David Niewolny

My latest for Forbes-- NVIDIA's work in healthcare is just getting started. I discuss NVIDIA's meteoric success, backed by decades of research and bold innovation, while specifically examining its progress spanning the fields of digital surgery, digital biology and digital health. The company's advancements in compute power have enabled significant progress in these fields. The analysis also looks at a variety of key areas that NVIDIA is innovating in, such as its #Holoscan platform, #DigitalTwins, #EdgeComputing, etc. Special thanks to Kimberly Powell for the great discussion and for helping articulate the passion and vision behind this important work. As always, keen to hear your thoughts and feedback! CC: Anthony Costa, Raghav Mani, Mona G. Flores, MD, Mahdi Azizian, David Bericat, Mukundan Govindaraj, Janette Ciborowski, Olivia Bass

219

2 Comments

In a recent episode of the NVIDIA AI podcast, Thomas Fuchs, co-founder and Chief Scientific Officer of Paige joined host Noah Kravitz to discuss how Paige is changing the trajectory of cancer diagnosis through AI-powered pattern detection. In addition to exploring this new technology, the episode highlighted a few reasons compliance is so crucial in enabling transformative AI-based technologies. AI’s evolution is outpacing our regulatory frameworks, creating a complex landscape for data privacy and data ethics challenges. As someone who spends large blocks of each day deeply immersed in data privacy regulations, I see several critical challenges ahead: 1. Regulatory Fragmentation: The global patchwork of privacy regulations (GDPR, CCPA, China's PIPL) creates compliance complexity for innovation with AI. How can we harmonize these frameworks while fostering innovation? 2. Data Minimization vs. AI Training: AI's data hunger conflicts directly with privacy principles like data minimization. How could techniques like continuous monitoring, compliance previews, and forward-looking encryption models offer protection at scale? 3. Explainability in Deep Learning: As models become more complex, so too becomes the ability to completely understand and explain what the model is doing. Advancing explainable AI is crucial not only for transparency but also for regulatory compliance and risk management. Is it possible to be confident that a model is compliant if it can’t be explained? 4. Bias in Historical Data: Historical, or “look back,” data often contains societal biases. How can we apply understanding bias to part of the data procurement and inclusion process, before model training begins? 5. Reimagining AI Governance: We have the opportunity to redefine corporate governance for the AI era. Will roles like Chief AI Ethics Officer and AI audit committees be granted real influence in companies to blend compliance and innovation? The recently published EU AI Act (Regulation (EU) 2024/1689) marks a significant step towards comprehensive AI regulation. With its phased implementation beginning February 2, 2025, and full applicability for most obligations by August 2, 2026, it sets a new standard for AI governance. The act's risk-based categorization of AI systems aligns with the growing need for clear and thoughtful regulation in the field. In the US, the proposed "Online Civil Rights Act" by the Lawyers' Committee for Civil Rights Under Law is a step in this direction, though it's just the beginning. Responsible AI is a competitive differentiator. It's not merely about avoiding fines—it's about building trust with both consumers and regulators. The future of AI regulation is exciting but will require unprecedented collaboration among technologists, ethicists, and policymakers. What challenges do you see or face with Responsible AI? https://lnkd.in/eYdSMYY5 #AIRegulation #DataPrivacy #ResponsibleAI #NVIDIAAI

28

1 Comment

Over 40 #healthcare and life sciences startups and enterprises are leveraging Meta Llama 3 NIM to develop and deploy applications that accelerate digital biology, digital surgery, and #digitalhealth 🧬 Learn more about use cases for NVIDIA NIM inference microservices in the blog >

3

This post is a great reminder of the significance of resilience. Success is seldom achieved without encountering obstacles. Looking back, there were numerous times when I thought I wouldn't make it through a situation. However, I emerged stronger and more confident each time. Every obstacle I faced was an opportunity to grow and learn. Let's all celebrate the power of resilience! #resilience #success https://lnkd.in/eP4e9cde

9

2 Comments

Over 40 #healthcare and life sciences startups and enterprises are leveraging Meta Llama 3 NIM to develop and deploy applications that accelerate digital biology, digital surgery, and #digitalhealth 🧬 Learn more about use cases for NVIDIA NIM inference microservices in the blog >

9

Over 40 #healthcare and life sciences startups and enterprises are leveraging Meta Llama 3 NIM to develop and deploy applications that accelerate digital biology, digital surgery, and #digitalhealth 🧬 Learn more about use cases for NVIDIA NIM inference microservices in the blog >

5

With NVIDIA NIM and NIM Agent Blueprints, medical researchers across the public sector can jump-start their adoption of state-of-the-art, optimized #AI models to accelerate their work. #healthcare #hc-digitalhealth

3

🌟Discover how AI-powered #ZeroShotLearning is accelerating treatments for rare diseases by identifying existing drugs in record time. 🧬💊 #AI #DrugRepurposing #HealthcareInnovation Read the full story ➡️

