Matthias Karner

Digitalization & IT leader. Avid reader. Dedicated Writer. Engineering PhD. Productivity enthusiast.

9 minute read

If you’re like me - constantly looking for ways to work smarter, not just harder - then you’ve probably heard about generative AI.

It’s everywhere right now, and for good reason. AI tools like ChatGPT, DeepSeek, Microsoft Copilot, Notion AI, and others are transforming how we approach work, making it faster, more efficient, and sometimes even more enjoyable.

But how exactly does generative AI boost productivity? And how can you use it to make your workday smoother? Let’s dive in.

image
What AI thinks cooperation with it looks like. Image created in cooperation with MS Copilot

Let’s make this more fun with a small quiz

Generative AI is great at writing texts. I like to use it to enhance my writing, create outlines, and other things. However, I don’t use texts solely created by AI—in my opinion, they lack personality and originality. However, for this article, I made an exception, and I’ll mix in some paragraphs solely created by AI: can you guess which ones?

What is generative AI?

Please, feel free to skip this section if you already know what generative artificial intelligence is.

Generative AI is a type of artificial intelligence that can create new content - like text, images, music, and even code - based on patterns it has learned from existing data. Instead of just analyzing information, it generates something new, making it a powerful tool for creativity and productivity.

  • Text AI (ChatGPT, …): Writes articles, emails, and summaries.
  • Image AI (MidJourney, …): Creates pictures and designs.
  • Music AI (AIVA, …): Composes music.
  • Code AI (GitHub Copilot, …): Assists in writing and debugging code.

7 tasks where I use generative AI

1. Summarize texts

Working with text is where generative AI shines. Models like GPT-4 were trained with billions of lines of text and they’ve learned how to work with text very well.

My very first use-cases where I applied generative AI tools were summarizing longer texts (articles, manuals, books) to quickly get to the essential information.

And it works very well.

I use it regularly now and it has become a part of my workflow.

The longer the texts are, the better it works. For smaller manuals, for example, a full-text search can also work well. However, if you are searching for more complex topics (which you cannot describe with simple keywords), generative AI has the big advantage that it understands the context and semantics of what you are searching for.

What I also really like about summarizing text with generative AI is that you can guide how the summary should look like: whether it is a 3-sentence summary or a quick 10-step guide - generative AI has you covered.

2. Find information quicker

Beyond single documents, generative AI can be a search engine on steroids.

This is true for web searches (be careful with the sources though), but is also very helpful for document search. If you are using a system that is connected to (all) your documents, it makes retrieving information, or finding the right documents much (much, much) easier.

Again the context-based search kicks in here: it doesn’t just search for keywords, but for information that is (or could be) relevant to what you are actually searching for.

3. Knowledge synthesis

This is one of my favorite use cases of generative AI as it combines its capability of processing a lot of text and its context awareness.

Let’s take an example from research: every research project needs some kind of literature analysis.

Looking back at my time as a PhD student, I remember reading hundreds of papers to put together a summary of the information I gathered. If I had generative AI back then, it would have made my work so much easier and faster!

With generative AI you feed the artificial intelligence algorithm several papers to process. After tokenizing the information you can make use of the chat functionality to make queries on the now internalized information and go as far as make the AI create a comparative literature analysis within seconds for you.

It works like a breeze.

Fine, it doesn’t replace the need to read and comprehend key papers yourself, but it allows you to cover a broader range of relevant literature much more efficiently.

But knowledge synthesis with generative AI isn’t just interesting for research. Think of project work for example: How often do you work on topics/projects / … where something similar has already been done before?

Synthesizing the available information (summarizing best practices and lessons learned) beforehand can save a lot of time and energy during project execution.

However, without well-curated knowledge-bases, it can be very cumbersome to unearth the findings from past projects. Generative AI makes this task much faster.

4. Creating images

This is a use-case I am not using that often, but sometimes it’s a great addition to visualize information for a presentation or an article. I even drew some AI caricatures for some of my colleagues’ birthdays.

If you work a lot with graphics design, generative AI might be a great addition to your toolset. I am sure it won’t replace the act of creativity - I’d consider it more of a tool to use for brainstorming with or to create first drafts.

5. Automating repetitive tasks

We all have those small, time-consuming tasks that add up: writing emails, summarizing meetings, generating reports.

AI can handle these in seconds.

Instead of spending 10+ minutes crafting the perfect email, you can use AI to draft one instantly and then refine it to add your personal touch.

For automating tasks, AI tools that are integrated within the software you are using anyway are great. This is why for MS Office users MS Copilot can be a mighty tool. It still has its flaws as I found out in my first tests (it doesn’t work that well if you don’t use it directly from the office app you want it to work in), but if used properly it drastically speeds up work with Office apps.

6. Accelerate text and document creation

Generative AI is not only great at creating summaries, it can create any form of text you ask it to.

