• Everyday AI
  • Posts
  • NVIDIA changes the future of computing, Google’s BIG AI shakeup – AI News That Matters

NVIDIA changes the future of computing, Google’s BIG AI shakeup – AI News That Matters

New U.S. AI chip export rules and NVIDIA’s criticism, U.K. unveils new AI plans, Google automotive AI agent, and more!

Author

Everyday AI
January 13, 2025

Outsmart The Future

Today in Everyday AI
8 minute read

🎙 Daily Podcast Episode: From NVIDIA’s busy week in AI to Google’s AI shakeups, we break down the AI news that matters. Give it a listen.

🕵️‍♂️ Fresh Finds: GitHub Copilot Workspace now generally available, Gemini Flash 2.0 native image output and Perplexity launches Perplexity Sports. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: New U.S. AI chip export rules and NVIDIA’s criticism, U.K. unveils new AI plans and Google automotive AI agent. For that and more, read on for Byte Sized News.

🚀 AI In 5: Is Google an AI winner or loser in 2024? We break down their recent AI moves and where they stand among competitors. See it here

🧠 AI News That Matters: NVIDIA, Google, OpenAI and more had a big week in the AI world. Wondering what you missed? We break it all down. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about CES 2025 AI announcements, NVIDIA unveiling new AI avatars, Meta reportedly willingly pirating books for AI training and Microsoft’s performance-based AI layoffs. Check it here!

AI News That Matters - January 13th, 2025 📰

Is Microsoft laying off thousands because of AI?

How did one small box from NVIDIA change the future of work?

What are Google’s big AI shakeups?

And why is OpenAI getting into humanoid robots?

So many AI questions. We’ve got the AI answers with the AI news that matters.

Join the conversation and ask Jordan any questions on AI here.

Also on the pod today:

• Microsoft's Open Source Model 🤖
• Google's AGI shift 🧠
• Zuckerberg on AI and Software Engineers 🧑‍💻️

It’ll be worth your 43 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – Revfx discovers look-alike accounts instantly, Eapy is a multimedia workspace for music creation and UpMeals is an AI operating system for food businesses.

Microsoft – Microsoft has opened its GitHub Copilot Workspace to the general public.

Amazon – AWS’ CEO sat down to talk AI, cloud and nuclear power

Google – Here’s a preview on how Gemini Live might work with Youtube videos.

Google has teased native image output with Gemini 2.0 Flash.

Trending in AI - Hugging Face has settled a lawsuit with AI startup FriendliAI.

Perplexity – Perplexity has launched Perplexity Sports.

AI Models – Researchers have open-sourced Sky-T1, a reasoning AI model.

1. Biden Administration’s New Rules on Exporting AI Chips 🇺🇸️

The Biden administration's fresh proposal on AI chip exports is stirring the pot, aiming to balance national security with economic interests. While the plan seeks to preserve U.S. leadership in AI by restricting chip access to 120 countries, industry leaders warn of potential disruptions to global supply chains and economic setbacks.

The framework has a 120-day comment window, leaving room for President-elect Donald Trump to shape the final rules amid concerns that it may unintentionally hinder innovation.

2. NVIDIA Challenges New U.S. AI Chip Rules 👀

NVIDIA is pushing back against new AI regulations announced by the Biden administration, which aim to limit AI chip exports to certain countries. These rules are designed to keep powerful AI technology under the control of the US and its allies, but NVIDIA argues they could stifle innovation and economic growth.

The company is eyeing the incoming Trump administration for more favorable policies that could enhance American competitiveness in AI. With key players like Microsoft ready to comply with these standards

3. U.K. Takes Bold Step into AI Future 🇬🇧

the U.K. government announced a sweeping "Plan for Change" to embrace AI, with plans for AI Growth Zones and a major £14 billion investment pledge from private firms. Culham, Oxfordshire, is set to become the first of these Zones, aiming to boost public sector efficiency and enhance the U.K.'s technological landscape with a supercomputer and National Data Library.

While investors welcome the move, they caution that success will require cohesive policy alignment and competitive taxation to avoid talent loss.

4. Microsoft Takes Legal Action Against AI Hackers 🧑‍⚖️️

Microsoft has filed a complaint against an unnamed group accused of bypassing the safety measures of its Azure OpenAI Service, according to a report by TechCrunch. The tech giant alleges that the defendants used stolen API keys to develop a "hacking-as-a-service" tool called de3u, which facilitated unauthorized content generation and avoided Microsoft's content filters.

As Microsoft implements additional countermeasures, this case emphasizes the critical need for robust security protocols to safeguard AI.

5. Google's New Automotive AI Agent Hits the Road 🚙

Google Cloud has introduced the Automotive AI Agent, an innovative generative AI tool designed for automakers, according to Google Cloud. This new agent, powered by Gemini with Vertex AI, is set to redefine in-car experiences by providing highly personalized and natural dialogue capabilities, surpassing traditional vehicle voice controls.

Mercedes-Benz is among the first to integrate this technology into their MBUX Virtual Assistant, set to launch with the new Mercedes-Benz CLA later this year.

NotebookLM 's Join Interactive Mode. The single BEST AI feature ever?

This one feature inside NotebookLM is getting slept on!

We think this is one of the most IMPORTANT features in AI history.

Yep. You read that right.

Why? We dive in and show you.

Check out today's AI in 5.

Is Microsoft laying off thousands because of AI? 

How did one small box from NVIDIA change the future of work? 

What are Google’s big AI shakeups? 

And why is OpenAI getting into humanoid robots? 

So many AI questions. 

Want to be the smartest person in AI at your company? 

It starts with wrangling all of the AI news and being able to translate how it impacts you. 

So, let’s get started shorties.

1 – NVIDIA’s RTX Series Rewrites the Rules 🎮

Last year's $1,600 GPU performance now costs $549. 

