions but capture at any time convenient and convert that into whichever conditions are needed. 

Generative AI can take search engine services to the next level. For example, text to image translation. The technology can be used for security services at airports and country borders as it can create front-on photos from photos taken at different angles and vice versa for face verification or face identification system. In the healthcare sector, Generative AI converts inputs that are semantic images or sketches to photo-realistic images. 

Going ahead 

Generative AI requires a considerable amount of training data required to generate outputs, else the output may turn out to be subpar or not good. But an enormous amount of work needs to go into securing the data to avoid any privacy concerns.  

Like any other AI domain, Generative AI also tends to grow with more and more applications across multiple industries.