How Generative AI is Revolutionizing Our Work

Share on Social Media

Since November 2022, you have probably heard the term artificial intelligence (AI) mentioned more than ever before. Google searches for it have jumped 400% since then. That's due to the development of a new kind of AI technology known as generative AI that has already begun revolutionizing the way we work. What is it, how does it differ from the technology that went before it, and what does it mean for you?

For years, AI has analyzed vast amounts of existing data and found patterns in it. It used these patterns to create a kind of digital reference manual that it could use to recognize things about new data. That reference manual was called a training model.

AI created training models for narrowly defined tasks. One model might focus on spotting damaged doors in pictures of used cars. Another might be good at understanding natural language queries. An AI model trained using audio files might be able to transcribe telephone conversations into text. This AI has been useful, but it only ever classified existing data. It would sort incoming information rather than producing any new information of its own.


What generative AI does that traditional AI can’t

In 2017, Google researchers pushed the boundaries of the possible. They published an academic paper called Attention is All You Need that proposed a new training method for AI. Known as the transformer model, it gave AI a far more detailed understanding of our existing data by analyzing it in new ways. The transformer model was the basis for generative AI, which went beyond traditional AI. A trained generative AI data model, known as a large language model (LLM), doesn't just interpret existing data. It also uses its enhanced understanding of that data to generate new things. For the first time ever, we have access to a creative form of AI.

two hands holding a holographic globe


Creating brand new information

Generative AI still uses machine learning and deep learning techniques, but the underlying training algorithm can now understand text or voice input and use it as a prompt to produce something else. You can ask it "suggest some nature activities for the kindergarten class that I am teaching tomorrow" and it will create some bullet points. Ask it to write you a brief history of Norway in the style of Shakespeare and it will happily oblige.

How ChatGPT changed the course of AI development

One of the earliest generative AI companies was OpenAI, which uses techniques based on the transformer model to produce its Generative Pre-trained Transformer (GPT) service. It used text found on the Internet to train a GPT model, expanding the volume of training data to produce more sophisticated versions.

GPT-3 goes public

OpenAI opened up commercial access to GPT in 2020 with its third version, GPT-3. This impressed people with its ability to not only understand language, but to produce convincing text based on external prompts. Its successors, GPT-3.5 and GPT 4, can match human performance for many tasks. They are also behind the publicly accessible ChatGPT service launched in November 2022.

How your inputs make generative AI smarter

Making generative AI services available to the broader public also allowed OpenAI to train the models further based on simple user feedback. That concept, known as Reinforcement Learning via Human Feedback (RHLF), refines ChatGPT’s responses to these diverse queries, making the service’s responses more accurate.

More than just text

AI scientists realized that the transformer model could analyze other kinds of data beyond text. That enables it to produce new content in other formats:

  • Images (Dall-E) · Audio (Boomy for music, and Lyrebird for overdubbing and podcast presenting)
  • Computer code (Codota for code completion, SourceAI for code review)
  • Video (Synthesia and Elai for presentations by AI avatars)

Companies are already using the technology to illustrate articles and assist with software development.

Generative AI has now made its way into many products, including enterprise tools such as chatbots and business intelligence software, along with consumer productivity software such as Microsoft's Windows operating system and Office applications. For many people, it is already making working life easier.

man holding a tablet

How generative AI can improve EWS

AI's abilities to streamline work and improve productivity have particular potential in enterprise workplace services (EWS). These are the support services that keep everyday office operations running, ranging from managing a busy reception desk through to organizing mail, creating presentations, and keeping records.

As generative AI solutions improve, they continue to offer more possibilities for automating and enhancing EWS. Some of the possibilities include:

Enhanced research

Searching through multiple files and databases to extract salient points is a time-consuming task, especially in organizations with large amounts of data. Generative AI's advanced summarization capabilities, combined with new, more advanced databases, can expedite this process, helping your information management teams to deliver the right information. AI engines can not only effectively comb through large amounts of data in multiple formats, including written documents, web sites, photos, PDFs, and videos, but they also allow users to request the information conversationally, like they would if they were asking a colleague.

Content creation

Generative AI's ability to produce new content makes it a great tool for quickly creating everyday business documents such as email or web copy. Users in marketing and sales can use it to produce basic documents and images that they can then tweak to suit their unique requirements, saving them hours of work.

Content repurposing

Generative AI excels at formatting content for different channels and platforms. Having helped to create a text asset, it can quickly tweak it into an appropriate format for channels such as emails, social media, and blog posts, expediting digital design services.

Providing feedback on content

Generative AI services can analyze content - especially text - and provide feedback on how to improve its quality. Consider these services as copilots that make suggestions as you work.

Fast data analysis

The days of using complex queries to pull insights from business intelligence systems are ending. Instead, natural language interfaces enable you to have conversations with your systems and data. You can ask a database questions about customer segmentation, refining the results incrementally using a conversational interface. Microsoft’s Copilot in Excel lets you ask natural language questions like ‘what happens to our sales if we discontinue this product?’ or ‘plot product sales by category over time’. This digital assistant capability produces valuable new insights quickly.

This exciting new form of AI technology excels in automating tasks, not full jobs, and doesn’t replace the need for human intelligence and creativity in an organization. Its use highlights the importance of critical thinking skills, as it is required to validate generative AI's output. Nevertheless, it carries great promise as a tool to augment employee capabilities and make work easier.


Generative AI is still in the early stages of development, currently sitting roughly where computing was in 1982, just after the launch of the personal computer (PC). As with the PC, the potential is enormous. McKinsey expects the technology to generate up to $7.9 trillion in value per year.

It is difficult to know exactly what generative AI will look like in 20 years, but one thing is certain: it will constantly improve and become embedded in more products and services. As it does so, it will make our jobs increasingly fulfilling, enabling us to concentrate on more meaningful aspects of our work while it takes care of the small stuff.


Placeholder image

Use-cases for generative AI

As explored in the first post in this series, generative artificial intelligence (AI) is a ground-breaking technology that uses large language models (LLM) to create...

Placeholder image

How can generative AI can transform productivity?

The personal computer (PC) was invented in the 1980’s. However, it was not until the late 1990’s, when unit costs had dropped and PCs started being connected to the internet, that this breakthrough started having a noticeable impact on productivity figures.

Placeholder image

Use Generative AI to talk with your data

If data is the new oil, then the key to unlocking its value lies in refining it. Without effective management and analysis, the insights in your data will go unrealized.