No, that's not a typo.

The RTX 50 series is using generative frames to double down on performance and not cost. AI's filling in the gaps instead of brute-forcing every pixel.

And the best part? These beasts are coming to laptops in March. Portable AI power just went mainstream.

What it means: 

GPUs run everything in the GenAI space. 

Big tech needs hundreds of thousands of GPUs to make their AI chatbots run for billions of users, and we all need GPUs to run LLMs locally. 

High-end AI hardware just went from "second mortgage territory" to "skip Starbucks for a few months.” 

2 – NVIDIA’s Project Digit: Desktop Supercomputer Goes Hard 🚀️

A petaflop of AI power. In a box on your desk. Three grand.

Wait, what? 

Surprise of 2025 so far might be NVIDIA’s recently announced Project Digit, which is essentially an insane supercomputer but for personal use. (If you’ve $3K to cough over.) 

Project Digit packs their Grace Blackwell super chip and 128GB of memory into a compact powerhouse that could theoretically run GPT-4o locally. 

(You can’t actually run a proprietary model locally, but this in theory possible) 

The secret sauce no one's talking about? It works with ANY Windows or Mac setup as well as running as a standalone device. No Linux required. Just plug in and watch your computer transform into an AI beast.

What it means:

 The "only big tech can do AI" era is officially DEAD. 

Small companies can run serious AI without bleeding cash on cloud credits. Local AI infrastructure's about to be as common as having a company server.

3 – NVIDIA Cosmos: Your Robot’s New Gym 🤖

NVIDIA trained Cosmos on TWENTY MILLION HOURS of real-world footage to revolutionize robotics and autonomous systems. Not simulated. Not synthetic. Real world data.

Companies like Agility, Figure AI, and Uber are already using it. Plus, the Isaac robot simulation platform now generates synthetic training data from limited examples.

What it means: 

Robot training just got a major upgrade. NVIDIA's not just building a gym - they're creating the whole CrossFit ecosystem for AI robotics. But like times 1,000. 

4 – NVIDIA Nemotron Family: Llama’s Glow-Up 🦙

DLast NVIDIA piece of news, we promise. 

(But yo… their announcements were massive. Go watch Jensen’s CES keynote if you haven’t already.) 

NVIDIA took Meta's Llama models and said "we can do better." Their new Nemotron family comes in small, medium, and large flavors, all optimized for enterprise.

The kicker? Their first version outscored the respective Llama variation on benchmarks. NVIDIA's engineers just out-Meta'd Meta.

What it means: 

This isn't just another model family - it's NVIDIA's play for enterprise AI dominance. Better models drive hardware demand. 4D chess moves only.

5 – Google DeepMind’s Real World AI Push 🌍

According to reports, DeepMind is creating a new team of AI researchers focused on developing "world models" to simulate physical environment

A few months ago, Google DeepMind snagged OpenAI's Sora co-lead Tim Brooks and is going ALL IN on world models. 

They're teaching AI how reality actually works.

The restructure's massive too - all Gemini teams now report to DeepMind. Big boss energy.

What it means: 

Language models know everything on the internet but can't explain how a bicycle stays upright. 

DeepMind's betting that understanding physical reality is the key to AGI.

6 – Microsoft’s Phi-4: The Giant Killer  🎯

Phi-4 just dropped on Hugging Face with full MIT license. 

Fourteen billion parameters of pure performance, smoking Gemini Pro on certain benchmarks.

What it means: 

Efficient, well-trained models are eating the big models' lunch. 

Microsoft's proving you don't need a trillion parameters to hang with the big dogs.

7 – Microsoft’s AI Talent Defense 💼

Microsoft's reportedly cutting 1% of their workforce but throwing cash and stock at AI talent like it's going out of style.

They literally added "Would losing this person hurt our AI game?" to performance reviews. (Well, not EXACTLY that, but that’s the gist.) 

What it means: 

The great AI workforce reshuffle isn't coming - it's HERE. If you're not working on AI at your company, you better start learning. Yesterday.

8 – Meta’s Mid-Level Engineering Automation 💻

Not entry-level jobs - we're talking positions paying HALF A MILLION per year.

The future of coding? Natural language prompts over traditional programming. Jensen Huang called it first.

What it means: 

Forget everything you heard about AI only taking entry-level jobs. Meta's targeting the expensive positions because that's where the ROI is. 

9 – OpenAI’s Robot Revolution 🤖

OpenAI's back in the robotics game with custom sensor suites and in-house AI models. They're hiring people to test prototypes with actual limbs.

With robotics pulling in $6.4B of VC money last year, this isn't just another side project.

OpenAI shut down its robotics team in 2021 due to challenges like limited training data but quietly relaunched it in 2024, aiming to integrate AI into versatile robots. Now, the division is getting ‘Internet Job Listing Official.’ 

What it means: 

Multimodal AI + advanced robotics = game over. OpenAI isn't just building smart chatbots anymore - they're building smart EVERYTHING.

Embodied AI is gonna be a booming sector in 2025. 

10 – OpenAI’s Strategic Power Play 📊

OpenAI dropped a 15-page economic blueprint warning that $175B in global funds could flow to China if the U.S. doesn't step up its AI game.

They're pushing for nationwide rules and export controls on AI models. The stakes? Only the entire future of global AI development.

What it means: 

The AI race just went from tech competition to geopolitical chess match. OpenAI's thinking way beyond product launches - they're playing for control of the entire AI ecosystem.

⌚

Numbers to watch

$105 Million

Qventus, an AI startup that builds tools to automate healthcare work, has raised $105M.

Now This …

What are your thoughts?

Vote to see live results

How much trust do you have in our reporting?

Login or Subscribe to participate in polls.

Reply

or to participate.