What is a large language model? A plain-English guide
How large language models like the ones behind ChatGPT, Claude and Gemini actually work, what they're good at, where they fall short and how businesses use them.
How the main branches of AI relate to each other, what each is good at, and which approach fits common business problems.
Artificial intelligence is the broad goal of machines performing tasks that need intelligence. Machine learning is a set of methods that learn patterns from data. Deep learning is a type of machine learning using large neural networks, excellent for images, audio and text. Generative AI uses deep learning to create new content such as text, images and code. Many business problems are still best solved with simpler machine learning.
AI terms are often used interchangeably, which causes confusion when planning projects. Understanding how they fit together helps you pick the right tool rather than the most fashionable one.
Think of nested circles:
Traditional machine learning uses algorithms such as linear regression, decision trees, random forests and gradient boosting. You provide examples with features, such as customer age, purchase history and region, and an outcome, such as whether they churned. The model learns patterns to predict outcomes for new cases.
Good for:
Advantages: often needs less data and computing power, trains quickly, and can be easier to explain.
Deep learning uses neural networks with many layers that learn complex features directly from raw data such as pixels, audio waves or text.
Good for:
Trade-offs: typically needs more data and computing power, and decisions can be harder to explain.
Generative AI models learn from enormous datasets and produce new content: text, images, audio, video and code. Large language models are the best-known example. Our guide to what a large language model is explains how they work.
Good for:
Trade-offs: can produce confident errors, cost per use can add up, and outputs need checking.
| Machine learning | Deep learning | Generative AI | |
|---|---|---|---|
| Typical data | Structured tables | Images, audio, text, large datasets | Huge text, image and multimedia datasets |
| Output | Predictions, scores, categories | Recognition, classification, detection | New content |
| Data needed to build | Moderate | Large | Usually use pre-trained models |
| Explainability | Often easier | Harder | Hard |
| Typical cost to use | Low | Moderate | Varies with usage |
Many solutions combine approaches: a generative assistant that calls a traditional forecasting model, for example.
Most businesses don't train large models from scratch. Options include:
Our comparison of open-source and proprietary AI covers the model choices.
All approaches depend on good data: accurate, relevant, representative and well governed. Biased or incomplete data leads to biased or unreliable results.
Patterns in data change: customer behaviour, prices, products and seasons. Models trained on old data gradually become less accurate. Monitor performance and retrain or update models when accuracy drops.
When AI influences decisions about people, such as lending, hiring or pricing, being able to explain outcomes is important ethically and often legally. Simpler models or explanation tools may be preferable in these cases.
Our AI solutions team helps choose between machine learning, deep learning and generative AI for your problem. Our web app developers integrate models into business systems, and our app development service builds AI features into products.
Yes. It's built on large language models, a form of deep learning, which is part of machine learning.
Often not. Traditional machine learning works very well on structured business data.
AI is the broad field; machine learning is a set of methods within it where systems learn from data.
This guide is general information, not legal, tax or financial advice, and details such as prices, features and rules can change. Product and company names mentioned are trademarks of their respective owners and are used only to identify those products; UK Technology Hub is not affiliated with or endorsed by them.
Tell us about your project and get a free, fixed quote.
Get a free quoteHow large language models like the ones behind ChatGPT, Claude and Gemini actually work, what they're good at, where they fall short and how businesses use them.
How open-weight and proprietary AI models compare on capability, cost, control, privacy, licensing and operations, and how to choose for your project.
The real differences between native, hybrid and cross-platform mobile apps in performance, cost, user experience and long-term maintenance.