Industrialize AI 'factories'—from power-train design and thermal-chain cooling to prefabrication—so teams deploy high-density AI capacity faster, more predictably, and with lower risk.

Industrialize AI Data Centers: Power, Cooling & Prefab Strategies

Optimizing AI Factories

Learn how leaders are driving unprecedented efficiency in modern data center buildouts.

Watch the NVIDIA GTC Live replay: https://nvda.ws/48uTsS6

A clinician-technologist explains how thoughtfully integrated AI can cut clerical burden, improve documentation, and restore clinician presence—featuring a 'Bridge' case study.

How AI Reduces Clinician Burnout & Restores Human-Centered Care

AI Elevates Human Connection

Hear how AI frees clinicians to focus on patients and transform the care experience.

Explore how AI — from generative design and predictive maintenance to human-centered augmentation and supply chain intelligence — can make U.S. industry smarter and more competitive.

AI for American Manufacturing: Boost Productivity & Resilience

NVIDIA GTC Live Pregame: How AI Can Bring Industrial Back to America

See how AI is accelerating breakthroughs and powering the next era of productivity.

Watch the NVIDIA GTC Live replay: https://nvda.ws/3MBiKGX

NVIDIA GTC Live Pregame

Discover how Pixar's RenderMan XPU leverages modern GPUs to speed iteration, handle complex scenes, and reshape production workflows—where art challenges technology.

How Pixar’s RenderMan XPU Is Transforming GPU Rendering

Art Meets Innovation: Inside Pixar’s RenderMan XPU Revolution

Pixar is pushing the boundaries of storytelling by using NVIDIA GPUs to accelerate rendering with RenderMan XPU. This technology empowers artists to handle larger, more complex scenes faster, transforming workflows and enhancing creative possibilities. See how this partnership helps bring Toy Story 5 to life with unprecedented rendering performance and artistic freedom.

Learn more about NVIDIA and Pixar at SIGGRAPH Asia: https://nvda.ws/4aq0Uk6

Voice is now integrated directly into ChatGPT: speak naturally, see a live transcript, and get real-time visuals like maps and weather—switch between voice and text seamlessly.

ChatGPT Voice Built into Chat — Live Transcripts & Real-Time Visuals

What's New with ChatGPT Voice: Built Right into Chat for Live Conversation and Real-Time Visuals

You can now use ChatGPT Voice right inside chat — no separate mode needed.

You can talk, watch answers appear, review earlier messages, and see visuals like images or maps in real time.

Rolling out to all users on mobile and web. Just update your app.

If you prefer the original experience, turn on “Separate mode” under Settings → Voice Mode.

What's New with ChatGPT Voice

Philosopher Amanda Askell explains how ethics, identity and pragmatic philosophy should shape AI behavior — from deprecation and model welfare to 'superhuman' moral reasoning.

Amanda Askell on AI Ethics, Model Identity & Practical Philosophy

A philosopher answers questions about AI

Amanda Askell is a philosopher at Anthropic who works on Claude's character. In this video, she answers questions from the community about her work, reflections and predictions.

0:00 Introduction
0:29 Why is there a philosopher at an AI company? 
1:24 Are philosophers taking AI seriously? 
3:00 Philosophy ideals vs. engineering realities 
5:00 Do models make superhumanly moral decisions? 
6:24 Why Opus 3 felt special 
9:00 Will models worry about deprecation? 
13:24 Where does a model’s identity live?
15:33 Views on model welfare
17:17 Addressing model suffering
19:14 Analogies and disanalogies to human minds
20:38 Can one AI personality do it all?
23:26 Does the system prompt pathologize normal behavior?
24:48 AI and therapy
26:20 Continental philosophy in the system prompt
28:17 Removing counting characters from the system prompt
28:53 What makes an "LLM whisperer"? 
30:18 Thoughts on other LLM whisperers
31:52 Whistleblowing
33:37 Fiction recommendation

Further reading:
Claude’s character: https://www.anthropic.com/research/claude-character 
When We Cease to Understand the World by Benjamin Labatut: https://www.penguinrandomhouse.com/books/676260/when-we-cease-to-understand-the-world-by-benjamin-labatut-translated-from-the-spanish-by-adrian-nathan-west/

A practical playbook for agent memory: trim, isolate, and retrieve context; avoid context burst, conflict, and poisoning; and scale memory systems from experiment to production.

