AI glossary · Fundamentals
Generative AI
In one sentence
Generative AI is artificial intelligence that creates new content, such as text, images, audio, video or code, from an instruction, using models trained on huge numbers of examples of that kind of content.
Also known as: GenAI · gen AI · generative artificial intelligence
What is generative AI?
For years, the most common kind of artificial intelligence was the kind that sorts and predicts: catching spam, recommending a product, forecasting demand. Generative AI does something different. Rather than choosing between options, it produces an output that didn’t exist before, whether that’s an email, a summary, an image or a voice recording.
Put it to work
An agent that writes, checks and publishes
From €49/mo + VAT
It writes the article, prepares each network's posts and publishes on schedule.
See the content agentIt went mainstream in late 2022 with chat assistants built on large language models, but the field is broader. Some models generate images, some generate speech and some work across several types of content at once.
Related to this term
UTIO Content
- Never publishes a figure without a source
- Blog, LinkedIn and Facebook on autopilot
- Your first campaign is free
From €49/mo + VAT
See the content agentHow it works
During training, a generative model learns what the content it has seen tends to look like. When you give it an instruction, it builds the answer piece by piece:
- It reads the instruction, which text models split into tokens.
- It predicts the most likely next piece: the next token in a text, or, in many image generators, an image gradually refined out of random noise.
- It repeats until the output is complete.
That explains both its strength and its weak spot. It writes fluently because it has seen an enormous amount of text. But it generates what’s plausible, not what’s verified: if it doesn’t have a fact, it may invent one. That’s why businesses pair it with their own sources, as RAG does, and with human review.
Example
A craft brewery in Kent wants to post on its blog and social channels every week, but nobody on the team has time to write. From a few notes about a new seasonal ale, generative AI produces a draft article, a short LinkedIn version and an Instagram caption.
The brewery doesn’t publish the draft as it stands. It checks the ABV, the hop varieties and any awards mentioned, adjusts the tone and signs it off. The AI removes the blank page; responsibility for what goes out stays with the brewery.
Generative AI at UTIO
UTIO’s content agent applies generative AI to exactly that job: it writes the article, prepares each network’s posts and publishes on schedule. What sets it apart from asking a chatbot for a post is control. It starts by auditing sources: no figure goes out without one, and it never invents customers or case studies. And you approve every campaign before it’s published.
The voice and WhatsApp agents also use generative AI to word each reply, grounded in your knowledge base and with whatever level of autonomy you choose.
Frequently asked questions
How is generative AI different from traditional AI?
Traditional, or predictive, AI analyses data and returns a label or a number: this email is spam, this customer may leave. Generative AI produces something new: a piece of text, an image, a voice. Many real products combine both.
Can you trust content made with generative AI?
Not by default. A generative model produces plausible text, not necessarily true text, and it can make up figures or sources. Check numbers, quotes and any claim about your business before you publish.
What are some examples of generative AI?
Chat assistants built on language models that draft and summarise, image generators that work from a description, text-to-speech that reads text aloud in a natural voice, and coding assistants that suggest code to developers.
Related terms
- LLM
An LLM (large language model) is an AI model trained on vast amounts of text to predict the next word. That single skill lets it draft, summarise, translate, classify and hold a conversation.
- Artificial intelligence
Artificial intelligence (AI) is the branch of computing that builds systems able to do tasks we associate with human intelligence, such as understanding language, recognising images or making decisions, by learning patterns from data.
- AI hallucination
An AI hallucination is a response from an AI model that sounds convincing but is false or not backed by any source: a figure, quote, price or reference the model invents while generating the most probable text.
- Prompt
A prompt is the instruction or text you give an AI model to get something done: a question, a command, a document to summarise or a mix of all three. Much of what the model gives back depends on the prompt.
- Multimodal AI
A multimodal model is an AI model that can take in or produce several types of data, such as text, images, audio or video, and connect them. For example, it can read a photo of an invoice and answer questions about it.
Updated on