[Chatbots, ChatGPT-3.5, GPT-4, GPT-4.o, Gemini 2.0 flash, Copilot, DeepSeek, Meta AI, Claude, AIthor research assistant, etc.]
GENERATIVE ARTIFICIAL INTELLIGENCE (A.I.)
Generative AI (Gen A.I) is a broader class of Artificial Intelligence systems that focuses on creating something new. This “something new” or content can be text, image, sound/audio, music or even video. A.I. technology make use of machine learning to return prompt output based on user’s inputted prompt. A.I development engineers commonly use supervised learning on a very large dataset (leaning from human-generated content) to train Gen A.I model, from where the model consults to determine the best possible response to a user’s prompt.
You can actually see Generative A.I. as a form of predictive artificial intelligence. Simply the use of A.I to create or generate something new. These A.I systems usually learn by examples from different kinds of datasets. Then, they can create original outputs that are similar to what they have seen or learnt from the training datasets.
For example, if you train a Generative A.I on images; it will be able to create new realistic photographs from a text prompt. Also, A.I. trained on musical notes can compose a tune and output similar musical sound.
Examples of Gen A.I, – ChatGPT, DALL-E, Sora, Midjourney, MusicLM (text to music A.I.), Gemini (MultiModal LLM for text, audio, image etc by google DeepMinds), etc.
LARGE LANGUAGE MODELS (LLM)
A Large Language Model (LLM) is a particular type (subset) of Generative A.I that focuses mainly on understanding, generating, and interacting with language to perform various natural language tasks, they also power chat-bots.
To understand human grammar, context, meaning, and intent, LLMs are being trained on massive amount of textual data from books, websites, articles, conversations or other documented materials created by humans. These models can write essays, answer questions, summarize documents, translate languages, write computer codes, and even engage in human-like conversations. Within the broader field of Generative A.I., you can consider LLM as a language specialist
Examples of LLM: Generative Pre-trained Transformer models like GPT-4, GPT-3.5 (by OpenAI), Claude(focused on constitutional AI), LlaMA, DeepSeek R1(reasoning model), Grok-2 (by xAI), Gemini (MultiModal LLM for text, audio, image etc by google DeepMinds).
CHATBOTS
A chatbot is a kind of conversational AI designed as a computer program (powered by LLM) to simulate interaction or conversation with the human users. This can take place through either text, voice or imagery interactions.
As a type of artificial intelligence (AI), they are designed with the ability of mimicking human-like conversations. It can be integrated and be used in a variety of applications, such as; a customer service application and information retrieval applications amongst others.
Chatbots can be general or specialized for use in banking systems, marketing, customer cares, aviation, health, education and other places. Examples: Duolingo (Language Learning Bot), JetBlue (Customer Service Chatbot), Leo (UBA), Fibani (firstbank ng), Babylon health, Florence health assistant, Healthily etc.
Pictorial similarity & difference between Gen AI and LLM below:
GUIDE ON HOW TO USE GENERATIVE AI (CHATGPT)
ChatGPT is a Generative Artificial intelligence model developed by OpenAI and launched in November 30, 2022. It utilizes Large Language Models (LLM) used in creating human-like responses in the form of text, speech, and images.
It has access to features like the ability to search the web, create image, write codes, using apps, and running programs. (powered by Engine => GPT 4o, 4.1).
You can use the free tier or upgrade to ChatGPT Plus/Pro/Team/Enterprise for more powerful models and features.
You can either:
1. Sign up or Login or Use without Sign up
2. Basic Usage
Type your customized prompt in the input box and hit Enter. You will receive a response in the chat window; you can accept the output or regenerate or refine your prompt.
Chats are automatically saved and listed in the sidebar, to recheck, click any of it to revisit, rename, archive, or delete.
3. Writing Good Prompts
Be clear: Be clear and specific. Include/indicate desired background, context, and the exact format you want in your expected response to appear (e.g., bullets, step‑by‑step or numbered).
Try to ask one question at a time to avoid overloading the model.
Use WH words: Use question words (“who, where, why, when, what) in order to clarify your intent for a relevant response from the Gen A.I.
Ask for multiple ideas: For creativity or brainstorming tasks, ask the Gen A.I. for multiple or varied ideas, which can be further refined it (e.g., “make these bolder”, “underline these”).
4. Advance features
Custom Instructions: Tell the GenAI (ChatGPT) exactly how you would like it to respond (in terms of tone, style, preferences).
Memory: It can remember your preferences and past context for personalized replies.
5. Useful Modes
Voice Mode: You may choose to talk (speak into your device mic) instead of typing a text prompt. ChatGPT responds to both text, speech and image.
Image/Document Upload: You can share images directly in chat for commentary or analysis.
Voice Mode: You may choose to talk (speak into your device mic) instead of typing a text prompt. ChatGPT responds to both text, speech and image.
6. Best Practices & Limitations
– Interrogate all responses: Fact-check chatbot outputs: Gen AI can sometimes hallucinate or produce incorrect information. AI hallucination refers to an output generated by an AI model that deviates from reality or lacks a factual basis.
– Use iterative prompting: ask follow-ups like “What if…”, “What about…”, “In an event of…” to refine the responses delivered to you.– Validate and oversee: Think of AI suggestions as a starting point not a conclusion of idea, your judgement actually matters most.
To read more on the applications of Gen A.I. in various industries, visit the link below:
What are the applications of generative AI?
Generative Artificial Intelligence (Gen A.I) is being applied in diverse industries, from healthcare sector, manufacturing, software engineering, financial services, entertainment and media, to marketing/advertising.