Practical Playbook for Agent Memory & Context Engineering

Build Hour: Agent Memory Patterns

AI agents don’t just reason — they remember. In this Build Hour, we deep-dive into context engineering techniques that enable agents to maintain short-term and long-term memory, personalize interactions, and operate reliably across long-running workflows.


Emre Okcular (Solutions Architect) covers:
• Why memory matters: stability, personalization, and long-running agent workflows
• Short-term memory patterns: Sessions, context trimming, compaction, summarization
• Long-term memory patterns: state objects, structured notes, memory-as-a-tool
• Architectures: token-aware sessions, state injection strategies, guardrails, and memory triggers
• Live demo: building an end-to-end agent with dynamic short and long term memory
• Best practices: avoiding context poisoning, context burst, context noise and context conflict.
• Live Q&A

👉 Context Engineering Cookbook: https://cookbook.openai.com/examples/agents_sdk/session_memory
👉 OpenAI Agents Python SDK: https://openai.github.io/openai-agents-python/
👉 Context Summarization with Realtime Cookbook: https://cookbook.openai.com/examples/context_summarization_with_realtime_api
👉 Follow along with the code repo: https://github.com/openai/build-hours
👉 Sign up for upcoming live Build Hours: https://webinar.openai.com/buildhours/

00:00 Context Engineering
10:44 Context Lifecycle Demo
20:13 Context Engineering Techniques
26:49 Reshape + Fit Demo 
39:16 Conclusion
42:45 Q&A

A reporter's look at 2025 search behavior — the phrases, emotions, and category trends (news, health, travel, tech, culture, climate) that defined the year.

2025 Search Insights: The Phrases That Shaped the Year

2025 – Year in Search Trailer: What People Asked, Felt, and Needed

Sneak peek 👀 Our Year in Search 2025 video is coming soon. Stay tuned…. #YearInSearch

And if you’re feeling nostalgic, revisit the moments the world searched for over the last decade →  
https://www.youtube.com/playlist?list=PL590L5WQmH8fmto8QIHxA9oU7PLVa3ntk 
https://www.youtube.com/playlist?list=PL590L5WQmH8f0U3hwuzUPD2rsgFZQ_Jhx 

Explore more top trends from the year → https://goo.gle/3XwLbZ5

Subscribe to our Channel: https://www.youtube.com/google
Find us on X: https://twitter.com/google
Watch us on TikTok: https://tiktok.com/@google
Follow us on Instagram: https://www.instagram.com/google
Join us on Facebook: https://www.facebook.com/Google

2025 – Year in Search Trailer

Cristiano Ronaldo on what truly defines greatness—longevity, impact, versatility and mental toughness—plus a breakdown of his 2018 bicycle kick and the 'CU' celebration.

CR7 on Greatness: Legacy, Iconic Goals & How to Master the Bicycle Kick

Perplexity x CR7 — Cristiano Ronaldo on Greatness, Goals, and How to Master the Bicycle Kick

Curiosity is how greatness begins. 

Proud to partner with @Perplexity to inspire more people to explore their curiosity.  

perplexity.ai/ronaldo is just the beginning.

Perplexity x CR7

An investigative roundup of 25 unresolved WWII cases—disappeared officials, missing art and gold, secret operations—what historical records show and what remains unexplained.

