Disqus - Latest Comments for dhdeans

Re: AI Adoption Is a Challenge. Here’s a Solution.

David H. Deans — Fri, 01 May 2026 15:47:17 -0000

In the enterprise software sector, my area of focus, AI adoption with individual high-performance practitioners that have a track record of investing in their skills are doing just fine. However, in contrast, the collective employee peer group is behind the individual early adopters. Most attended an employer-led training class, but few applied what they learned. Moreover, this mainstream vendor workforce is uninterested in AI skills development at this time. Perhaps they will be impacted by the ongoing tech industry layoffs. And yet, most won't expect to be cut.

Re: Why AI Disclosure Matters at Every Level

David H. Deans — Fri, 23 Jan 2026 10:17:44 -0000

I see this issue from a different perspective. Some large enterprise employees say they use AI in their work to appear skilled and competent. However, they lie about their AI knowledge, and that's unethical. Middle managers seem to be the most egregious.

Re: How Can Companies Incentivize AI Adoption?

David H. Deans — Thu, 15 Jan 2026 17:47:29 -0000

We know from empirical research that employee training for Applied-AI is much more effective when it is followed by 1:1 coaching. Otherwise, people tend to forget because they haven't applied what they've learned in the class setting. That said, publicly acknowledging the pioneer learners also helps to increase peer adoption and ongoing usage.

Re: NVIDIA Unveils Multi-Agent Intelligent Warehouse and Catalog Enrichment AI Blueprints to Power the Retail Pipeline

David H. Deans — Fri, 09 Jan 2026 09:08:47 -0000

What stands out here is the shift from dashboards that merely describe yesterday’s performance to agents that can explain why something is off, surface the evidence, and recommend actions such as rebalancing tasks, tuning safety policies, or filling sparse catalog attributes at scale. Retailer CIOs can apply this insight to guide their AI investment approach.

Re: Marine Biological Laboratory Explores Human Memory With AI and Virtual Reality

David H. Deans — Fri, 02 Jan 2026 10:17:04 -0000

Turning a large dataset bottleneck into an interactive workflow is a great example of how an applied-AI mindset, not just raw compute, is what converts massive data into real scientific traction on questions like protein localization and memory encoding. That's insightful.

Re: Adopting AI: Starting Small While Thinking Big

David H. Deans — Tue, 30 Dec 2025 11:58:35 -0000

Sovereign Intelligence might gain momentum in 2026, as more large enterprise leaders deploy private small language models trained with their subject matter experts' industry-specific domain knowledge. Public LLMs will still be the preferred tool for General Intelligence use cases where ownership won't matter as much.

Re: Now Generally Available, NVIDIA RTX PRO 5000 72GB Blackwell GPU Expands Memory Options for Desktop Agentic AI

David H. Deans — Fri, 26 Dec 2025 09:06:12 -0000

Kudos to the NVIDIA team and early adopters leaning into this direction. Pairing this kind of workstation-grade hardware with a clear, Applied-AI strategy — governed by real KPIs and change-management plans — will be a key differentiator for forward-thinking enterprises and studios that want to turn AI potential into a durable competitive advantage. Well done.

Re: UC San Diego Lab Advances Generative AI With NVIDIA DGX B200 | NVIDIA Blog

David H. Deans — Fri, 19 Dec 2025 10:27:27 -0000

What really stands out here is the shift toward disaggregated inference; separating prefill and decode across different GPUs to unlock low-latency, scalable serving without sacrificing quality, a critical ingredient for production-grade AI agents and applications. An Applied-AI mindset builds on exactly this kind of work: treating hardware, scheduling, and serving strategies as levers for tangible outcomes such as faster experimentation cycles, more responsive interfaces, and better ROI, rather than just peak benchmark numbers.

Re: NVIDIA and AWS Expand Full-Stack Partnership, Providing the Secure, High-Performance Compute Platform Vital for Future Innovation

David H. Deans — Fri, 12 Dec 2025 09:18:50 -0000

Nemotron open models on Bedrock, GPU-accelerated vector indexing in Amazon OpenSearch with cuVS, and co-engineered tooling for agent development and observability all aim to shorten the distance from hardware to impact. Organizations that adopt an Applied-AI mindset will be best positioned to translate these capabilities into measurable outcomes in productivity, customer experience, and new revenue. Well done!

Re: NVIDIA Awards up to $60,000 Research Fellowships to PhD Students

David H. Deans — Fri, 05 Dec 2025 10:47:31 -0000

The combination of a substantial fellowship and a built-in internship is a powerful way to bridge cutting-edge academic work with real-world impact. It reflects how deeply intertwined hardware, software, systems, and human interaction have become in this next era of AI.

Re: FLUX.2 Image Generation Models Now Released, Optimized for NVIDIA RTX GPUs

David H. Deans — Fri, 28 Nov 2025 09:18:30 -0000

This FLUX.2 release feels like an important update for local image generation on RTX hardware. The combination of FP8 quantization, VRAM savings and weight streaming in ComfyUI meaningfully lowers the barrier for creators who don’t have datacenter-class GPUs at home. Much appreciated, indeed.

Re: How Are Companies Using Gen AI in 2025?

David H. Deans — Mon, 24 Nov 2025 10:41:34 -0000

As more C-suites fund Applied-AI Initiatives, the large enterprise focus will continue to shift away from GenAI tools and IT infrastructure. That said, too few companies offer practitioner skills development resources that enable the achievement of desired business outcomes. Hopefully, skills will be the priority in 2026.

