Practical No 3

Generative AI

Abdul Kabir Hussain SOLIHU

Introduction

Generative Artificial Intelligence (Generative AI) is a field of AI capable of generating new content of text, images, video, audio or other data using generative models, based on patterns learned from existing data. this essay covers Generative AI definition, machinery, capability, applications that can be used for Generative AI.

Definition and Machinery of Generative AI

Generative AI refers to a category of Artificial Intelligence algorithms that generate new outputs based on the data they have been trained on. Unlike traditional AI systems that recognize patterns and make predictions, generative AI creates new content in various forms, including images, text, audio, and more (see Fig. 1).

Generative AI employs Deep Learning (DL) techniques that can learn from varied data sources and generate outputs within a satisfactory scope. This adaptability enables the same foundational model to be leveraged for multiple applications. ChatGPT is one instance that can now process visual and textual inputs. This subset of Machine Learning (ML) is already utilized to make creative digital artwork, envision novel virtual settings, compose music, formulate written material, help with medication development by anticipating molecular formations, code software, and produce realistic video and audio segments. Rouse M. (2024). Studies have shown that in 2023, four in five teenagers in the United Kingdom are using generative AI tools while about two-thirds of Australian employees report using Generative AI for work (Thomson & Thomson, 2013).

Generative AI Capability

Generative AI has the capability to process huge amounts of data and generate new, business insights. This means that in the near future, employees will be able to "talk" to AI about any information or insights within their company. Some companies have already implemented these abilities with great success. The international law firm Allen & Overy uses AI to access large volumes of legal content and produce insights, suggestions, and forecasts. Meanwhile, Syntegra, a healthcare technology company, leverages generative AI to share statistical patterns from medical data without revealing patients' private information.

In short: While it's easy to foresee thousands of small generative AI use cases, company boards need to encourage thinking bigger. Forrester predicts that spending on AI will double from $33 billion in 2021 to $64 billion in 2025. This presents an opportunity for companies to challenge or change their business models. Companies should consider ways to take advantage of their unique data sets, optimize operations, or differentiate within their industries. Ultimately, successful companies will embrace strategies that increase their competitive edge (Sibio, 2023).

Generative AI Applications and Tools

The following are some applications/tools that are currently available for use for Generative AI and their classification:

Text Content: Large language models (LLMs) like ChatGPT, GPT-4, Copilot, and Gemini are chatbots that generate human-like responses (see Fig. 2 & Fig. 3).
Image Content: Systems like Stable Diffusion, Midjourney, and DALL-E produce artificial images from textual descriptions.
Video Content: Sora is an example of a text-to-video AI generator.
Music, Code, and More Content: Generative AI can also create music, software code, and other forms of content (Kothari, 2024).

Conclusion

In summary, generative AI is a powerful tool that opens up exciting possibilities and prospects across various domains and disciplines, but it also requires careful handling to address its challenges and ethical implications. Generative AI can pose serious challenge in and be misused for cybercrime, fake news, and deepfakes. By putting in place the code of ethics that will guide the use Generative AI, its misuse could be curtailed. The ethical code of Generative AI should also ensure that it does not displace humans in the workplace. It is human minds that created it in the first place; it must not make human minds irrelevant otherwise it becomes self-defeating.