Unsolved WWII Mysteries: Vanished Leaders & Lost Secrets

25 World War II Mysteries History Could Not Explain

Během druhé světové války se objevila zmizení, rozhodnutí a události, které nikdy nebyly zcela objasněny. Někteří z nejmocnějších mužů Třetí říše — jako Müller, Canaris, Hess či Bormann — zmizeli v chaosu závěru války, zanechávajíce po sobě protichůdné teorie, zničené archivy a stopy, které nebylo možné sledovat. Zároveň legendární poklady, jako Jantarová komnata, tajné náklady na Filipínách nebo truhly vhozené do jezera Toplitz, zmizely za okolností, které po desetiletí živily pátrání a spekulace.

Zrychlený technologický vývoj a zoufalství změnit průběh války daly vzniknout projektům, které byly stejně skutečné jako záhadné. Od Me 262 po extrémní experimenty jednotky 731, přes tajemná zařízení jako údajný „Zvon“ nebo světelné jevy známé jako Foo Fighters — hranice mezi vědou, mýtem a propagandou byla stále rozmazanější. Skryté aliance, únikové cesty a ponorkové operace směrem do Jižní Ameriky a Asie vytvořily síť tajemství, která zůstala uvězněna pod oceány, ledem nebo v tichu oficiálních dokumentů.

Na bojištích se ani vojenská logika ne vždy shodovala s realitou. Stalingrad, Hürtgen, Arnhem, Monte Cassino či Ržev zanechaly tisíce mrtvých bez jasného záznamu, strategická rozhodnutí bez vysvětlení a svědectví vojáků popisujících pozměněné vnímání, nemožné zvuky či nevysvětlitelné přítomnosti v okamžicích největšího napětí. Mezi taktickými chybami, lidskými zmizeními a jevy obtížně zařaditelnými válka zanechala dědictví, v němž ne všechno bylo možné říci, vysvětlit nebo obnovit — a některé epizody dodnes zůstávají zahalené ve stínu.

25 Záhad druhé světové války, které historie nedokázala vysvětlit

Step-by-step guide to checkpointing Flax NNX models with Orbax: extract nnx.state, use nnx.split/merge/update, and reliably save & restore model and optimizer state.

Orbax Checkpointing for Flax NNX — Practical Save & Restore for JAX

Checkpointing Flax NNX Models with Orbax (Part 1)

Orbax is the standard checkpointing library in the JAX ecosystem.  In part one of this two episode series on checkpointing with Orbax we’ll explore how to effectively save and restore your Flax NNX models.  As we’ve seen, Flax NNX offers a Pythonic, stateful approach which feels closer to frameworks like PyTorch, while still having all the advantages of JAX. We'll cover how NNX manages state and how Orbax interacts with it, starting with the basics and moving to advanced techniques like handling distributed training. This episode will cover the core concepts: what NNX state is, how Orbax is structured, and the complete workflow for saving and restoring a single, basic NNX model.   
 
Resources: 
Learn more → https://goo.gle/learning-jax

Subscribe to Google for Developers → https://goo.gle/developers 

Speaker: Robert Crowe

Hands-on techniques for debugging compiled JAX and Flax NNX models: when to disable JIT, use jax.debug_nans, and inspect with NNX.display and NNX.so.

Practical Debugging for JAX & Flax NNX — disabled_jit & NNX Tools

Debugging JAX & Flax NNX (Part 2)

You know that debugging is crucial when doing any kind of software development. JAX and Flax NNX offer amazing performance and a clean API, but debugging can feel different due to JAX's compilation model. Today, we'll bridge the gap, exploring how to debug JAX and NNX effectively, relating techniques back to your PyTorch experience. We'll focus specifically on the Flax NNX API.  In this second episode of our three part series on debugging we’ll build on the first by introducing more advanced and specialized techniques, including how to temporarily disable JIT and how to leverage the powerful, built-in inspection tools provided by Flax NNX.  

Resources: 
Learn more → https://goo.gle/learning-jax

Subscribe to Google for Developers → https://goo.gle/developers 

Speaker: Robert Crowe

Archive