Přehled z OpenAI Podcastu: jak se trénují a hodnotí zdravotnické AI modely, role kliniků a HealthBench, a jak AI bezpečně zlepšuje péči pro pacienty i lékaře.

AI ve zdravotnictví: bezpečné modely, trénink a klinické přínosy

Budování AI pro lepší zdravotnictví — OpenAI Podcast Ep. 14

Healthcare systems around the world are under strain, and both patients and clinicians are feeling the impact. OpenAI's Head of Health Dr. Nate Gross and Karan Singhal, who leads Health AI Research, discuss how AI can help address the biggest challenges. They cover how OpenAI is training models to handle sensitive health questions in collaboration with physicians, and how that foundation is unlocking a new generation of tools for patients, clinicians, and healthcare systems.

Chapters:
00:00:38 – Origins of Nate and Karan’s interest in AI and healthcare
00:05:01 – Strategy for building AI tools for clinicians
00:06:57 – How AI models are trained for health use cases
00:10:15 – How OpenAI is able to score well on health evals
00:14:21 – Key challenges deploying AI in healthcare
00:21:05 – Collaboration with hospitals and healthcare systems
00:23:05 – Practical everyday uses of AI health assistants
00:26:43 – Biggest “wow” moment during development
00:28:46 – Feedback from clinicians and early users

Building AI for better healthcare — the OpenAI Podcast Ep. 14

Nová beta funkce Skills pro ChatGPT Business a Enterprise: opakovatelné workflowy, integrace s kódem a API, automatické použití v konverzacích a doporučené postupy pro bezpečnost a governance.

Jak Skills v betě mění ChatGPT pro firmy — přehled a výhody

Funkce Skills v betě pro ChatGPT Business a Enterprise

We’re introducing Skills in beta for ChatGPT Business and Enterprise.

Skills are reusable, shareable workflows that tell ChatGPT exactly how to do a specific task, enabling ChatGPT to do that task better and more consistently. A skill can bundle instructions, examples, and even code. Once a skill has been created and installed, ChatGPT can automatically use a skill (or multiple skills) when it’s helpful.

A skill can include instructions and supporting resources you want ChatGPT to use whenever you ask it to do a specific task. And for more structured work, skills can include a set of steps that run the same way every time.

Skills are available in beta in ChatGPT Business and Enterprise. Skills are also supported in Codex and the API. While they don’t sync across products yet, OpenAI skills follow the Agent Skills open standard—so you can download them from one product and install them in another.

ChatGPT skills in beta for ChatGPT Business & Enterprise

Analýza GTC pregame: proč krátké představení před keynote nastavuje narativy o AI, jak AI může generovat obrovskou ekonomickou hodnotu a jak regulace a geopolitika změní trhy.

GTC Pregame: Jak předzápas formuje budoucnost AI, regulací a byznysu

NVIDIA GTC Live Keynote Pregame: Super Bowl technologické transformace

As we gear up for what’s next, we’re rolling the highlights. Relive the biggest moments, boldest takes, and behind-the‑scenes energy that made NVIDIA GTC Live pregames a fan favorite.

Stay tuned for the 2026 line-up on: hhttps://nvda.ws/3MI8Ljw

🗓️ Monday, March 16 | 8 a.m.–11 a.m. PDT

#NVIDIAGTC

NVIDIA GTC Live Keynote Pregame Teaser

Google otevřel Platform 37 v King's Cross: The AI Exchange — veřejné i výzkumné centrum spojující vývoj AI, vzdělávání a komunitu, inspirované tahem Move 37.

Platform 37 v King's Cross — Google otevírá The AI Exchange

Platform 37: Google otevírá nové centrum pro inovace v oblasti umělé inteligence v Londýně

Today, we’re unveiling the name of Google’s landmark King’s Cross building in London, Platform 37, and plans for The AI Exchange, a new public space dedicated to deepening understanding of artificial intelligence.

The name is a nod to the building’s location next to Kings Cross station—and to ‘Move 37’, a pivotal play made by DeepMind’s AI system AlphaGo in a now legendary 2016 match against Go world champion Lee Sae Dol. Go is incredibly complex, with more possible board configurations than the number of atoms in the known universe, and has long been a proving ground for AI research. 

Platform 37 is Google’s most ambitious and innovative building to date. Learn more: https://blog.google/company-news/inside-google/around-the-globe/google-europe/united-kingdom/platform-37-the-ai-exchange/

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

Platform 37: Google’s New AI Innovation Space in London

Zjistěte, jak Rakuten integroval OpenAI Codex do monitoringu a CI/CD, zkrátil MTTR až o 50 % a zrychlil nasazení. Praktické workflowy a lessons learned.