If you have an AI tool like MS Copilot that is built into your applications, it can even support you in formatting your documents, creating images, or structuring text-based information.

As I highlighted in (1) I don’t like to create all my text with generative AI as it feels stale and lacks personality. However, if you are using your own information as input and enhance the text after its creation, the output can be quite good.

At work, I use the following process for creating documents like protocols, or official text.

  1. Collect notes (manually).
  2. Process notes to create a document draft (generative AI).
  3. Read and enhance the draft (manually)

I’d estimate that the use of generative AI saves up to 75 % of the time needed for creating documents.

7. Creating software code

Generative AI also is good at creating software code. As code is also text-based it has even become a core-competence of text-based-generative AI. Tools like GitHub Copilot or Codeium (which is free) are purpose-built AI models to enhance software development.

But even ChatGPT can support you in creating software code. I did a small test and asked ChatGPT to program a JavaScript version of Snake. The first code didn’t work, but just with one additional prompt, I ended up with a version that actually worked.

That impressed me.

I mean: yes you can just google it and find working source code within a bunch of minutes, but the example shows how potent these language models are. If you can describe what a function (part of a software program) should do, generative AI can create the code for you - in whatever programming language you want.

However, it still has its flaws and the code should be reviewed before it is put into production. Otherwise, it may result in a lot of debugging.

4 tasks where I avoid generative AI

Generative AI can be a supercharger for productivity. However, it isn’t a solution to all productivity needs and sometimes it is faster / easier / better to do things without generative AI.

1. Reading the news

News and media are areas that are critical even without a generative AI tool.

Why?

Unfortunately, not all media houses are independent, and finding truly objective news can be difficult. Therefore, if I want to inform myself, I go to platforms I trust.

I don’t completely avoid generative AI here. It can definitely be handy to summarize longer news articles. However, if I do, I provide the sources to the AI model manually.

2. Handling confidential information

While AI can be a powerful productivity tool, it’s not the best choice for handling sensitive or confidential information. Here’s why:

  1. Data privacy risks: Most AI models, especially cloud-based ones, process data externally, meaning your input is sent to remote servers. If you’re working with personal data, business secrets, or confidential documents, there’s always a risk that this information could be stored, accessed, or even used to further train AI models.
  2. Compliance and legal issues: Many industries have strict data protection regulations (e.g., GDPR in Europe, HIPAA for healthcare in the U.S.). Using AI to process customer or employee information may violate these regulations if the tool doesn’t meet the required security standards.
  3. Lack of security guarantees: AI tools do not provide the same level of security as encrypted, enterprise-grade systems. If an AI service is hacked or compromised, any sensitive data submitted could be at risk. Additionally, many free AI tools do not guarantee data deletion after use.

3. Starting my research on a new topic

If I research something new, I don’t directly go to a generative AI solution as I don’t want to start with information that might be just a hallucination.

Instead, I usually start with some trustworthy sources - depending on the research I am doing this might be, for example for scientific research Scopus or Springer Link, or just Google (and looking for trustworthy sources there) for other kind of research.

After I’ve made myself familiar with some basics, I’d start adding in some generative AI tools to help e.g. with summarizing texts and speeding up my research.

4. Making decisions

AI can analyze data, but it can’t replace ethical decision-making.

Generative AI can help gather and organize information that is essential for making informed decisions. And I used it for that. However, I never delegate decision-making to AI.

Conclusion and further thoughts

Generative AI can boost productivity to a level I wouldn’t have imagined five years ago. It streamlines text-related tasks and offers support across various domains.

It’s here to stay, and sooner or later, knowledge workers worldwide will adopt it - because without it, keeping up will be a challenge.

The caveat?

The skillset we humans need will change in the future. Repetitive tasks will slowly fade out, but there is one particular skill that will become increasingly important: critical thinking.

Generative AI can hallucinate or just use wrong/untrustworthy sources. If you blindly trust it, it has the potential to misguide you. However, if you use it with a good sense of critical thinking it can extend your capabilities - a lot!

Thank you for reading! What do you think about generative AI? Where are you using it and when do you try to avoid it? Let me know on LinkedIn or Medium!

Quiz solution

So which parts of the text were written by AI?

  • The introduction is created completely by AI. I just added AI tools to the list.
  • All text (except the first sentence) in “What is generative AI?”.
  • The first paragraph of “Automating Repetitive Tasks” is 100 % AI-generated.
  • The section “Handling confidential information” is generated by AI.
  • The section “Making decisions” was partially created by AI.
  • The first two sentences in the conclusion are “optimized” by generative AI.

You may also enjoy

Good Power: Leading with Purpose

1 minute read

When you hear the word “power,” what comes to mind? Many people think of control, authority, or even fear. But power doesn’t have to be about dominance - it...

A Deep Dive into the 12 Week Year

6 minute read

As a productivity enthusiast, I’m always on the hunt for new methods and tools to try. Recently, while watching yet another Ali Abdaal video, he mentioned a...