Generative AI for everyone
Generative AI is a powerful, multi-purpose tool for the workplace. Learn the basics and start experimenting in your job.
No time to read? Get answers to your generative AI questions by asking our GenAI Helper bot.
What is generative AI?
Generative AI (GenAI) refers to a category of AI designed to generate content, data, or other outputs that are similar to human-created content.
GenAI tools have the capability to generate new, original data or content, often in the form of text, images, audio, or even video.
Generative AI can be used for tasks such as creating human-like text, generating artwork, and completing code.
How does generative AI differ from traditional AI?
Generative AI specializes in creating new content, such as text, images, or music, based on patterns it learns from large datasets.
Traditional AI, in contrast, focuses on analyzing data and making decisions based on predefined rules or algorithms.
Generative AI is more adaptable and creative, excelling in tasks that require innovation and a human-like understanding of context.
Traditional AI is often more efficient and reliable for structured tasks, like data processing or logical problem-solving, where clear rules can be applied.
How does generative AI work?
-
It learns from massive amounts of data.
Generative AI starts its journey by learning from a massive library of information. This library includes books, articles, images, videos, and more. It's like a student going through an extensive course of study, absorbing knowledge on a wide array of topics.
-
It looks for patterns as it learns.
As it learns, the model begins to understand how things are related. It uses probabilistic and statistical principles to notice patterns, like how sentences are structured or what makes an image appealing. This stage is akin to the artist practicing and understanding different styles and techniques.
-
It generates new content and adapts to feedback.
After learning and understanding, the model can create new content. It can also adapt to feedback. If you tell it to change a part of what it created, it can do so, learning from your preferences to make something closer to what you want.
Go deeper.
Explore this visual explainer on how ChatGPT works. ChatGPT is one of the most popular generative AI tools.
What is generative AI good at?
Summarizing information
Condense lengthy reports, emails, or documents into digestible summaries.
Distill complex research papers into key findings and implications, extracting main points from scientific articles for non-experts.
Data analysis
Generative AI can identify trends and patterns in small datasets, offering insights that guide strategic planning.
Example: Analyzing consumer behavior patterns from sales data or sentiment in customer feedbac,k.
Transforming information
Convert data from one format to another, like turning spreadsheets into visual presentations.
Translate technical jargon into layman’s terms, making information accessible to a wider audience.
Explaining and teaching
Create personalized learning materials,, generate custom curriculums for online courses.
Experimenting with explaining complex concepts in simple terms, aiding in training and development. Example: Breaking down new software features for non-technical staff.
Generating new content
Create marketing materials, from ad copy to social media posts. Go further with creative content like stories, videos, audio, based on specific prompts or themes. Learning to code? Use generative AI to generate code and troubleshoot errors.
Organizing information
Categorize emails and documents, improving workflow efficiency. Example: Sorting customer feedback into relevant categories for action.
Generate summaries and indexes for large collections of documents, making them easier to navigate.
Explore generative Ai use cases for your industry.
Dive deeper into how to use generative AI in your industry with Deloitte’s interactive use case tool.
What are some generative AI tools used in workplace software?
-
Canva
Canva is using generative AI to generate copy, images, and create digital avatars for training videos and presentations.
-
Adobe
Upload documents to Adobe, get summaries of your PDF, and ask questions about it.
-
Microsoft
Use Microsoft Copilot to create a first draft, inspire ideas, transform your content and more.
-
Slack
Slack offers personalized search responses, key highlights from channels, and one-click summaries of lengthy conversations.
-
Monday
Monday, a project management software, suggests complex formulas that best fit your use case.
-
Zoom
Zoom AI Companion can draft emails and chat messages, summarize meetings and chat threads, improve brainstorming with colleagues.
What are generative AI tools I can experiment with?
ChatGPT
The product that introduced the masses to generative AI. Built by OpenAI, ChatGPT can answer your questions, write copy, draft emails, and more.
MidJourney
Midjourney generates images from text instructions (prompts) Similar products includes DALL-E and Stable Diffusion.
Beautiful AI
Create presentations, brainstorm ideas, and generate text and images.
Writer
An AI-powered writing assistant that can create blog posts, marketing content, and more while adhering to your brand guidelines.
Perplexity
A search chatbot using generative AI that offers exact and complete responses to user inquiries.
Pi
A personal AI, designed to be supportive, smart, and there for you anytime. Ask it for advice and answers.
Runway
Runway is like a new video editor. It's a complete AI-powered video production studio, capible of creating videos from text.
Gemini
Similar to ChatGPT, but using a different model, Gemini is a chatbot that can create, brainstorm ideas, and search the internet.
Lensa
An all-in-one image editing app that allows anyone with no professional experience or editing skills to get great results.
Prompt writing structure from Writer.ai guide to prompting.
How can I get better at prompting?
-
Find prompt libraries and use templates. Customize the templates for your use case.
Prompt library from Claude -
Play with the prompts. Prompts don’t have to be written perfectly. When in doubt, ask a chatbot like ChatGPT to write a prompt for you by describing the problem you are trying to solve.
-
Your first output won’t be great. Keep trying by giving the model as much context as possible: who, what, how, when, why
What is generative AI not good at?
-
Large Language Models, the technology that powers ChatGPT and other popular text generators, does not product accurate results consistently. Often call “hallucinations,” generative AI products with LLMs, do not always provide correct answers. However, the products like ChatGPT will not tell you incorrect - it will simply provide an incorrect answer with confidence. The result is that it is challenging to know when a response is innacurate.
-
-
Generative AI models are intricate. They makes decisions based on a lot of hidden steps and rules that aren't easily visible or explained, making it hard to understand how it reaches its conclusions
Bias and generative AI
Generative AI amplifies bias and stereotypes found in the training data.
Will generative AI take my job?
The answer is complicated. Read on to learn the nuances.
Additional reading
A collection of resources to keep updated on generative AI and its impact.
Online course for generative AI
Sign up to be the first to learn when our new course, Generative AI for Everyone, launches in May 2024.