Marie Brueser

Europe has the talent — now it’s time to bet on ourselves. When Sandra Navidi and I met in Davos during World Economic Forum, one question kept coming back: What will it take for Europe to build global tech champions? We continued that conversation on ntvNachrichten’s BIZ & beyond podcast with Ulrich Reitz. From building Europe’s first venture studio, finleap, in Berlin to now leading early-stage investing at Motive Ventures in New York, that vantage point shapes how I see Europe’s path forward. Here’s what I believe: ▪️ Europe has great ideas — but we lack the structures, speed, and capital to scale them. ▪️Especially in later stages, we hesitate to go all in — and when capital finally comes, it’s often not European. ▪️We should learn from the U.S. — but not everything that shines is gold. ▪️Our real strengths, like quality of life and global talent, are already here. ▪️Capital matters. But culture decides. If we want to build global champions, we need belief. Belief that we can build big — and the urgency to work for it. That mindset is one of the few things Europe should take from the U.S. In short: If Europe wants to become a true tech force, the time to act is now. And it starts by betting on ourselves. A big thank you to Sandra and Ulrich for an insightful, thought-provoking discussion. And about that headline calling me “Tech-Wunderkind” — it made me smile. Feels good to be young again 😄 🎧 Listen via the link in the comments (German only). #Capital #Entrepreneurship #Europe #Leadership #Fintech

176

13 Kommentare

📢 VOICES OF STARTUP BERLIN 📢 – a conversation with Thorsten Lubinski (Co-Founder) and Catalina Plesmann (Head of Marketing) at DiaMonTech AG – who won the DeepTech Award in 2024! > In simple terms: How does your technology make blood sugar monitoring easier for patients? We developed a small device that lets you measure your blood sugar non-invasively by simply placing your finger on our sensor - no needles, no blood, no pain. People living with diabetes can check their levels more often, more comfortably, and without disposable supplies.    > What’s the biggest challenge in developing non-invasive medical tech – and how did you overcome it? The hardest part is precision. Glucose is present in very small amounts and hard to isolate without breaking the skin. We solved this by usiung a quantum cascade laser that targets glucose molecules specifically. After years of development and a successful clinical study, we were able to prove that our non-invasive approach is medically accurate. > How do you see your solution changing lives — especially for people living with diabetes long-term? Frequent glucose monitoring is key to managing diabetes, but many people avoid it because it’s painful or inconvenient. Our technology removes that barrier. Over time, this could mean better glucose control, fewer long-term complications, and a higher quality of life. Not just for people living with diabetes, but also for those at risk or those who take longevity seriously.    > What impact did winning the Deep Tech Award have on your visibility or credibility in the healthcare space? It put a spotlight on our work. The Deep Tech Award gave us recognition beyond the scientific community and helped position us as a serious innovator in both the MedTech and digital health space. It helped us connect with new partners and contributed to growing trust in our approach. > What role does #Berlin play in your journey as a MedTech innovator? Berlin is where science meets entrepreneurship. The city gave us access to leading research institutions, public funding programs, and a vibrant healthtech ecosystem. It’s also home to an open-minded community and strong infrastructure, the ideal base to grow a medical startup with global ambitions.    📣 Big news: DiaMonTech have just started a new Crowdinvesting-Kampagne campaign. Find more info in the comments 👇   Curious about the 2025 winners of the deep tech award? Keep an eye on Deep Tech Berlin for updates! 🚀

50

3 Kommentare

🚨 New “foundation model” for reasoning on text and time series by Stanford Biodesign Digital Health, Google and ETH Zürich. Spoiler alert: OpenTSLM lacks evidence to support its claims, does not show evidence of working beyond a narrow problem set, and is not more accurate than classical models. 👉 OpenTSLM is marketed as a breakthrough multimodal foundation model for temporal reasoning with order-of-magnitude gains in medical reasoning and beyond. ‼️But... no evidence to support those claims is to be found. The paper overstates results, omits key literature, and cherry-picks baselines, falling short of minimal standards for methodological rigor and transparency. Here are 🔟 points that I find problematic and share in the spirit of academic exchange. ⚠️ Core concerns: 1️⃣ Misleading framing: Tasks are mostly templated classification, not real temporal reasoning. No forecasting, imputation, or anomaly detection. 2️⃣ False claims: “order-of-magnitude gains in temporal reasoning” is simply false and misleading. 3️⃣ Missing literature: Complete omission of related work. E.g. Chronos by Amazon , TimesFM by Google, Moirai by Salesforce, Toto by Datadog, TabPFN by Prior Labs, TiRex by NXAI, Moment by Carnegie Mellon University... 4️⃣ Weak baselines (cherry-picking): self constructed datasets and omission of standard baselines. (E.g., Sleep-EDF results are far below models from the last decade) 5️⃣ Template dependence — Models were trained to solve one very particular task and don't show any signs of generalization. 6️⃣ Compute inefficiency — Enormous complexity and compute for subpar results. 7️⃣ No forecasting, classification or anomaly detection: Using ChatGPT-style captioning is not forecasting. The authors do include code to handle the Monash Dataset and The UCR Archive but completely omit any reference in the paper. 8️⃣ Missing results, logs, checkpoints, and relevant ablations — No breakdown of curriculum, rationale, or encoder impact. 9️⃣ Code irregularities — >122 files deleted before publishing; including evaluation notebooks and dataset generation script. Also the code history shows that some standard benchmarks (e.g. PAMAP2) were introduced, presumably evaluated and then replaced. 🔟 Suboptimal quality — The paper is signed, endorsed, and shared by researchers from from top institutions. Such an amazing set of authors can and has to produce better work. 🔍 Transparency, quality, and rigor matter — especially in fields like healthcare. 🎖️ If we want trustworthy foundation models we need to raise the bar, not lower. Patrick Langer, Thomas Kaar, Max Rosenblattl, Paul Schmiedmayer, Filipe Barata Welcome to the time series field, and #happyforecasting. 💬 Thoughts, evidence, or corrections welcome. If I’ve missed something, please point it out — I’ll happily adjust or retract. #AI #MachineLearning #TimeSeries #FoundationModels #Transparency #MLResearch #AcademicIntegrity

53

13 Kommentare