There's more to AI than just generative AI. Figure 1 shows how generative AI relates to AI, automation and technology. Let's define these terms.
Figure 1
Generative AI (GenAI)
Generative AI is a type of AI that creates content from existing patterns of data such as text, video and audio. Prompt engineering is about writing clear instructions or prompts to get the best outputs from the generative AI tool.
Artificial intelligence (AI)
AI is the automation of cognition, that is automating the collection, processing and interpretation of information. AI is also a wide field of study which includes:
- Machine learning - Teaching computers to learn from data.
- Natural language processing - Teaching computers to interpret and generate human language.
- Computer vision - Teaching computers to interpret images and videos.
AI algorithms, models and systems
An AI algorithm is a step-by-step guide for a computer to follow (like a recipe). An AI model is a representation of patterns in data, created using algorithms.
Some AI models use probabilities to handle uncertainty, which means their outputs can sometimes be wrong – as with anything designed and built by people.
A Large Language Model (LLM) predicts which words come next in sentences. It uses neural networks to recognise patterns in language from large amounts of data. A neural network is a way for a computer to learn patterns from data by changing its internal connections, inspired by how brains work.
Generative AI tools like ChatGPT, Claude and Copilot are built on large language models that can be adapted for many tasks. A model trained on large datasets that can be adapted for many tasks is also known as a foundation model or general-purpose AI system.
Automation and hyperautomation
Automation is about using technology to do tasks without (or with reduced) human assistance. AI, Robotic Process Automation (RPA) and even washing machines are examples of automation. RPA is used for repetitive tasks that people do on computers, like copying and pasting data between spreadsheets.
Intelligent automation (or hyperautomation) is about connecting technologies to execute business processes automatically on behalf of knowledge workers. For example, automatically updating a new joiner’s details on HR, payroll and IT systems when onboarding. Others say that hyperautomation takes it a step a further by automating a whole business process as much as possible. In onboarding, this could include automating the update of details on several systems through to the sending of relevant information to the line manager and new joiner.
Agentic AI is about setting a goal without prescribing the steps, so the AI works out how to achieve it and adapts as needed within the boundaries given. Agentic here means giving AI some agency or autonomy to act independently, while ensuring people review key decisions and approve important actions. Widespread use of generative AI tools has revived interest in agentic AI because you can tell the AI what you need in human language. Building on the onboarding example, if there’s no spare laptop, the AI onboarding agent might order an approved laptop model, inform the line manager and update the onboarding progress tracker.
Technology
Technology is the application of knowledge or processes to achieve practical goals that are reproducible. This includes automation and other tools which aren’t necessarily about reducing human involvement. Examples include communication technologies like the internet and videoconferencing tools.
Digital transformation
Digital transformation is the ongoing work of aligning technology, people, culture, structure and tasks so that an organisation can thrive. Kane et al describe this continual realignment as becoming more digitally mature as an organisation.
