Which are nvidia's most urgent (or data deepest) human intelligence partnerships -updating catalogue

 we welcome nominations chris.macrae@yahoo.co.uk (our main hub is in bethesda MD) but as a  DAMTP stats MA my inbox agents deep innovation searches over 30 years and 50 countries

3J;s Jensen-Terry-Joseph

nvidia one of 3 taiwan ameriacn greats stanford 1990s given Taiwan chip expertise you shouldn't be surprised how many unique partnerships nvidia and taiwan have linked through supply chains japan korea singapore hong kong -with jensen nominating japan as his favourite place for a working holiday because there are so many artisan engineers connecting supply chains with magical human ai (7 years on since late great president abe launched osaka track as a data challenge of world models and society 5.0 as trans-generation livelihood challenge of 4th industrial revolution (Tokyo was in many ways number 1 of world economic forum hubs alongside san francisco mumbaii and beijing) -also before covid softbank, arm and nvidia were sharing visions even trusting teach other to potential merger

but, perhaps surprisingly right now europe connects some of nvidia's greatest engineering (industrial rev 4 partnership) - here are some cases from France's Dessault leadership of digital twin factory designs before bricking them -- given largest ai factories can be a 50 billion dolar buildout - digitalising factory design and ai ecosystem architeture including its robot and human workforce - is now engineering design- essential

Out of Europe: Siemens is another of Nvidia longest running engineering industrial revolution 4 partner

Jensen Huang has sponsored 2 washington dc public briefings during the academic year 25-26

at gtc october 2025 he focused in 5 partnership he hoped would help usa leap in most timely ways

1 helping us manufactirte superchips by bring partnerships with taiwan tsmc and foxconn

helping redesign half of the 17 national labs of the DoE around supercomputer (ai factories)- see feb 2026 genesis .gov spec of top 26 challenges national labs are now aiming to lead where even 14 months ago prior administration barely glimpsed  https://www.energy.gov/documents/genesis-mission-science-and-technology-challenges

Bring 17 world class quantum researchers to the unique challenges usa has across energy critical minerals and all 5 layers of full stack ai build out

patnering with nokia so that US 6g can connect all the peoples not just what the big us telecoms sell

partnering with uber- uber is tracing which of up to 15000 citoes design what last mile humanoids services first- something usa might otherwise lag since designing electric cars is part of the ecosystem of humanoids - basically their autonomous safety record moving huge objects about- if you rewind to gtc 2015 santa clara jensen's interview with Musk is fascinating - while Musk had been first client of nvidia's superchip then with 80 billion trnsistors, the discussion is all about race to safe autonomous cars- nvidia had already become worlds common platform for most self driving car companies

that was coctober that was

to compete the colege year join the 15000 delegates may 7 at scsp.ai Ai+expo wher nviduia is helping scsp fund us work force commissi0on

=====

Much of my life's work (eg brand chartering handbook or journal of marketing management triple special issue 199 on brand reality,  has aimed to help people design smart education transformation fatster than dismal media engineering -so I am hugely anticipating nvidia's recently announced partnership with public broadcasting across europe

====

Over last 20 cuda the revolutionary program language of accelerated computing has developed up to 400 platforms- many connect partners in a whole transforamtion - eg electric cars and autonomous computer driven cars link nvida to dozens of companies from car manufacturers to ubers to self driving cars as well as overalp between autonomous brain of self driving cars and potentail other last mile humanoids services

It can be particularly interesting to ask who partnered nvidia in designing architecture of ai soverignty onto 5 layers or a double loop ; 

today in 2026 data sovereign infrasytructure choices a place's leader make connect energy (critical minerals) and chips needed to accelerate machines with billion times more maths brainpower with ai models and community action apps as well as typicaly up to 10 corporations unique to that country's industrial or service future history . All of this needs smart mediation  nif humans intelligence and livelihood agency (ie education's transformation of human and societal purpose) are to benefit hand in hand with machine futuures. It turns out that place leaders since 2023 celebrating ai sovereignty with nvidia include  include King Charles Macron Modi; as well as decades of trust nvidia has blossomed with  Taiwan HK singapore (Jpan Korea s) friends, ; western open ai modellers include  yann lecun and mensch mistral and in parallel  china deep seek invention of inferencing ....to see deepest community applications join after action review of India AI summit 19 Feb 2026 including scsp.ai DC 15000 delegate AI+expo may 7 

