Portable Computer přináší lokální AI pro NVIDIA DGX Spark, která pracuje s citlivými soubory bez jejich odesílání do cloudu. Plánuje vícekrokové úlohy, propojuje dokumenty a volí cloud jen pro nutné kroky.

Local First AI v praxi: Portable Computer na NVIDIA DGX Spark

Portable Computer: lokální AI na NVIDIA DGX Spark má pracovat s citlivými soubory bez jejich odesílání do cloudu

Portable Computer brings Perplexity Computer to NVIDIA DGX Spark, so more of your work can run locally on your own machine.

In this Short Session, PM Joey Zhong shows how Portable Computer keeps the model orchestrator, planner, tool router, and local connectors on your desktop. It can read local files, take actions on your machine, and run dictation locally. Local work doesn’t use Computer credits.

Watch Portable Computer analyze a folder of private financial documents, including tax returns, 1099s, and brokerage statements. It reconciles records and calculates year-by-year returns after fees and taxes without sending those files to the cloud. When the task requires current market data or frontier reasoning, Portable Computer automatically escalates only that step to a cloud model with Perplexity Search, then brings the sourced results back into the local workflow.

You’ll see how Portable Computer can:

• Work across sensitive files stored on your machine
• Plan and execute multi-step tasks locally
• Analyze and reconcile financial documents
• Use local models without consuming Computer credits
• Escalate individual steps when cloud capabilities are needed
• Combine private local work with current, sourced web research

Short Sessions is a Perplexity series featuring product demos, quick takes, and perspectives on the future of AI.

Learn more about Portable Computer: https://www.perplexity.ai/hub/products/portable-computer

#PerplexityAI #PerplexityComputer #PortableComputer #NVIDIA #DGXSpark #LocalAI #AIAgents #PrivateAI #AgenticAI

Portable Computer demo | Short Sessions

Claude Fable 5.1 (Anthropic) cílí na komplexní práci: udrží souvislosti v dlouhém procesu, lépe odolá řetězení chyb a zvládá scénáře jako finanční modely, matematické důkazy i smlouvy s křížovými odkazy.

Claude Fable 5.1: AI pro dlouhé, vícekrokové úkoly

Claude Fable 5.1 přichází jako AI pro nejtěžší a dlouhé pracovní úkoly

Claude Fable 5.1 is the latest upgrade to our most capable class of models. It excels at complex, long-running tasks, and its research capabilities offer an early glimpse of how AI will contribute to scientific progress.

It excels at complex, multi-step work, like a long mathematical proof or a contract with hundreds of cross-references. It avoids shortcuts and is smart enough to fix the root causes of software issues.

Across our benchmarks, Fable 5.1 sets a new standard, and at lower effort levels it can achieve similar or better results than Fable 5 at a much lower cost.
Claude Fable 5.1 is available everywhere today.

Read more: https://www.anthropic.com/claude-fable-and-mythos-5-1

Introducing Claude Fable 5.1

Přečtěte si, jak NTT DATA nasadila Codex a dosáhla rychlého růstu na 10 000+ aktivních uživatelů. Automatizujte analýzu incidentů a tvorbu reportů až za 30 minut – při zachování odborného úsudku.

Jak Codex zrychluje incidenty v NTT DATA: od analýz k reportům za 30 minut

Co Codex odemyká pro NTT DATA: od dvoudenních analýz k výsledkům za 30 minut

After introducing Codex, adoption grew to more than 10,000 active users and became one of the largest internal community in just a few months. 

The NTT Data sales teams uses Codex to automate tasks like customer list maintenance, which has created net new workflows and report creation that used to take 2 days now takes 30 min 💥

What Codex Unlocks for NTT Data

Nejčastější otázky o přírodě v Olympijském národním parku: deštný les vs. oceán vs. hory, proč je rozklad pomalý, co je ghost pipe a jak bezpečně objevovat přílivové tůně (tide pooling).

Olympijský národní park: průvodce přírodou, ekosystémy a tide poolingem

Nejčastější otázky o přírodě zodpovězené v Olympijském národním parku

We’re celebrating National Park Week with a deep dive into Olympic National Park in Washington State. Join us as an official U.S. Park Ranger answers top-trending searches around National Parks and the great outdoors. We’ll explore tidepools, rainforests and glacial mountain ranges, looking out for the special flora and fauna that make the Olympics so unique. Put on your hiking boots, it’s the next episode of Searching For.

00:24 The Olympic Rainforest 
01:48 What is tidepooling? 
02:23 Are starfish predators? 
03:09 What causes glaciers to form? 
03:56 Which national park has the most wildlife? 
04:42 What do you do if you encounter a black bear? 
05:05 How many national parks are there? 

