Objevte, jak GPT‑5.3 Instant snižuje 'over‑caviating' — méně zbytečných caveatů, lepší čtení kontextu a plynulejší odpovědi při zachování bezpečnosti.

GPT‑5.3 Instant: méně zbytečných varování, přirozenější konverzace

Snižování "over‑caviating" v GPT-5.3 Instant

OpenAI researcher Blair Chen explains how GPT-5.3 Instant reduces unnecessary disclaimers, making ChatGPT more directly helpful and smoother to use.

Reducing Overcaveating in GPT-5.3 Instant

Podrobná recenze Pixel 10a: Gemini Live, Camera Coach, Auto Best Take, flush kamera, Satellite SOS a proč je tento odolný A‑Series telefon za $499 zajímavý pro každodenní použití.

Pixel 10a — AI supertelefon za $499: praktická recenze funkcí a bezpečnosti

Pixel 10a: $499 AI Superphone — podrobný průzkum

The Google Pixel 10a is officially here! Join Product Manager Laura Castilani as she reveals how Google re-engineered the A-Series to include a zero-millimeter flush camera bar and flagship AI features. 

In this episode of the Made by Google Podcast, we dive into the Pixel 10a’s massive battery upgrades, the arrival of Satellite SOS, and why this is our most durable and sustainable phone yet. From the vibrant "Berry" color to 7 years of software updates, discover why the Pixel 10a is the ultimate $499 powerhouse. 

00:00 - Introduction to Pixel 10a
02:14 - Redesigning the Camera Bar
03:44 - Advanced AI: Gemini Live & Photography
06:47 - Safety and Longevity
10:19 - Sustainability and Vibrant New Colors

#Pixel #MadeByGoogle #TechPodcast 

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 $499 AI Superphone: Pixel 10a Deep Dive | Made by Google Podcast S9E1

Poznejte 'screenshot strategii' Onera Biberkoku a zjistěte, jak nástroj Flow umožňuje iterativní AI filmařství, lokalizaci a demokratizaci tvorby bez štábu.

Screenshot strategie a Flow: Jak AI mění filmařství pro tvůrce

Oner Biberkoku o AI filmařství s Flow | In the Flow

Filmmaker https://www.youtube.com/@Oner_Biberkoku shares his "screenshot strategy" for building a full narrative from a single starting point during this In the Flow interview. Thanks Oner for the partnership!

"At the end of the shot, you can take a screenshot, give it again [to Flow], and you now have a new image of that character in a different situation... So I think, with one image, you can stay in the flow, and make a story."

Try Flow for yourself: https://flow.google  
Learn how to create in Flow: https://youtu.be/9nVEfjmDlVk?si=DKlab6LdBgjaOCX4 
See our Flow playlist on YouTube: https://youtube.com/playlist?list=PL590L5WQmH8cZy80Ko7dzpH2MyvNJFE-p&si=JziPldokD6tevISg 
Missed earlier episode of In the Flow? Watch now: 
https://youtu.be/xjXhJcWZkDE
https://youtu.be/zsyqt_KnYR8
https://youtu.be/nkX8j9472lg
https://youtu.be/oRqEL0t67fo
https://youtu.be/Bb0oFbZ35qM
https://youtu.be/hZSDKiM8LvA

#Flow #AIFilmmaking #GenerativeAI #GoogleDeepMind 

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

Oner Biberkoku on AI Filmmaking with Flow | In the Flow

Objevte, jak prompt caching snižuje latenci a náklady AI aplikací — KV cache, prahy, prompt_cache_key a konkrétní tipy pro vývojáře pro lepší cache hit rate.

Praktický průvodce prompt caching: zrychlení a úspora nákladů u AI

Build Hour: Prompt Caching — jak zrychlit a zlevnit AI aplikace pomocí kešování promptů

Build faster, cheaper, and with lower latency using prompt caching. This Build Hour breaks down how prompt caching works and how to design your prompts to maximize cache hits. Learn what’s actually being cached, when caching applies, and how small changes in your prompts can have a big impact on cost and performance. 

Erika Kettleson (Solutions Engineer) covers:
• What prompt caching is and why it matters for real-world apps
• How cache hits work (prefixes, token thresholds, and continuity)
• Best practices like using the Responses API and prompt_cache_key
• How to measure cache hit rate, latency, and token savings
• Customer Spotlight: Warp (ttps://www.warp.dev/) led by Suraj Gupta (Team Lead) to explain the impact of prompt caching


👉 Prompt Caching Docs: https://platform.openai.com/docs/guides/prompt-caching
👉 Prompt Caching 101 Cookbook: https://developers.openai.com/cookbook/examples/prompt_caching101
👉 Prompt Caching 201 Cookbook: https://developers.openai.com/cookbook/examples/prompt_caching_201
👉 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
02:37 Foundations, Mechanics, API Walkthrough
12:11 Demo: Batch Image Processing
16:55 Demo: Branching Chat
26:02 Demo: Long Running Compaction
32:39 Cache Discount Pricing Overview
36:03 Customer Spotlight: Warp
49:37 Q&A

Build Hour: Prompt Caching

Jak fyzická AI kombinuje velké modely, simulace a digitální dvojčata pro bezpečné, rychlé a škálovatelné nasazení robotů ve výrobě, stavebnictví a autonomní dopravě.