5

May be not! $NVDA CEO JENSEN HUANG SAYS QUANTUM COMPUTING IS EXCITING BUT LIKELY 15–30 YEARS AWAY -- CASTING A REALITY CHECK ON THE HYPE AROUND $QUBT, $QBTS, $RGTI & $IONQ 👀 That’s a well-reasoned argument, and I’d like to delineate why Spectral Capital is uniquely positioned to play a meaningful role in the quantum computing ecosystem today. Jensen’s point about ‘useful quantum computers’ highlights the long-term aspirations of the field, but as you aptly noted, we don’t have to wait decades for quantum technologies to deliver value. Companies like Spectral Capital demonstrate this by leveraging hybrid quantum-classical systems to solve real-world problems today. Spectral’s focus on plasmonics and hybrid cloud systems is particularly significant. Plasmonics, which utilizes the interaction of light and free electrons on a metal surface, is an innovative approach to overcoming some of the current limitations in quantum computing hardware. By complementing this with hybrid cloud systems, Spectral is effectively enabling quantum to sidestep the reliance on cryogenic supercomputers and deliver solutions that are practical, scalable, and accessible in the near term. This approach aligns with the broader trend of “staircase innovation,” where incremental progress builds toward the eventual vision of large-scale, fault-tolerant quantum systems. It’s easy to dismiss the industry as a bubble when focusing solely on the distant future, but Spectral and its peers are proving that value creation is happening in the here and now. Optimization, logistics, material discovery, and even financial modeling are being reshaped by these technologies, and Spectral’s contributions are helping lay the groundwork for a more quantum-enabled world. In short, #SpectralCapital is not just betting on the future of quantum computing—it’s actively participating in building it, one step at a time. #quantumcomputing Sean Brehm Spectralcapital.com

20

8 Comments

The short video can help prepare you for any physician’s appointment to make the most of your limited time.

2

A thoughtful interview on the future of ML and some strategic pressure in the ecosystem from an R&D and business standpoint. Thanks Yariv Adan for calling it out

2

💡 A Major Leap in AI Infrastructure: Ayar Labs Shines Bright with Optical Innovation! 🚀 Ayar Labs recently secured $155M in funding from industry giants like AMD, Intel, and NVIDIA to address the growing demand for scalable and cost-effective AI infrastructure. 🖥️✨ Their cutting-edge optical I/O solutions promise to revolutionize data communication by delivering unprecedented speed, efficiency, and scalability — a crucial step as AI systems grow larger and more complex. With this strategic backing, Ayar Labs is positioned to redefine the future of AI, cloud, and HPC systems. 🌐🔗This is a significant move towards bridging the gap in high-performance data communication. 🔗 Read More: https://lnkd.in/gUx-myd4 ▶️ Watch - Bloomberg Coverage: https://lnkd.in/gY8UGPDk #AIInfrastructure #Innovation #TechFunding #AI #CloudComputing #HPC #OpticalIO #FutureOfAI #TechEcosystem #Scalability #TechNews

5

Inspiring to hear Jensen Huang's lecture at Stanford University delved deeply into his philosophies on people, culture, and leadership, emphasizing personal and organizational growth. Here are some of the subtler, heart-centered themes he discussed: 1. Importance of Pain and Suffering: Huang openly talks about pain and suffering at Nvidia, stating that greatness comes from overcoming challenges. He encourages embracing difficulties as they build resilience and character, which are essential for long-term success. 2. Empathy and Connection: He spends much of his time brainstorming with managers and leaders, helping them navigate tasks and opportunities. This approach fosters a deep connection and understanding within the team, promoting a collaborative and supportive work environment. 3. First-Principles Thinking: By breaking down complex problems into their basic components and rebuilding solutions from the ground up, he encourages innovative thinking that goes beyond conventional methods. This approach not only drives technological advancements but also instills a mindset of curiosity and continuous learning within the organization. For those interested in exploring these insights further, I highly recommend watching the full lecture.

11

As AI grows, so does its demand for power. NVIDIA AI's director of accelerated computing, Dion Harris, and Penn Engineering expert Benjamin Lee discuss ways to minimize AI energy demand and grid impact in the new episode of the Energy Policy Now podcast.

4

With NVIDIA NIM and NIM Agent Blueprints, medical researchers across the public sector can jump-start their adoption of state-of-the-art, optimized #AI models to accelerate their work. #healthcare #hc-digitalhealth

3

Welcome to 2025 ! LLMs? Old news. Agentic AI ....so what ? The spotlight is now on WFMs (World Foundational Models). I’m betting every VC will become a WFM expert in the next 90 days—mark my words! 😏 Jokes aside, NVIDIA's bold move into open-sourcing their models is a game-changer. This move is setting the stage for intense innovation among the major players—OpenAI, Gemini, Meta (Llama), and others. NVIDIA is showcasing the art of the possible. By pushing use cases that demand even more data and compute power than LLMs, they want to push these players to consume more compute. The bar has been raised. Who’s ready to rise to the challenge? #AI #WFMs #Innovation #NVIDIA #OpenSource

18

1 Comment