Sign In

Rasmus Rothe

Building and Investing in AI @ Merantix

Rasmus Rothe is an accomplished individual who is deeply involved in the AI industry, aiming to address global challenges by bridging the gap between AI research and practical application.

Rasmus is a co-founder of Merantix, a company focusing on building and investing in AI ventures across various sectors like healthcare, biotech, climate, manufacturing, and automotive.

In addition to Merantix, he is also the co-founder of the AI Campus, a not-for-profit entity dedicated to fostering AI ecosystems, with a particular focus on the European landscape.

Rothe is highly engaged in advocating for AI development in Europe and is a co-founder and co-leader of the German AI association (KI Bundesverband e.V.), a key representative body for AI startups in Germany.

Rasmus Rothe's educational background is extensive, including studying computer science at prestigious institutions like Oxford, Princeton, and ETH Zurich, culminating in a PhD in computer vision and deep learning.

His academic accomplishments are notable with over 15 papers published in reputable conferences and journals such as CVPR, ICCV, and IJCV.

Prior to his current roles, Rasmus scaled a deep learning service analyzing faces to a vast user base, also founding Europe's largest hackathon, HackZurich.

His professional journey includes stints at Google and The Boston Consulting Group, further enriching his experience in the tech and business spheres.

Rothe's diverse organizational affiliations underscore his commitment to advancing AI and technology, having been involved in various capacities with entities like AI House Davos, Deltia, Briink, Cambrium, and more.

Overall, Rasmus Rothe is a highly accomplished individual in the AI domain, with a strong focus on practical application, ecosystem building, and advocacy for AI advancement.

His combination of academic expertise, entrepreneurial endeavors, and strategic collaborations positions him as a key figure in the European AI landscape.

Highlights

Sep 12 · linkedin

Can AI Generate Novel Research Ideas? A Major Study Offers New Insights!

A recent study from Stanford University tackles an intriguing question: can large language models (LLMs) match human experts in generating research ideas? The findings are both surprising and thought-provoking.

The study recruited over 100 NLP researchers to compare AI-generated ideas with those from human experts. Remarkably, the LLM-generated ideas were judged as more novel than those of humans. However, while AI excelled at creativity, it fell short on feasibility, suggesting that while LLMs can introduce innovative concepts, they often need human refinement to make these ideas practical.

What does this mean for AI?

AI as a Catalyst for Innovation: AI has the potential to inject fresh, unconventional perspectives into research, helping scientists think beyond established boundaries. This could significantly accelerate the ideation phase of scientific work, where creativity is critical.

Human-AI Synergy: Rather than replacing researchers, AI may enhance their ability to explore new ideas. The study points toward a future where AI acts as a brainstorming partner, generating bold concepts, while human experts assess their feasibility and execution.

Challenges and Next Steps: Despite its creativity, the AI struggled with diversity and evaluating its own ideas. This highlights the need for improvements in AI’s ability to balance novelty with practicality.

This study marks an important milestone in the evolving relationship between AI and research. While AI shows real promise in driving creativity, human insight remains essential to bring those ideas to life. Together, they could transform the future of scientific discovery.

Read the full paper here: https://lnkd.in/dseP34AY

#AI #Research #NLP #ArtificialIntelligence #ScientificDiscovery #Innovation #FutureOfResearch #LLMs

Sep 10 · linkedin

26% More Productive: How Generative AI is Transforming Software Development:

There’s been debate about whether generative AI truly enhances productivity for knowledge workers. Now, thanks to a randomized control trial, we have real-world evidence that proves its effectiveness.

In a study conducted across Microsoft, Accenture, and a Fortune 100 electronics company, software developers using an earlier version of GitHub Copilot, powered by GPT-3.5 (Technology from 2021!), achieved a remarkable 26% increase in completed tasks. The most notable gains were seen among junior and less experienced developers, demonstrating AI's ability to elevate productivity for those early in their careers.

This large-scale, real-world experiment is one of the first to provide hard evidence that generative AI is more than just a futuristic concept—it’s already driving tangible results in high-skilled work environments. And with newer models on the horizon, the potential for even greater impact is clear. Exciting times ahead for software development!

Read the full paper here:  https://lnkd.in/dqvPuBAh

#GenerativeAI #AIProductivity #SoftwareDevelopment #AIResearch #TechInnovation #GPT #AI

Rasmus Rothe
Add to my network

Location

Berlin, Berlin, Germany