Each week, we bring you 5 stories that resonated the most in our internal Slack channel #AI-news. We write the newsletter using various AI tools because we're an AI company and our marketing wants to move with the times too. 😎
Today you're reading the 66th issue in a row.
#1
Windsurf 💼: AI coding tool may end up under OpenAI 🤖💻.
OpenAI is reportedly in talks to acquire Windsurf 💼—an AI-powered tool 🤖 that helps developers write code 💻. The deal is said to be worth around $3 billion 💰. Why? OpenAI has recently been falling behind 🐢 in the programming space compared to rivals like Google and Anthropic, whose latest models are more powerful when it comes to coding ⚙️🚀. Instead of building a tool from scratch, OpenAI aims to fill the gap through acquisition 🔄.
Windsurf fits into this strategy well 🧩—it boosted its revenue from $40 million to $100 million this year 📈 and is becoming a favorite among enterprise developers 👨💻🏢. OpenAI previously tried something similar with an attempt to acquire Cursor 💡, but that tool is currently valued at $10 billion 💸 and the deal fell through ❌.
If Windsurf accepts the offer ✅, it could help OpenAI gain traction among developers and catch up with the competition in the “AI for coding” arena ⚔️. Coding tools powered by AI are quickly becoming a key battleground 💣—and OpenAI doesn’t want to be left out 👀.
#2
Colossus 2 🚀: Musk’s supercomputer with one million NVIDIA GPUs 🎮🏗️.
Elon Musk’s startup xAI 🚀 is planning to raise up to $25 billion 💰 to build a supercomputer called Colossus 2 🧠. It’s designed to run on one million NVIDIA graphics cards 🎮—nearly ten times the compute power currently driving the AI system Grok 🤖. If successful, it would be the largest AI supercomputer housed in a single building 🏢 and likely one of the most powerful in the world 🌍.
Unlike companies like OpenAI or Anthropic 🏢 that rent computing power, xAI is taking a different route 🛠️—building its own hardware. It’s an ambitious and extremely costly move 💸, focused on achieving maximum control 🎯. But the project faces major challenges: from GPU shortages ⚠️ to high demands for electricity ⚡ and cooling ❄️, as well as the complexity of getting one million GPUs to work efficiently together 🔄.
Despite these hurdles 🔧, investors see Colossus 2 as a long-term bet on Musk’s ability to scale fast and at massive scale 📈. The supercomputer is currently being built in Memphis 🏗️—with the city itself involved in the project 🏙️. This is far from just another chatbot 💬; it’s a demonstration of technological power on an industrial scale 🏭.
#3
Perplexity 💡: AI assistant to be pre-installed on Motorola phones 🤖📱!
Perplexity, the AI-powered question-answering tool 🤖, has partnered with Motorola 📱. Their app will now come pre-installed on millions of devices 🌍—including the iconic Razr model 🔄. This means users will have instant access to AI right after turning on their phone, capable of answering questions, booking a restaurant 🍽️, sending an email ✉️, or playing music 🎵—even when the phone is folded 📲.
What makes this partnership unique is that it’s not just about bundling an app—it’s about deep integration of AI into both the software and hardware of the phone ⚙️🔗. Perplexity will be more than an add-on; it will be part of the overall user experience 👤. As a bonus 🎁, new Motorola users will get three months of free access to Perplexity Pro, which includes advanced models like GPT-4o and Claude 3.5 💡.
In short: an always-on AI assistant built right into your phone 📞🤖 is becoming reality ✅. This move marks Perplexity’s bold push forward 🚀—with ambitions to outpace tech giants like Google and Apple ⚔️.
#4
DeepMind 🤖: Created a realistic digital fruit fly using AI 🪰🤖!
DeepMind has used its physics simulator MuJoCo 🧪 to create a digital fruit fly 🪰 that behaves incredibly realistically. This virtual fly, powered by artificial intelligence 🤖, flaps its wings like a real insect and uses virtual eyes 👁️ to navigate and respond to its environment. All of this happens in a 3D environment 🌐, and at first glance, it looks eerily lifelike 🧠.
The training was based on real video recordings of fruit flies 📹. The neural network learned on its own how to move, where to go, and how to react to stimuli 🧬. The result is a digital creature that closely resembles its living counterpart—not just in appearance, but also in behavior ⚙️🪰.
Why try to replicate an insect? Because such simulations help scientists study how the brain 🧠, body, and environment interact—without needing to dissect hundreds of living creatures 🔍. After previously modeling a rodent 🐁, DeepMind has now focused on the fly, with a zebra finch 🐦—a small bird that shares surprisingly much of our genetic code 🧬—next in line. AI in neuroscience? Maybe a bit unusual, but incredibly useful 🤝.
#5
AI in Schools 🇺🇸: Trump signs executive order on AI education 🖊️🎓!
President Trump has signed an executive order 🖊️ aiming to introduce artificial intelligence into classrooms starting in elementary schools across the U.S. 🇺🇸. The goal is to teach students not just how to use AI 🤖, but also to understand it 🧠 and critically evaluate its impact ⚖️. A federal task force 🏛️ will be created to develop a nationwide strategy for AI education. Apprenticeship programs 🧰 and special high school certifications 🎓 are also being considered.
This move isn’t entirely new—it builds on what countries like China 🇨🇳, Singapore 🇸🇬, and Estonia 🇪🇪 are already doing. The U.S. doesn’t want to fall behind ⏳. But bringing AI into the classroom is no simple task: schools will need well-trained teachers 👩🏫, high-quality learning materials 📚, and sufficient funding 💰—especially in regions already struggling with tight budgets ⚠️.
In a broader sense, this is about more than just technology 🧑💻. It’s about shaping the next generation of workers 🧑🔧—whether they’ll be able to build and shape AI, or simply consume it passively 👀. The success of this initiative could determine what role today’s children will play in the tech-driven world of tomorrow 🌍🚀.
