Speakers:

Daniel Braun, Professor at University of Twente

Letitia Parcalabescu, PhD Researcher at Heidelberg University, Youtuber

David Stutz, Scientist at Google DeepMind

Moderation: Anne Schreiter

»At conferences we should not only talk about WHAT our research is about, but also HOW we do research.« – David Stutz

Key Messages:

There are tons of new tools out there and more and more are popping up.

AI applications are a complex and ever-changing market, even old tools come with AI extensions, e.g., PPTX, Photoshop etc.

Most tools are not trained on scientific data/texts and research seems to be not as profitable yet for companies (e.g. Mendeley has an outdated interface whereas tools for planning your vacation are state of the art).

The tool is secondary – it’s more important to know what you are using it for and why. For instance, there are many calendar tools out there, but you need to know how to manage your time and deal with your tasks.

In the future, AI tools might help to speed up scientific discoveries. However, it’s important to understand their principles in order for them to be co-pilots, not the pilot.

»Embrace life-long learning and expect to be on a journey.« – Daniel Braun

Criteria for skills of researchers might change, e.g., the importance of good writing skills. Teaching and assessing competencies will change.

AI will not replace humans anytime soon; keep improving your critical thinking skills, your fact-checking ability, and a flexible mindset.

Helpful Tools and Tipps:

→ When you don’t know which tools to use:

• Figure out what task you spend the most time on (writing, coding, proofreading,…) Choose the areas that most likely benefit from automation. Focus on those, ignore the rest.
• Talk to colleagues about what they find helpful – every field and discipline is different, that is how you find out relevant tools.

→  Helpful non-AI tools:

• Trello or other Kanban tools for organizing your work flow
• Calendly for scheduling meetings
• Notion as a workspace

→  If you are concerned about data security:

• Download/use applications on your local device.
• Legally speaking, don’t put any personal data in any app – that’s true for any other sensitive data. However, you have to weigh in what that means practically for you.

→  Pitfalls:

• Don’t get fooled by advice that promises to “power charge” your PhD by just using the right prompts – you might have the illusion to write the perfect abstract, but you haven’t really understood the point.
• Try to teach concepts and how to critically assess them, don’t try to ignore the tools (on-premise exams because of ChatGPT are not a good idea). 

»It’s easy to feel overwhelmed, but you are doing just fine. Remember that you are a scientist because you can do science, not because you know how to best use ChatGPT.« – Letitia Parcalabescu