This is the third installment in the series of getting to know the DeepJudge team.
Meet Kevin Roth, Co-Founder & COO. Like the other co-founders of DeepJudge, Kevin earned his PhD in machine learning from ETH Zürich and has experience working at Google. Despite these shared achievements, each team member brings unique and diverse backgrounds and interests to the table. Join us as we ask them some questions and learn more about the people behind DeepJudge.
I hold a masters degree in theoretical physics and a doctorate in machine learning / AI from ETH Zurich. I was fortunate to have been surrounded by exceptional fellow students and mentors who inspired me to pursue a PhD in ML/AI at a time when the big “deep learning” paradigm shift was just unfolding—way before AI became the hot topic it is today.
My PhD research focused on understanding and enhancing the robustness and trustworthiness of deep neural networks, in particular of generative AIs. During those four-and-a-half years, I also spent time at Microsoft Research in Cambridge and Google Brain in Berlin, and got to know my co-founders.
It is rewarding to be working with bright minds on building powerful AI tools that deliver real value for our customers and users.
The legal industry is predestined for AI. With our Knowledge Assistant, we enable legal professionals to harness the entire collective knowledge of their firms, allowing them to focus on the truly strategic aspects of their work. I am pleasantly surprised to see how tech-forward the industry has become in recent months.
Even though we have already experienced remarkable advances in Al research and application, we are still only at the very beginning of this new era, starting to gain deep and fascinating insights into the immense potential of AI.
AI will change not only the way we work, but also the way we think, our culture, and society as a whole. The more you get into it, the more you start to see it as a clear logical succession of little steps.
Over the next five years, we will gain a much better understanding of where we are on the data vs. compute vs. intelligence surface and how easy / hard it is to build super-human intelligent systems (“AGI”) by simply scaling things up.
When I’m not working, I love to spend time mountaineering, rock climbing and backcountry skiing in the alps with my brother and friends. Unfortunately, not as much as I’d like to these days!
Few people know that I’ve been desk neighbors with two of the authors/inventors of the famous “Transformer” architecture, the foundation of Generative Pre-trained Transformers (the “T” in GPT).