Fyzická AI v průmyslu: simulace a digitální dvojčata mění robotiku

Co bude dál v robotice? Jak fyzická AI, simulace a digitální dvojčata mění průmysl

By combining decades of real-world data with advanced AI, simulation and digital twins, teams are rapidly training, validating, and deploying intelligent robotics at scale.

Physical AI is reshaping how intelligence moves from data into the real world with leaders like AgiBot, Franka Robotics, Caterpillar, and PlusAI driving the transformation. 

From construction and manufacturing to autonomous operation, this fusion of simulation and AI is unlocking new capabilities for industrial machines and redefining what robotics can do in the physical world.

Explore NVIDIA Robotics: https://nvda.ws/4rYamRr
Build with NVIDIA Isaac: https://nvda.ws/4cx3llY
Unlock world foundation models with Cosmos: https://nvda.ws/4cw5uhN
Learn more about Physical AI: https://nvda.ws/4qJBSRF 

Thumbnail Image Courtesy of Caterpillar

What’s Next in Robotics?

Naučte se, jak Lyria 3 v Gemini generuje plně aranžované soundtracky z textu nebo nahrávky. Praktické tipy na prompty, nastavení parametrů a použití v videích, podcastech a hrách.

Vytvořte soundtrack za pár sekund s Lyria 3 v Gemini

Lyria 3 | Generování hudby přichází do Gemini

Create custom soundtracks in Gemini 🎼Describe your vision or upload from your camera roll and watch a track come to life in seconds. The soundtrack to your imagination is just a prompt away. Explore the new music generation feature in Gemini. 

https://gemini.google.com      

#Lyria #Lyria3 #GoogleGemini

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

Lyria 3 | Music Generation comes to Gemini

Proč 'cost per token' nahrazuje FLOPS a jak extreme co‑design (hardware, síť, software) snižuje náklady na inference u LLM a MoE modelů.

Snižte cost‑per‑token: Extreme co‑design pro efektivní tokenomics a škálovatelnou AI

Extreme Co-Design pro efektivní tokenomics a škálovatelnou umělou inteligenci

As AI moves into the era of real-time reasoning, performance alone is no longer enough. The true measure of AI at scale is efficient tokenomics—how much it costs to generate each token of intelligence.
Reasoning models like mixture-of-experts (MoE) architectures generate massive volumes of tokens, improving answer quality but placing simultaneous pressure on compute, memory, networking, storage, and software. In this new paradigm, the hidden costs of communication and routing matter just as much as raw FLOPS.

This video explores why extreme co-design—engineering the entire stack as one unified system—is the key to lowering cost per token and maximizing AI ROI.  

You’ll learn:
- Why cost per token is becoming the defining metric for reasoning AI
- The networking and communication challenges behind MoE inference
- How rack-scale systems like GB200 NVL72 deliver breakthrough token efficiency
- How Blackwell and Ruben integrate silicon, interconnect, networking, and software to power AI at scale

Featuring insights from NVIDIA, Signal65 Microsoft Azure, and CoreWeave, this discussion makes one thing clear: End-to-end system design is the most powerful lever for delivering efficient tokenomics and scaling reasoning AI.

Extreme Co-Design for Efficient Tokenomics and AI at Scale

Jak Codex provádí self‑verification: spouští testy, nasazuje aplikaci a kontroluje logy, aby refaktoring nezpůsobil ztrátu observability a zkrátil rutinní práci.

Codex ověří kód a ochrání observability při refaktoringu

Codex zkontroluje svou práci za vás

Javi walks through a logging refactor and shows why Codex's self-verification is a step change: the model runs the app, finds the right session, and proves logs still flow.

Takeaways:
- Codex can validate its work by running tests and launching the app.
- It excels at broad refactors that touch many files.
- The model can find session IDs and query tools on its own.
- Verification collapses a risky manual loop into minutes.

When the agent can prove correctness, you can move faster with less risk.

Chapters:
00:00 Why Codex has been a step change
00:18 Self-verification: run tests and launch the app
00:52 The task: a logging refactor across many files
01:10 The risk: do not break observability
01:28 How this used to be verified manually
01:35 Ask the model to verify logs end-to-end
01:50 It finds the session ID and queries logs MCP
02:03 Proof: logs still pipe, task done fast

Codex checks its work for you

Deep research v ChatGPT nyní běží na GPT‑5.2: cílené prohledávání aplikací/domén, sledování průběhu v reálném čase, fullscreen reporty a přesnější citace. Postupné nasazení uživatelům.

Deep research v ChatGPT přechází na GPT‑5.2 — rychlejší a přesnější rešerše

Aktualizace deep research v ChatGPT: Deep research nyní poháněno GPT-5.2

Deep research in ChatGPT is now powered by GPT-5.2.

