Special Issue - How language AI can transform project management and boost efficiency and creativity

Newsletter

AI-driven language tools may become valuable assistants in project management routines. The article suggests a number of use cases as well as examples of prompts to try out.

Project management is a challenging task that requires managing international collaborations, handling multiple deadlines, maintaining communication channels, making data-driven decisions, and being adaptable to unexpected situations, such as budget constraints and changes in plans.

If you have been keeping an eye on the tech scene, you have probably seen headlines like, "The Top XX Mind-Blowing Applications of ChatGPT", "215+ Can't-Miss ChatGPT Prompts," or “Top 10 Most Insane Things ChatGPT Has Done.” To stay ahead in the game, project managers need to keep themselves updated with the latest technological advancements. Large Language Models (LLMs) like ChatGPT, Bard, and others can be game-changers in this regard. By incorporating these AI tools into their toolkit, project managers can revolutionize their workflow and simplify their tasks. In this article, we will explore some practical examples of how ChatGPT and similar tools can be used to streamline project management tasks.

Jonas Krebs,

a personal account

Jonas Krebs currently serves as the scientific project manager and coordinator of the projects area at the "Strategy and Funding" office in the Centre for Genomic Regulation (CRG) located in Barcelona. He obtained his PhD in plant molecular biology in 2010 from the University of Potsdam, Germany, and has since held various roles in project development, research management, internationalization, and professional networking at different institutions. In 2014, he focused on European research funding and earned the certificate of "EU-Liaison officer" from the German Federal Ministry of Education and Research. Since 2015, Jonas has been responsible for developing and managing EU projects at CRG, with a primary emphasis on Marie Skłodowska Curie Actions. He is an active member of the Marie Curie Alumni Association (MCAA) and co-founded and currently chairs its Research Management Working Group.

Image by Midjourney, prompt by Fabrizio Martina
Image by Midjourney, prompt by Fabrizio Martina

Say Hello to Crystal Clear Communication

Effective communication is at the heart of successful project management. LLMs can be your ally in crafting direct, easy-to-understand messages, emails, and reports. By leveraging their capabilities, we can sidestep lengthy emails, cryptic jargon, and misinterpretation. Plus, they're an excellent tool for brushing up our English language skills, especially for those of us who are not native speakers.

Example prompts:
• "Summarize [text] in exactly xx words"
• "Shorten this email by 30%”
• "Rephrase this text to make it [more casual/first person/humorous]"

Banish Writer's Block Forever

Chances are, you have experienced this all too familiar scenario: you would like to draft an initial piece of text - be it a tweet, a blog post, or a script for a dissemination video. You stare at the blank page, unsure of how or where to begin. But fear not! LLMs are here to banish your block. Just provide a few key points and some context, and you can get a suggested structure as a starting point. Even if it is not perfect, it’s much simpler to revise such a skeleton than to create it from scratch. At least, that has been my experience.

Example prompts:
• "Address this heading/question by incorporating the following points"
• "Turn this text into a tweet/LinkedIn post"
• "Suggest a (video) script/Twitter campaign about … and highlight….”• “How can I structure a section about …"

Expert Tip: If you find that the output from an LLM doesn't quite match the tone or style you were aiming for, don't worry! There are several ways to address this. For instance, you can copy and paste your own previous text samples that you liked and ask the LLM to analyze the tone and style. Then, use the received attributes as additional instructions for your next prompt. Another option is to use any other text as input to mimic a particular style, such as that of your favorite writer, journalist, or scientist.

Ace Those Difficult Conversations

As a project manager, you're tasked with managing challenging discussions among consortium members or colleagues. Addressing performance problems, disputes over intellectual property rights, or budgetary concerns can be awkward and tough to handle. With only a few clicks, your AI-powered LLMs can quickly provide a personalized blueprint of what to say, role-play the conversation, and even identify potential compliance issues.

Example prompts:
• "How can I address a consortium member who constantly misses deadlines without causing resentment?"
• "Help me script a conversation where I have to inform the project officer about a delay in our project delivery"
• "How can I approach a conversation with upper management about the need for more resources for our project?

Unleash Your Creativity

ChatGPT can be a great asset to project managers when it comes to brainstorming tasks. By providing alternative perspectives and fresh ideas, LLMs can help project managers to overcome unexpected challenges and roadblocks. LLMs can also suggest new activities and approaches based on data and past experiences, giving project managers access to a wider range of potential solutions

Example prompts:
• "How can I actively involve our Early Stage Researchers in…"
• "How can I structure an interactive 2-day workshop on…"
• "How can our project… promote its achievements to the public in a comprehensive and engaging manner?”

Excel in Excel

As a project manager, I frequently find myself reaching the limits of my Excel skills, whether it's creating budget forecasts, sorting survey data, or other tasks. This is where LLMs can be a game-changer. LLMs are capable of understanding natural language queries and can provide appropriate Excel formulas in response. By simply asking the LLM for the formula needed to accomplish a specific task, the model can quickly provide the correct formula or function, saving time and reducing the likelihood of errors.

Image by Midjourney, prompt by Fabrizio Martina
Image by Midjourney, prompt by Fabrizio Martina

Example prompts:
• "How can I extract from column A… and show in column B only…"
• "Which formula allows me to compare/ identify/ forecast…”
• "Suggest tips for using Excel's conditional formatting to better manage and monitor project expenses."

Integration of LLMs into other Apps

LLM-powered chatbots are already incredibly powerful, but in this last point, I want to showcase two examples that take LLM integration to the next level. These apps combine LLMs with audio and virtual reality (VR) technology, demonstrating the full potential of these tools.

Read AI:
• Read AI is a dashboard for virtual meetings that leverages AI and LLMs to document the meeting and measure engagement, performance, and sentiment among participants.”
• it creates a full transcript of your virtual meeting, but also key questions that were discussed, a summary with action points, and an engagement report of attendees.

We had a first pilot test of the tool in our working group coordination team, and such Apps have a high potential to safeguard you from the annoying task of writing the minutes and/or summary reports.

VirtualSpeech:
• VirtualSpeech is an AI-powered training tool for improving communication skills.
• In short: you can put on your VR headset and can practice conversation with ChatGPT-enhanced avatars, amongst other functionalities.

I discovered the tool in my role as the project manager of the ITN PROTrEIN and currently explore how we can use it on the consortium level. We meet already for more than a year with all ESRs in virtual reality-assisted tools and platforms.

Conclusion

I hope this article provided you with some fresh ideas on how AI-driven LLM tools can assist project managers. ChatGPT provided helpful advice in writing this article. In our Research Management Working Group, we had so far, no systematic exchange on the experiences and potential of LLMs in our daily work. But it is definitely about time to facilitate such a discussion. We have established a great platform for it.

It is important to acknowledge that the constant influx of new tools and advancements in LLM models and algorithms also raises ethical concerns, which are addressed by other articles in this newsletter. I very interestedly follow the debate on “intellectual input” vs. pure “content creation”. In a research funding system that is mostly based on grant writing, I sincerely wonder whether LLM-generated proposals become the norm, and if so, how this will impact the current model of funding allocation. The rapid development of LLMs highlights the need to re-evaluate existing research funding models and assess if they're still appropriate. Might we soon see a competition between AI text generation and detection tools that aim at identifying syntactically generated text?

Jonas Krebs
Centre for Genomic Regulation
jonas.krebs@crg.eu
Twitter @_JonasKrebs