See what else we’ve been Searching For:
https://www.youtube.com/watch?v=dWH8n-FkprM
https://youtube.com/shorts/BodGtkvX3FY
https://youtu.be/zASY1bewxbs 
https://www.youtube.com/shorts/7l9kV1ndFS8 
https://youtu.be/I0V14dTS9JQ 
https://youtube.com/shorts/bDaPXwzEeqE

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

Footage and Graphics Courtesy of: 
Virginia Aquarium & Marine Science Center
National Parks Service 
Passport To Your National Parks® is a registered trademark of Eastern National. All rights reserved.

The world’s top nature searches answered in one epic national park | Searching For

Jak lokální AI agent (Hermes) běžící na NVIDIA RTX Spark pomáhá vývojářům při pádu služby: od zachycení incidentu přes prioritizaci až po přípravu změn a ověření opravy. Praktický pohled na nový pracovní postup.

Lokální AI agent pro incidenty: RTX Spark a rychlejší debugování

Debugování s lokálním AI agentem během kávy: jak NVIDIA RTX Spark mění pracovní postup vývojářů

This demo shows how a local Hermes agent running on NVIDIA RTX Spark can streamline software development by monitoring communication channels, prioritizing urgent issues, and coordinating debugging and QA.

RTX Spark is built to run demanding AI models and always-on agents locally and simultaneously. By offloading repetitive tasks, RTX Spark gives developers more time to focus on building.

Learn more about RTX Spark: https://www.nvidia.com/en-us/products/rtx-spark/ 

Follow NVIDIA RTX Spark on social:
X: https://x.com/NVIDIARTXSpark 
Instagram: https://www.instagram.com/nvidiartxspark/ 
TikTok: https://www.tiktok.com/@nvidiartxspark 
Facebook: https://www.facebook.com/NVIDIARTXSpark

Debugging with a Local Agent While You Get Coffee, Powered by NVIDIA RTX Spark

Jak Artemis II připravuje čtyřčlennou posádku na pilotovaný let kolem Měsíce: testování Orionu v extrémních podmínkách, role akustické komory, průběh mise a návrat k Zemi.

Artemis II v kostce: Orion, posádka a testy pro let k Měsíci

Artemis II: Co se děje za zavřenými dveřmi testovacího centra NASA

Ever wonder what happens behind the classified doors at NASA? Step inside a top-secret test facility as we answer the top trending questions about the Artemis II mission. Explore the extreme testing grounds, cutting-edge hardware, and meet the crew behind the historic journey back to the Moon. All systems go on the next episode of Searching For.

00:22 What is the artemis ii mission?
00:33 Who are the artemis ii astronauts? 
00:47 How long is the artemis ii mission? 
01:02 How big is artemis ii?
01:17 Why is NASA going back to the moon? 
01:27 How to watch the artemis ii space launch 

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

Go behind closed doors at the Artemis II test facility | Searching For

Jak se z měření na reálné platformě NVIDIA Vera Rubin (NVL72) od CoreWeave promítá počet tokenů za sekundu na megawatt do ekonomiky AI továren, výkonu a udržitelnosti. Čtěte.

Ekonomika AI infrastruktury: proč rozhoduje 10× tokenů na megawatt

Více tokenů za sekundu na megawatt: ekonomika platformy NVIDIA Vera Rubin

10X more tokens per second per megawatt as measured by CoreWeave on the NVIDIA Vera Rubin NVL72 platform.

The next generation of AI factories will be measured by how efficiently they turn compute into tokens, and tokens into revenue. Achieving that requires extreme co-design across the entire stack, paired with an open ecosystem built to innovate and scale.

NVIDIA’s Dion Harris breaks down 5 critical questions every AI infrastructure leader should be asking across compute, networking, storage, software, and security — and connects the technology behind Vera Rubin to the economics shaping the future of AI infrastructure.

Learn more:

https://www.cio.com/article/4209776/5-critical-questions-that-define-ai-factory-economics.html

https://blogs.nvidia.com/blog/vera-rubin/

More Tokens per Second per Megawatt: The Economics of NVIDIA Vera Rubin

Zjistěte, jak používat Computer Analytics v Perplexity: denní spotřebu kreditů, organizační pohled na adopci, žebříček uživatelů a jak z analytiky udělat praktické provozní rozhodování.

Computer Analytics v Perplexity: měřte adopci a náklady AI v organizaci

Jak používat Computer Analytics v Perplexity: přehled využití AI napříč organizací

Perplexity Computer Analytics shows exactly how Computer is being used across your organization — credit consumption, top users, most-used tools, and where the work is landing.

Computer Analytics has two views. Organization admins see analytics across every member of the org. Individual members and consumer users see their own usage. 