Jak Rakuten zkrátil MTTR o 50 % nasazením Codexu — případová studie

Rakuten opravuje problémy dvakrát rychleji díky Codexu

Rakuten operates one of the world’s largest digital ecosystems across e-commerce, fintech, and mobile services. Shipping fast without sacrificing reliability is critical.

Teams use OpenAI Codex to support incident response, code review, and development workflows. The agent works inside monitoring systems and CI/CD pipelines to diagnose issues, review changes against internal standards, and advance projects from partial specifications.

The result is faster recovery and faster shipping. Rakuten reports up to a 50% reduction in mean time to recovery, and projects that once took a quarter can now ship in weeks.

As Codex handles more generation, engineers focus on clearer specifications and rigorous verification. The shift helps teams move quickly while keeping safety built into every release.

Learn more: https://openai.com/index/rakuten/

Rakuten fixes issues twice as fast with Codex

AIMS ukazuje, že AI jako „druhý čtenář“ v mamografii může snížit zátěž radiologů, zkrátit čekání pacientek a zachovat kvalitu detekce.

Jak AI zrychluje mamografický screening a snižuje zátěž radiologů

Pomoc radiologům při odhalování rakoviny prsu pomocí AI

Breast cancer affects 1 in 8 women worldwide, and in the UK, a shortage of radiologists puts immense pressure on staff in radiology departments to deal with the backlog of people waiting for scans. To help reduce the burden on resources and staff, new ways of screening need to be investigated. 

Google partnered with Imperial College London and three NHS trusts on the Artificial Intelligence in Mammography Screening (AIMS) study to explore how an AI reader, integrated in place of a second human reader in the UK’s existing two reader plus arbitration system, performs in a real-world setting.

Results of the AIMS study, recently published in “Nature Cancer,” showed that an AI reader could successfully integrate in to the UK’s existing system in place of a human reader, reducing the overall workload of radiologists and allowing them to focus their time and care on patients who needed it the most. 

Note: The patients featured in this video did not participate in the study and are sharing their personal experiences with the mammography process. They were compensated for their time.

To learn more about the AIMS study, visit https://www.nature.com/articles/s43018-026-01128-z   

To learn more about how AI can improve breast cancer detection in the UK, check out The Keyword: https://blog.google/innovation-and-ai/technology/health/google-ai-breast-cancer-detection/

Subscribe to our Channel: https://www.youtube.com/google
Find us on X: https://www.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

Helping Radiologists Detect Breast Cancer with AI

Jak Codex desktop, GPT‑5.4 a nová API primitiva (hosted shell, code mode, WebSocket) umožní delegaci úloh agentům a jak Agent Legibility Score zlepší čitelnost repozitářů.

Připravte repozitář a tým pro agenty — Codex, API a Agent Legibility Score

Build Hour: API & Codex — jak připravit kód a tým pro agentní delegaci

Build agent-powered workflows for real engineering work with Codex and OpenAI APIs. In this Build Hour, you'll learn how teams are moving beyond pair programming to agentic delegation, where AI systems can take on entire engineering tasks from planning to execution. 

Charlie Guo (DevEx), Ryan Lopopolo (Future of Work), Mitch Troyanovsky (Cofounder, Basis) cover how to:
• Ship new features faster using Codex, with real examples from Basis: https://www.getbasis.ai/
• Build agent-driven systems using new API capabilities like hosted shell, skills, and websockets
• Evaluate and improve agents using the Agent Legibility Score framework
• Apply Harness Engineering techniques to make agent workflows reliable in production
• Use GPT-5.4 for large-context reasoning, computer-use tasks, and tool-based agent workflows
• Reduce manual overhead by turning complex logic into reusable agent-driven flows

👉 Codex docs: https://developers.openai.com/codex
👉 API docs: https://developers.openai.com/api/docs
👉 Harness Engineering Blog: https://openai.com/index/harness-engineering/ 
👉 GPT-5.4 Blog: https://openai.com/index/introducing-gpt-5-4/ 
👉 Follow along with the code repo: http://github.com/openai/build-hours
👉 Sign up for upcoming live Build Hours: https://webinar.openai.com/buildhours

00:00 Introduction
01:58 What's New: Codex
05:01 What's New: API
08:33 Demo: Agent Legibility Score
21:53 Harness Engineering
32:32 Customer Spotlight: Basis
47:57 Q&A

Build Hour: API & Codex

Příběh Dr. Marian Croak: od dětské zvědavosti k Bell Labs a Google. Jak její VoIP a text-to-donate proměnily komunikaci, humanitární pomoc a vedení.