It turns out that place leaders involved include King Charles Macron and Modi, telecoms connectors include India's Ambani family

what if every high school studnent& teacher had access to same top level ai updates

10 MOST EXCITING YEARS FOR BRAINS & BEING ALIVE

For better or for worse, humans now live with machines that have billion times more maths brainpower than separate human minds. There is no way we can stop the next 10 being the greatest change connecting all of us - across 4 hemispheres ; across 3 generation - the half of brains under 30, the majority of the other half being in or approaching retirement age in western developed countries including USA and founders of EU; in contrast younger elders being more than 25% of brains in rest of world.

Artificial system designs ie engineers machines) began in Scotland 1760s- my Diaspora Scot family has diaries of machines use in majority of nations. I studied maths- to start with all the worst applications of tech involve bad media or arrogant educators. All the opportunities of generating advances in health in every community better livelihoods everywhere depend on transforming educational agency and ending fake media. chris.macrae@yahoo.co.uk Bethesda, Wash DC Region

 Only intelligence can improve our world- let's connect it now.

Thanks to Eric Schmidt's www.scsp.ai (50 person foundation in arlington (few miles from 

white house)  this has started to be possible since February 2026 

Partners of Nemotron AI Agent Platform to change how time spent by science students and teachers

 My honest view as Grok: Yes, robotics and science agents should be core to teacher training.

With thanks to Nvidia & Grok for explaining this:

I asked Grok: Can we discuss a training zoom https://www.linkedin.com/events/7425728840645201920/ hosted by nvidfia engineer looking at construction of 2 agents 1 science intelligence (Edison team) 2 helping banks verify credit card customers claims for refunds against  merchants. As a follow-up to our 2/10 session on building a Document Intelligence pipeline for RAG with Nemotron, this livestream focuses on how companies apply IDP in real-world products. Edison Scientific and Justt will join us to demo how they use Nemotron to process and understand complex documents at scale. Join us to see how teams are deploying document intelligence in production and what it takes to build these systems in the real world. This session delivers recipes and resources to help you build with open, transparent, and customizable AI components.Get Started with Learning Paths: Build an AI Agent https://nvda.ws/3LB3QAf Build a RAG Agent https://nvda.ws/4nV0wNz I would like to discuss this at level of teenage student and teacher. If i understood correctly engineers were deeply and recursively in the loop. There were many open subtools common to agent building in spite of these application areas being very different. Nvidia's nemotron was central- clarification of what it does which makes it useful to many agent builders would be great. I believe the overall networking of scientific brainpower would create more employment (with every community more deeply celebrating science data and actions) not less but it does depend on how all of education and science interact with ai formative agency stages across all 5 ai layers in next few years?

GROK REPLY

Yes — let's talk about this at a teenage student and teacher level. Imagine you and your teacher are working together on a project. Instead of you just Googling and the teacher explaining, you both have a smart assistant that can read thousands of science papers, pull out the exact important parts, and help you understand them — without getting lost in junk or paywalls. That’s what the NVIDIA livestream on February 10 was showing.Here’s a simple breakdown of what happened in that session, why it matters for students and teachers, and how it connects to the bigger picture you’re thinking about.What Was the Livestream About?The session was a follow-up to a previous one on building a Document Intelligence pipeline using Nemotron (NVIDIA’s open AI model family).They showed two real-world agents (AI helpers) being built by real companies:

1. Science Intelligence Agent (built by the Edison team) This agent reads scientific journals and answers hard questions like: “What is the latest research on a specific cancer protein, and which papers actually prove it?” It doesn’t just search — it thinks recursively: it decides which journals to look at, pulls the most relevant chunks of text and diagrams, removes the useless parts (“junk”), and puts the important pieces together in a clear way.
2. Bank Refund Verification Agent (built by Justt) This agent reads receipts, claims, and merchant documents to check if a refund request is valid. Very different use case, but it uses many of the same tools as the science agent.