Now in deep research you can:
- Connect to apps in ChatGPT and search specific sites
- Track real-time progress and interrupt with follow-ups or new sources
- View fullscreen reports

Rolling out starting today.
https://chatgpt.com/features/deep-research/

Updates to deep research in ChatGPT

Objevte fun-first strategii Allie K. Miller: jak AI nasadit hravě a efektivně — ne jen pro úsporu času, ale pro lepší práci, nové možnosti a rychlé experimenty.

Fun-first přístup k AI: praktický návod od Allie K. Miller

The Fun-First Framework for AI Adoption s Allie K. Miller

Allie K. Miller has a rare superpower: she makes the cutting edge of AI feel clear, practical, and genuinely fun. Recently named to TIME’s list of the most influential people in AI, Allie has become one of the internet’s most trusted guides—breaking down new models, tools, and workflows for millions, from curious beginners to seasoned builders. In this episode, she joins Cameron Adams to unpack what’s coming next: agentic browsers, “AI-first” reinvention, and why productivity alone can be a trap. Along the way, she shares a simple prompt structure that unlocks better outputs, how to turn fear into agency, and why creativity is often the fastest on-ramp to adoption. Plus: a delightfully weird prediction about how we’ll perceive the world.

Cam's notes on Substack: promptedwithcam.substack.com

Timestamps: 
00:00 - Introduction
01:45 - The Prompt
04:07 - Creative All Hands
13:39 - Use AI to simulate future scenarios when convincing teams to change.
15:30 - Let small teams move ahead instead of waiting for company-wide adoption.
17:37 - Use AI to enable work that was previously impractical to do manually.
20:23 - Use creative AI tasks as an entry point before work use cases.
23:32 - Encourage skeptics to try AI even if they expect to dislike it.
24:40 - Ask permission before introducing AI into group conversations.
26:21 - Try this prompt structure: who you are, your objective, then context.
28:07 - Add goals, fears, constraints, and resources to improve AI responses.
31:29 - Photograph handwritten notes or post-its and upload them as AI context.
43:29 - Use AI to analyze patterns across multiple past conversations.
52:10 - Turn one detailed input into multiple formats like decks or videos.
1:01:08 - Design Reveal
1:02:48 - Key Takeaways

Links
Subscribe to the Prompted Newsletter: https://promptedwithcam.substack.com/ 
Cam on LinkedIn: https://www.linkedin.com/in/themaninblue/ 
Allie on LinkedIn: https://www.linkedin.com/in/alliekmiller/ 

Link to Allie, Tisha and AI’s Creative Collaboration: https://promptedpodcast.com/episode-9

The Fun-First Framework for AI Adoption with Allie K. Miller

Praktický postup pro PM: jak pomocí Codex skills rychle najít příčinu selhání v Buildkite, nainstalovat token a vyřešit PR selhání bez zbytečného pingování vývojářů.

Codex pro produktové manažery: rychlá diagnostika CI bez rušení inženýrů

Jak produktoví manažeři používají aplikaci Codex

Alexander Embiricos (a Product Manager on the Codex team) shows how he uses Codex skills to make a small product change, diagnose a Buildkite failure, and improve the skills so the next PR goes faster.

Takeaways:
- Skills are a shortcut for repeated workflows like Buildkite logs.
- When a skill fails, fix the root cause and update the skill.
- The real win is compounding: the codebase gets easier over time.

This is the loop: ship the fix, then teach the workflow.

Chapters:
00:00 PM context: careful before tagging engineers
00:16 A confusing button and a quick team check
00:32 Delete the button, then hit a PR failure
00:46 Use the Buildkite skill instead of digging through logs
01:09 Install the Buildkite token
01:20 Update the skill so it works next time
01:53 Inductive loop: feedback - fix - improve the skill
02:06 Compounding payoff: Codex gets better over time

How PMs use the Codex app

Rozhovor s Asadem Awanem o tom, jak ChatGPT zavádí reklamy bez ovlivnění odpovědí: technické a vizuální oddělení, ochrana soukromí a ad‑free pro předplatitele.

Reklamy v ChatGPT: jak zachovat důvěru, soukromí a design

Epizoda 13 – Myšlení za reklamami v ChatGPT

How should advertising work in an AI product? Asad Awan, one of the ad leads at OpenAI, walks through how the company is approaching this decision and why it’s testing ads in ChatGPT at all. He explains how ads are built to stay separate from the model response, keep conversations with ChatGPT private from advertisers, and give people control over their experience.

Chapters
00:00:29 — Mission and principles
00:04:01 — Separation between ads and answers
00:07:31 — Who will see ads
00:08:52 — Internal input and decision-making process
00:11:06 — Controls and how ads will work
00:15:53 — Guardrails for sensitive conversations
00:17:33 — Skepticism about ads
00:20:26 — Helping small businesses
00:24:13 — Future of ads

AI and Technology News

Latest