From Technology to Practical Use
We have learned what Generative AI is and how it can create different types of content.
Now we can ask a more practical question:
What can people actually use Generative AI for?
The answer is broad.
Generative AI can assist with communication, education, research, software development, marketing, customer support, creative work, analysis and many other activities.
AI as an Assistant
A useful way to think about Generative AI is as an assistant.
The AI can help with parts of a task while a person remains responsible for deciding what the final result should be.
For example, a person might ask an AI system to create a first draft of an email and then review and edit it before sending it.
Writing and Communication
One of the simplest applications is helping people write.
Generative AI can assist with:
- Email drafts
- Reports
- Articles
- Summaries
- Meeting notes
- Announcements
- Social media drafts
- Translations
The user can provide the subject, audience, tone and other requirements.
Example: Writing an Email
Imagine that an employee needs to write a polite email asking a customer for additional information.
Instead of starting from a blank page, the employee could describe the situation to an AI assistant.
The AI could produce a first draft.
The employee can then review the draft, correct any errors, and make sure that it accurately represents the situation.
Summarization
Generative AI can also help reduce large amounts of information into a shorter form.
For example, an AI system might summarize a long document, meeting transcript, article or collection of notes.
This can save time when a person first needs to understand the main points.
However, summaries can leave out important details, so the original material may still need to be consulted.
Education
Generative AI can act as a learning assistant.
A student can ask for:
- A simple explanation of a difficult concept.
- Examples of a topic.
- Practice questions.
- A quiz.
- A summary of study material.
- An alternative explanation of something they did not understand.
This can make learning more interactive.
Personalized Explanations
Different learners need different explanations.
A student who understands basic mathematics may want a more technical explanation of machine learning.
A complete beginner may need an everyday analogy.
Generative AI can adapt its explanation based on the instructions supplied by the user.
Research Assistance
AI can help researchers and knowledge workers organize information, summarize documents, brainstorm questions, and identify areas that may deserve further investigation.
When an AI system is connected to reliable external sources, it can also help retrieve and organize current information.
However, AI-generated research should not automatically be treated as authoritative evidence.
Business Applications
Businesses can use Generative AI for many routine knowledge tasks.
Examples include:
- Drafting business documents.
- Summarizing meetings.
- Preparing customer responses.
- Creating marketing ideas.
- Analyzing and explaining documents.
- Generating internal documentation.
- Brainstorming product ideas.
Customer Service
Generative AI can assist customer-service teams by helping draft responses to common questions.
An AI system can receive a customer's question and generate a suggested response based on company information supplied to it.
A human employee can review the response before sending it, particularly when the request involves an unusual or sensitive situation.
Marketing
Marketing teams can use Generative AI for brainstorming and drafting.
For example, AI can help generate:
- Campaign ideas
- Product descriptions
- Advertising variations
- Blog outlines
- Social media ideas
- Audience-specific messaging
Human review remains important because marketing content represents a brand and must be accurate and appropriate.
Software Development
Generative AI has become an important assistant for software developers.
AI coding tools can help with:
- Writing functions.
- Explaining existing code.
- Finding possible bugs.
- Generating tests.
- Writing documentation.
- Converting code between languages.
- Creating database queries.
AI Does Not Replace Testing
A developer should not assume that generated code is automatically correct.
Generated code needs to be reviewed, tested, and evaluated for security and performance.
This is especially important when the code will be used in a production system.
Creative Work
Generative AI can also support creative activities.
Writers can brainstorm stories.
Designers can explore visual concepts.
Musicians can experiment with ideas.
Video creators can develop scripts and concepts.
In these situations, AI can be used as a creative partner or brainstorming tool.
Document Analysis
AI systems can help people work with large collections of documents.
An application might allow a user to upload documents and ask questions about their contents.
The system can retrieve relevant information and use a language model to explain it.
This approach can be particularly useful when employees need to work with policies, manuals, reports or other large collections of information.
Internal Knowledge Assistants
A company can build an AI assistant that works with its own internal information.
For example, an employee could ask:
"What is our procedure for handling a customer complaint?"
The application could retrieve the relevant company policy and provide it to the AI model as context.
The model can then generate an easy-to-understand response.
Productivity
Many Generative AI use cases are ultimately about reducing the time required for routine knowledge work.
AI can help people start faster, organize information, create drafts, and explore alternatives.
The biggest benefit is not necessarily replacing an entire job.
Often, the benefit comes from helping a person complete parts of a task more efficiently.
Accessibility
Generative AI can also support accessibility.
Examples include generating descriptions of visual content, simplifying complex text, converting information between formats, and assisting with speech or language.
These applications can help people interact with information in ways that better suit their needs.
Personal Productivity
Individuals can use Generative AI for everyday tasks.
Examples include:
- Planning a trip.
- Creating a study schedule.
- Organizing notes.
- Brainstorming ideas.
- Writing a first draft.
- Explaining a technical concept.
Where AI Is Less Suitable
Generative AI is not automatically the best tool for every problem.
A simple calculator is usually better for straightforward arithmetic.
A reliable database is better for retrieving an exact stored record.
A specialist professional may be necessary when decisions have serious consequences.
The best solution often combines AI with other tools rather than using AI for everything.
AI Plus Traditional Software
Modern applications can combine Generative AI with traditional software.
For example, an application might use:
- A database to store information.
- Traditional software to apply business rules.
- An API to retrieve current information.
- A language model to explain the results to the user.
This combination can be more reliable than expecting the AI model to perform every part of the task.
Human-in-the-Loop
A human-in-the-loop system includes people in important parts of the AI workflow.
For example, an AI model may generate a suggested customer response, but an employee reviews and approves it before it is sent.
This approach can help reduce the impact of AI errors.
Measuring Success
Using AI is not automatically beneficial simply because the technology works.
An organization should consider whether the AI application actually improves the task.
Useful measurements might include:
- Time saved.
- Accuracy.
- Cost.
- Customer satisfaction.
- Employee productivity.
- Error rates.
The Importance of Good Data
Many useful AI applications depend on reliable information.
If an AI assistant is expected to answer questions about company policies, the underlying documents should be accurate and up to date.
Better information generally provides a better foundation for the AI application.
The Importance of Good Instructions
The quality of the prompt can also affect the usefulness of the output.
A clear request with appropriate context usually gives the model more information about the desired result.
We will explore this topic much more deeply in the Prompt Engineering module later in the course.
AI Use Cases Are Expanding
Generative AI is still developing rapidly.
New applications continue to appear as models become more capable and easier to integrate into software.
The most useful applications will depend on the specific needs of users and organizations.
The Big Picture
Generative AI is best understood as a flexible technology that can assist with many different tasks.
It can help people create, summarize, explain, analyze, translate, brainstorm and transform information.
However, useful AI systems usually combine the model with good data, appropriate software, clear instructions, safeguards and human judgment.
What You Should Remember
- Generative AI can assist with writing, communication, education, research and business tasks.
- It can help developers write, explain, test and document code.
- AI can support creative work and brainstorming.
- AI applications can work with company documents and internal knowledge.
- Traditional software, databases and external APIs can complement AI models.
- Human review is important when errors could have meaningful consequences.
- The best AI solution is not always the one that uses AI for every part of a task.
- Good data, good instructions and appropriate evaluation are important for successful AI applications.
What Comes Next?
We have now seen some of the many ways Generative AI can be used.
In the next lesson, we will look more closely at how Generative AI can affect education and work, including both opportunities and practical challenges.