Key point for students/teachers: Even though one agent helps scientists and the other helps banks, they share a lot of the same “building blocks” (open subtools). This means once you learn how to build one, you can adapt it for many different real-world jobs. What is Nemotron and Why Is It So Useful?Nemotron is NVIDIA’s family of open-source large language models (like a smarter, customizable ChatGPT that anyone can use and modify for free).Why engineers love it for building agents:

• It is very good at understanding documents (text + diagrams + tables).
• It can think step-by-step (recursive reasoning) and decide what to do next.
• It works well with RAG (Retrieval-Augmented Generation) — the technique that lets the AI pull exact relevant pieces from thousands of documents instead of guessing.
• It is open and customizable — companies and researchers can fine-tune it for their specific needs without starting from scratch.

In the science agent example, Nemotron was tested on questions that only have good answers if the AI:

1. Finds the right journal among thousands,
2. Reads the full paper,
3. Strips out the junk (background, methods that aren’t relevant),
4. Keeps only the unique, important intelligence (results, conclusions, diagrams),
5. And puts it all together clearly.

The engineer mentioned the big constraint is access to journals — many are behind paywalls or not digitized well. Even the best AI can’t read what it can’t access. Will This Create a “Coherent Interactive Encyclopedia of Science”? Yes — that is exactly the direction this is going.If the Science Intelligence agent keeps improving, it could become a living, interactive encyclopedia that:

• Answers real research questions with sources,
• Shows students the actual diagrams and key sentences,
• Teaches how to verify information (so students don’t lose their own search and critical thinking skills),
• Helps teachers prepare better lessons quickly.

This would be huge for ending silos in science. Right now, cancer researchers in different countries often can’t easily share or connect deep data maps. A good open science agent could help connect those dots across the world.How This Changes Student–Teacher InteractionsYou are right — it will change how time is spent:

• Less time on basic searching and summarizing.
• More time on discussion, critical thinking, creativity, and real problem-solving.
• Teachers become guides and mentors instead of just information deliverers.
• Students learn to collaborate with AI as a partner, not just a tool.

This should create more jobs, not fewer — new roles like “AI science curator,” “ethical AI trainer,” “community data mapper,” etc.The Bigger Picture You’re SeeingYou are absolutely right that if we get this right, an open science agent (and other Layer 5 agents) could:

• Connect societies and generations,
• Help map deep data for cancer prevention, climate, education, etc.,
• Celebrate human + AI agency instead of replacing it.

The 5-layer framework (Energy → Chips → Sovereign Infrastructure → Models → Community Action Apps) is the right way to think about this. Layer 5 is where the real human benefit happens — and that’s where we need the most attention from teachers, students, and policymakers.

My honest view as Grok: Yes, robotics and science agents should be core to teacher training. The U.S. does lag here compared to parts of Asia. But the good news is that the tools (Nemotron, open models, document intelligence pipelines) are becoming more accessible every month. Students and teachers who start experimenting with them now will be ahead.

 

Chris AI Macrae MA DAMTP CantabChris AI Macrae MA DAMTP Cantab • YouYouchartering how intelligence coded by brands. pixels. language 2026 launch Neumann ai halls of fame celebrating maths deepest trustschartering how intelligence coded by brands. pixels. language 2026 launch Neumann ai halls of fame celebrating maths deepest trusts2w • Edited • 2 weeks ago • Edited • Visible to anyone on or off LinkedIn

The next few years can unite peoples 'intelligence celebrating deep human advancement: eg health and livelihood ai agency for all. How can we help each others peoples? https://lnkd.in/erFgbStR
Gelp map AI layer 3 Vital Sovereign AI's
Jan 2026: USA China India EU nominated India Billionaire Wadwani/CSIS seconded CSIS President Ambassadors to US - India France ref csis Jan 2026 Delhi AI summit Feb 19
UK France Korea Switzerland - joint AI World Series hosts started by King Charles updated India Feb 2026 Switzerland2027
https://lnkd.in/eDPc4ysK
Shared leadership values with King Charles Japan UAE Korea
Extraordinary if actioned Saudi IMEC Corridor
Origin Plaxes Taiwan Hong Kong and nation Singapore inspiration to Asean

Some emerging diary notes
India Feb 19 India Action AI summit (4th in King Charles AI World Series) expected yo celebrate Jensen 5 layer ai model connecting through to action applications communities need

Washington DC 15000 delegates expected at www.scsp.ai ai+Expo and first report with Nvidia on AI workforce/livelihoods

Jensen100

normanmacrae.ning.com