Panel: The Future of AI

on YouTube ↗31m

" summary

Panelists report concrete advances: teams train a 530B-parameter model with NeMo/Megatron on Azure using NVIDIA A100 clusters and InfiniBand, and roadmap H100 Tensor Core GPUs and Superpod architectures to scale to trillion-parameter models. Production tooling includes Azure Machine Learning, interpret-ml and Fairlearn, PyTorch/TensorFlow and Ray for fine-tuning, plus domain wins in image-driven drug discovery. Microsoft publishes a Responsible AI standard, mandates impact assessments and controls (for example gating custom neural voice), and emphasizes full-stack hardware‑software validation.

AI-written summary · about

" chapters (6)

1. 0:00Welcome and introductions
2. 0:45Panelist backgrounds
3. 5:51Current state of AI
4. 14:14The next decade of AI
5. 20:08Industry evolution and AI
6. 23:04Responsible AI practices

" related talks

• 28mHarnessing generative AI with NVIDIA AI and Microsoft AzureTheresa Campbell, Jimmy Della, Mary Howell, Eva Wasielewski, Uttara Kumar, Dave Salvator, Matt Vegas · Microsoft Build 2023
• 16mUnlocking AI Innovation: A Conversation with Eric Boyd and Asha SharmaEric Boyd, Asha Sharma · Microsoft Build 2024
• 46mMultimodel Multimodal and Multiagent innovation with Azure AIMark Schoennagel, Marco Casalaina · Microsoft Build 2024
• 22mQuickly deploy NVIDIA optimized generative AI on Azure AI StudioVineeth Kalluru, Manuel Reyes-Gomez · Microsoft Build 2024
• 43mBeyond the AI Hype: What's Real, What's NextRichard Campbell · Build Stuff 2025
• 24mIndustry PanelChris Lattner, Michael Suo · AI Infra @Scale 2024
• 54mThe Future of AI: Building the Most Impactful Technology TogetherLeandro von Werra · PyCon DE & PyData 2025
• 29mBuild Customize and Deploy LLMs At-Scale on Azure with NVIDIA NeMoTheresa Campbell, Jimmy Della, Mary Howell, Eva Wasielewski, Hugo Affaticati, Joey Conway, Vinh Nguyen, Shashank Verma · Microsoft Build 2023