Admins open organization settings, then Analytics, then Computer, and get a dashboard that breaks credit usage down by day so spikes, trends, and anomalies stand out. A leaderboard surfaces the top Computer users in the org — the power users, the champions, and the people behind your next internal case study.

You'll learn how to:

-Open the Computer Analytics dashboard from organization settings

-Read daily credit usage and spot spikes, trends, and anomalies

-Use the leaderboard to find power users and internal champions

-Review artifacts generated, top connectors, top skills, top projects, and average task duration

-Filter across 7, 30, and 90 days and group usage by model

-Export any view for stakeholder decks, custom analysis, or finance review

-Enable the Computer Analytics API and generate an org-scoped key

Enterprise organizations can pull the same data programmatically through the Computer Analytics API. The endpoint returns credits, connectors, artifacts, skills, projects, workflows, and task durations as a bucketed time series, ready to flow into Snowflake, Hex, Metabase, a spreadsheet, or an internal finance system. The dashboard shows what happened. The API brings that data into the tools where planning decisions already happen.

Subscribe for more tutorials on Computer, Connectors, Skills, and agentic workflows.

🔗 https://www.perplexity.ai/enterprise
📚 Perplexity Academy Playlist: https://www.youtube.com/watch?v=yoD6LUiTuH8&list=PLKwRkjCH760NNEPqk75Ds6FfIMDmFwHIK

#PerplexityComputer #PerplexityAcademy #PerplexityAI #ComputerAnalytics #PerplexityEnterprise #PerplexityConnectors #PerplexitySkills #AIAnalytics #AIWorkflow

How to use Computer analytics | Perplexity Academy

Zjistěte, jak převést jednu jasnou otázku do obhajitelného strategického decku: propojení tržního kontextu, zákaznických dat, interních podkladů a lidskou kontrolou výsledku s ChatGPT Work.

Proměňte obchodní otázku v strategický deck s ChatGPT Work

Jak proměnit obchodní otázku ve strategický deck s ChatGPT Work

Turn a business question into a clear recommendation—without spending days gathering research and building slides.

In this demo, Arvind from OpenAI’s Strategy & Business Operations team shows how he uses ChatGPT Work to pull together market research, customer data, and past strategy work; propose a storyline; and create a leadership-ready strategy deck.

You’ll see how he reviews the sources and numbers, refines the recommendation, and approves the final readout.

https://chatgpt.com/work

How to Turn a Business Question Into a Strategy Deck With ChatGPT Work | Tutorial

Zjistěte, jak pomocí ChatGPT Work propojit roztříštěné podklady (produkt, messaging, výzkum i zákaznické poznatky) a vytvořit rychle kvalitní první návrh launch blogu pro další kontrolu a úpravy.

Jak urychlit první návrh launch blogu pomocí ChatGPT Work

Jak vytvořit kvalitní první návrh blogu s ChatGPT Work

Turn scattered launch materials into a solid first draft.

Sahil from OpenAI’s Marketing team shows how he uses ChatGPT Work to bring together product documentation, research, meeting notes, and messaging; follow an established blog template; and create a structured launch blog draft.

He then reviews, refines, and keeps the draft current as launch details change.

https://chatgpt.com/work

How to Create a Solid Blog Draft With ChatGPT Work | Tutorial

Zjistěte, jak ChatGPT Work pomáhá obchodníkům připravit brief před schůzkou a po ní vytvořit zápis, follow-up e-mail a aktualizace v CRM—rychleji, přehledně a pod lidskou kontrolou.

Jak zrychlit přípravu na schůzky se zákazníky v ChatGPT Work

Jak se připravit na schůzky se zákazníky s ChatGPT Work

Spend less time preparing for customer calls and following up afterward.

Alex from OpenAI’s Sales team shows how he uses ChatGPT Work to gather account context and create a meeting brief. After the call, ChatGPT Work drafts notes, a follow-up email, and proposed CRM updates for his review..

The result: less time managing the process and more time with customers.

https://chatgpt.com/work

How to Prepare for Customer Meetings With ChatGPT Work | Tutorial

Zjistěte, jak ChatGPT Work pomáhá neztrácet kontext, zvýraznit úkoly k rozhodnutí a navazovat na práci týmu. Čtěte praktické principy kontinuity a prioritizace.

Jak udržet plynulost práce v týmu s ChatGPT Work: upozornění na klíčové úkoly

Práce se může hýbat dál: jak ChatGPT Work upozorňuje na důležité úkoly a navazuje na práci týmu

ChatGPT Work keeps projects moving. Monitoring what matters, surfacing what needs attention, and picking up where your team leaves off.

Archive