Re: The Largest Digital Zoo: Biology Model Trained on NVIDIA GPUs Identifies Over a Million Species

David H. Deans — Fri, 21 Nov 2025 11:01:35 -0000

The BioCLIP 2 project is a good example of how foundation models can drive real-world impact beyond traditional AI domains. By linking biological insight with large-scale visual learning, this work demonstrates how accelerated computing can help close the long-standing data gaps in biodiversity and conservation science. Well done, indeed.

Re: AI On: 3 Ways to Bring Agentic AI to Computer Vision Applications

David H. Deans — Fri, 14 Nov 2025 10:46:04 -0000

What stands out to me is the practical approach to augmenting, rather than replacing, existing CNN-based deployments. By making visual data searchable and alerting systems more contextually aware, organizations not only gain operational agility but also experience tangible benefits in compliance and safety. Well done, indeed.

Re: GeForce RTX GPUs Power Creative Workflows | NVIDIA Blog

David H. Deans — Fri, 07 Nov 2025 09:29:20 -0000

The ability to run demanding AI effects and generate professional-level outputs faster is a real breakthrough for anyone serious about content creation. Major kudos to the NVIDIA team for driving the ecosystem forward and empowering creators at every level. Much appreciated.

Re: AI-Powered Mobile Clinics Deliver Breast Cancer Screening to India’s Rural Communities

David H. Deans — Fri, 31 Oct 2025 09:10:09 -0000

This is an inspiring demonstration of tech serving communities — bringing AI-powered breast cancer screening directly to rural regions where early detection is often out of reach. The results from MedCognetics and "Health Within Reach" show that when advanced AI platforms are combined with local expertise and telehealth workflows, real lives are changed for the better. Truly, well done.

Re: NVIDIA GTC DC: Live Updates on What’s Next in AI

David H. Deans — Sat, 25 Oct 2025 08:17:42 -0000

The convergence of public sector leaders, developers, and innovators signals how quickly AI is moving from concept to national capability. It’s encouraging to see discussions on agentic AI and quantum alongside actionable demos and workshops — a rare blend of vision and implementation.

Re: Placeholder IMAX | NVIDIA Blog

David H. Deans — Fri, 10 Oct 2025 11:20:13 -0000

NVIDIA’s continued push for full-stack, open-source optimized AI infrastructure is inspiring. The performance and efficiency metrics reported for Blackwell in the InferenceMAX benchmarks highlight innovation leadership and also a deep understanding of the economic pressures facing enterprise AI deployments. The collaboration with model builders and open frameworks will help to democratize access and accelerate innovation across sectors.

Re: How to Get Started With Large Language Models on NVIDIA RTX PCs

David H. Deans — Fri, 03 Oct 2025 14:39:15 -0000

The optimizations highlighted for Ollama, AnythingLLM, and LM Studio — especially the collaboration with open frameworks and the new support for models like gpt-oss, Gemma 3, and Nemotron Nano — showcase NVIDIA’s commitment to an open and high-performance edge AI ecosystem. The proliferation of hands-on tools for custom workflows (from study buddies to certification prep) is a good step forward for independent learning and productivity.

Re: Open Secret: How NVIDIA Nemotron Models, Datasets and Techniques Fuel AI Development

David H. Deans — Fri, 26 Sep 2025 11:35:38 -0000

The practical integrations highlighted clearly demonstrate the real impact of an open and transparent ecosystem. It's encouraging to see technologies like NVFP4 and collaborations with projects like Llama and Qwen further shape next-generation GPU architectures and reasoning capabilities. This is an essential path toward further open innovations.

Re: Operationalize AI Accountability: A Leadership Playbook

David H. Deans — Thu, 25 Sep 2025 17:52:48 -0000

The Applied-AI Initiative is a helpful source of additional insights

Re: The UK’s ‘Goldilocks Moment for AI’: NVIDIA, UK and US Leaders Highlight AI Infrastructure Investments

David H. Deans — Fri, 19 Sep 2025 09:58:38 -0000

The collaboration between NVIDIA, leading UK and US officials, and a robust network of academic and venture partners demonstrates a commitment to fostering innovation at scale. Investing in national infrastructure, talent development, and cross-sector partnerships is essential to ensure the United Kingdom's place in a rapidly emerging AI-enabled economy.

Re: Use Generative AI For Content Creation With ComfyUI | NVIDIA Blog

David H. Deans — Fri, 12 Sep 2025 11:23:16 -0000

NVIDIA’s ongoing partnership with ComfyUI is empowering both creators and developers by bringing major speed and efficiency gains to GenAI workflows. The integration of TensorRT-optimized diffusion models, alongside impressive new video, image, and 3D generative capabilities like Wan 2.2 and Hunyuan3D 2.1, sets a new bar for local creative production. Very helpful, well done!

Re: AI On: 6 Ways AI Agents Are Raising Team Performance — and How to Measure It

David H. Deans — Fri, 05 Sep 2025 11:53:24 -0000

Measuring success by adoption, engagement, task completion, and business results is essential, and it’s great to see a clear framework for leaders aiming to maximize AI agent ROI. The focus on supporting high-value human work, personalizing education, and guiding operational excellence will empower organizations at every level. Well done!

Re: How Do You Teach an AI Model to Reason? With Humans

David H. Deans — Fri, 29 Aug 2025 10:12:51 -0000

The collaborative efforts of the diverse data factory team, along with rigorous data curation and reinforcement learning, set a new benchmark for responsible and human-centered AI development. It's encouraging to see how these pioneering innovations continue to shape the future of intelligent systems.