Marian Croak — průkopnice VoIP, sms-darování a lidsky orientované inovace

Shaping the Limitless Possibilities of Dr. Marian Croak

The Art of Possible is a documentary tracing the legacy of Dr. Marian Croak, a pioneer whose career was shaped by family, colleagues, mentors, and collaborators. These formative relationships and moments drove her innovation forward, establishing Dr. Croak as the crucial link between foundational technology and the future of responsible AI. She pioneered Voice over Internet Protocol (VoIP)—the technology powering every internet call today—holds over 200 patents, and invented critical applications like text-to-donate technology. Now, as VP of Responsible AI at Google, she continues to ensure that advanced research is built on ethical foundations for the benefit of humanity. Join us as we celebrate Dr. Croak.

0:52 – A Foundation of Curiosity
3:09 – Innovating for Social Good
6:45 – Mentorship and the Next Generation

A Creative Theory Agency Production

Directed by
Trivell Miller

Produced by
Alexandra César

Executive Producer
Rachael Hardy

Creative Producer
Ciara Montgomery

Producer
Hannah Strickland

Director of Photography
Othello Banaci

Production Manager
Jose Brown

Associate Producer
Ashlee Green

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

Přehled, jak umělá inteligence, datová centra a robotika mění zdravotnictví, zemědělství a zásobovací řetězce — klíčové přínosy, výzvy a doporučení pro přípravu na éru AI.

Jak AI a datová centra přetvářejí zdravotnictví, zemědělství a průmysl

Éra AI inovací | Kde žije internet

This trailer previews the mini-documentary series for Season 5 of the Webby Award-winning Where the Internet Lives.  

Season 5 of the series focuses on how advances in AI and data center infrastructure bring solutions in every sector, from medicine to creativity. 

Check out the full series: https://www.youtube.com/playlist?list=PL590L5WQmH8dLLdbnJ775gPNvMeVWHLxT

Learn more about Google Data Centers here: https://datacenters.google/ 

#AI #Data #WhereTheInternetLives

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

The Era of AI Innovation | Where the Internet Lives

Objevte, jak NVIDIA AODT vytváří fyzikálně přesná digitální dvojčata rádiových sítí — ray tracing, EM simulace, AI agenti a škálování v cloudu pro rychlejší a bezpečnější nasazení.

NVIDIA AODT: fyzicky přesná digitální dvojčata pro 6G sítě

Jak NVIDIA AODT (Aerial Omniverse Digital Twin) posiluje digitální dvojčata sítí

With NVIDIA Aerial Omniverse Digital Twin, developers design, validate, and iterate systems in simulation, before they enter the real world.
The platform is modular, designed for an ecosystem, and is already delivering.

Learn more: https://developer.nvidia.com/blog/5-new-digital-twin-products-developers-can-use-to-build-6g-networks/

How NVIDIA AODT Powers Network Digital Twins

Project Rio využívá NVIDIA Isaac k vytvoření fyzikálně přesného digitálního dvojčete, teleoperaci, Minik‑Gen a Cosmos Transfer pro bezpečnou, škálovatelnou automatizaci v nemocnicích.

Project Rio: Jak NVIDIA Isaac vytváří trénovatelné nemocnice

Building Hospital Automation Using NVIDIA Isaac pro zdravotnictví: Jak Project Rio mění nemocnice

Hospitals today face chronic staff shortages and increasing patient demand. Project Rheo is a blueprint for smart hospital automation, using physical AI and advanced simulation to train robots that can safely work alongside doctors and nurses. Robots are trained in high‑fidelity simulation with NVIDIA Isaac tools before ever touching real patients, improving safety, efficiency, and capacity in smart hospitals

Follow us:
X: https://nvda.ws/4hi2sxB
LinkedIn: https://nvda.ws/4eSlE3B
Subscribe to the NVIDIA AI Healthcare Newsletter: https://nvda.ws/4eS1F54

Building Hospital Automation Using NVIDIA Isaac for Healthcare

Poznejte Kosmos — systém AI-agentů využívající Nematron a NVIDIA (CUDA, cuML/cuDF) k zkomprimování šestiměsíčního výzkumu do jednoho dne. Aplikace v genetice, neurovědách a návrhu léčiv.

Kosmos: Jak AI od Edison Scientific zrychlí vědecký cyklus na jeden den

Jak Kosmos od Edison Scientific provede šestiměsíční výzkum za jeden den pomocí Nemotronu

Scientific research is slowing under the weight of papers, code, and complex data, so Edison Scientific built Kosmos, an AI scientist on NVIDIA that compresses six months of work into a single day. In this video, see how Kosmos uses NVIDIA GPUs and foundation models to read thousands of papers, run large-scale analyses, and turn evidence into new hypotheses across genetics, neuroscience, and drug discovery.

Discover how NVIDIA Nemotron™ Parse, cuML, and NeMo RL power Kosmos’ literature, analysis, and molecule-design agents—and what this AI stack means for the future of scientific labs.

Learn more: https://developer.nvidia.com/case-studies/scientific-literature-ai-nvidia-nemotron

#AIScientist #AgenticAI #NVIDIAGTC

Archive