AI's Role in Shaping Careers in Data Science: A Conversation with SAP's Chief Data Scientist

Yeah, so let's talk about Gen AI a bit more.

So as we've kind of discussed or touched on, generative AI is a part of everyday life and work.

What are some practical ways you think that people could get the most out of these tools and Are there any prompt writing tips or habits that you feel could make a real difference?

Yeah, it's really something that's hit mainstream.

I think everyone is now aware and is using tools like ChatGPT or Gemini.

It seems that AI agents and AI tools are popping up pretty much everywhere in all kinds of products.

So I think it's really been something from maybe experts in research working with AI to something that everyone just uses, like everyone is using Google and other search engines.

So as the technology evolves, I think also the way and how easy it is to use it has been shifting very quickly.

I think in the beginning, it was still quite difficult to make these AI models do what you want.

So there was a lot of emphasis on prompt engineering and how do you have to write it in the way that the model does.

But they get better very quickly.

So now I think this is less important if you can just naturally express what you want to one of these AI tools.

And they often get it right.

But still, it's important to provide the right context.

I think it's important for people to understand these are still tools and machines in the end.

So if you don't provide the right context, they might not necessarily give you what you were expecting, right? So this prompt, basically what you give as input to the model, should hopefully provide all the necessary information, maybe the constraints, what the model, what the AI should do, should not do.

And the users are basically endless now, right? So people use it for, I don't know, suggesting some recipes or things to do on the weekend and their personal life to very highly professional tasks, for example, in writing software.

One of the jobs that is then changing maybe the most now is our own jobs as people who write code and write software.

I think all of us are using AI agents now, and this has really been accelerating how quickly people can write and prototype software.

But the other thing is, I think we're all still learning along this journey as the technology changes very quickly.

So in the ways it's kind of level sets everyone a little bit to label a playing field, because everyone now, for example, has to figure out how to use AI in software engineering, because the processes that we used to have established maybe now change dramatically as smaller teams can work much more rapidly with the help of AI.

So I think my best practical advice that I can give is really to be very hands-on and try these things out because things are moving very quickly, things are changing very quickly, and nobody really now has all the answers.

So the best way is often just to try out maybe some problems that you used to have and say, hey, what if I use one of these AI tools, whether it's ChatGPT or Cloud or Gemini? And then also see where the boundaries are, because you need to also be aware that things make mistakes, they hallucinate facts.

So it's dangerous if you just over-rely on them.

And the best way is often just to get some experience yourself rather than just listening to a podcast or reading a book or a website where somebody says whether AI is great or AI is still failing here.