🤖DeepMind’s AI Is Making Robots Smarter

Today we will discuss:

• Sponsored: Carousel Studios—The fastest way to create LinkedIn carousels using AI

• DeepMind’s AI upgrade for robots—DeepMind’s latest models are enhancing robot reasoning and dexterity

• Nvidia’s GTC conference—Expect AI breakthroughs, next-gen GPUs, and quantum computing insights

• In other AI news today—Apple is reportedly developing AirPods with built-in cameras for AI features, Snap debuts AI Video Lenses powered by its own generative model, OpenAI urges the Trump administration to prioritize speed and light regulation in its AI strategy, and Alibaba releases an AI model that reads emotions from video

• Don’t miss this week’s Workflow Wednesday—AI and creativity unlocked with Sudowrite.com, Suno.com, and Descript.com; now available at only $20/month! 

• Saved the best for last—9 must-try AI tools

Create LinkedIn Carousels in <5 mins

Building a personal brand on Linkedin? You need this. (9500 creators already use it!)

Carousel Studio lets you create LinkedIn carousels directly in Canva in <5 mins with AI. 

Here’s the kicker: It's 100% FREE for you!

Creating carousels manually sucks because you have to:

- Think of a good copy

- Mess around with the layout

- Pick the right colors, fonts, theme

Carousel Studio takes all the hard work out:

1. Give it a topic, pick some pre-defined colors

2. Click “Generate” and it will create a 4-slide carousel perfect for Linkedin for you

3. Just edit & post!

Get it here

Google DeepMind unveils AI models to make robots smarter and more capable

Key Points 

• Google DeepMind introduced Gemini Robotics and Gemini Robotics-ER to enhance robot adaptability, interactivity, and dexterity in real-world tasks.

• DeepMind is prioritizing safety with AI-driven risk assessment and new benchmarks, building on its Asimov-inspired 'Robot Constitution.'

♨️News - Google DeepMind is rolling out two new AI models designed to help robots take on a broader range of real-world tasks. The first, called Gemini Robotics, is a vision-language-action model that can interpret new situations, even if it hasn’t been specifically trained on them.  

🤖Here's the lowdown - Built on Gemini 2.0, Google’s latest AI model, Gemini Robotics combines multimodal understanding with physical actions. It improves three key areas essential for robotics: generality (adapting to unfamiliar scenarios), interactivity (engaging with people and environments), and dexterity (handling precise tasks like folding paper or opening a bottle cap).  

🤹‍♂️What's more? DeepMind is also launching Gemini Robotics-ER (Embodied Reasoning), an advanced visual language model designed to help robots better understand and navigate complex real-world situations.  Senior director Carolina Parada gave an interesting example: when packing a lunchbox, you need to recognize where everything is, open the box, pick up items, and place them correctly—Gemini Robotics-ER is built to tackle that kind of reasoning. It can also integrate with existing robotic systems, adding new AI-powered capabilities.  

🤓What about safety? DeepMind researcher Vikas Sindhwani says the company is taking a “layered approach” to safety. Gemini Robotics-ER models are trained to assess whether an action is safe before executing it. DeepMind is also introducing new safety benchmarks and frameworks, expanding on its previously announced “Robot Constitution,” a set of Asimov-inspired rules for AI-driven robots.  

🚀What's next? DeepMind is working with robotics company Apptronik to develop the next generation of humanoid robots. It’s also giving early access to Gemini Robotics-ER to companies like Agile Robots, Agility Robotics, Boston Dynamics, and Enchanted Tools. As Parada puts it, DeepMind’s focus is on building intelligence that can not only understand the physical world but also act within it.

Read the full story here. 

Big AI and GPU updates expected at Nvidia’s GTC conference

Key Points 

• CEO Jensen Huang’s keynote on Tuesday will focus on AI, accelerated computing, and potential updates to Nvidia’s GPU lineup.

• Blackwell Ultra and Rubin GPUs could take center stage, with Nvidia also hosting a dedicated session on quantum computing.

☕News – GTC, Nvidia’s biggest conference of the year, kicks off Monday and runs through Friday.

✨Here's what to expect – CEO Jensen Huang will deliver his keynote on Tuesday at 10 a.m. Pacific, focusing on AI and accelerated computing. Nvidia is also teasing major announcements in robotics, sovereign AI, AI agents, and automotive, alongside 1,000 sessions featuring 2,000 speakers and nearly 400 exhibitors.

Hardware news will likely take center stage. Nvidia has a history of using GTC to unveil its latest hardware, and this year, all eyes are on an upgraded iteration of its Blackwell chips. The Blackwell B300 series, known as Blackwell Ultra, is expected in the second half of the year, featuring more memory (288GB) for demanding AI workloads.

Looking further ahead, Nvidia’s next-gen GPU series, Rubin, is expected to get a mention. Set for release in 2026, Rubin is being positioned as a major leap in performance, with Huang describing it as a “big, big, huge step up.”

Beyond GPUs, Nvidia has scheduled a dedicated “quantum day” at GTC, bringing in industry leaders to discuss how the technology is progressing toward real-world applications.

Read the full story here. 

We’ve just launched the Tenth edition of Workflow Wednesday for AI-minded professionals like you—actionable AI workflows delivered directly to your inbox.

This week’s topic: AI-Powered Creativity

Join us here