Learn to build and deploy GenAI pipelines in "Orchestrating Workflows for GenAI Applications", built in partnership with Astronomer and taught by Kenten Danas, the company's DevRel Senior Manager, and Tamara Fingerlin, developer advocate. Many GenAI applications require

🚀 New short course: Orchestrate GenAI workflows with Apache Airflow Notebooks are great for prototyping, but what happens when your GenAI app needs to scale, run reliably, and recover from failure? In our new short course built in collaboration with Astronomer, learn to

This week in The Batch, Andrew Ng discusses what skills he looks for when hiring GenAI application engineers. Plus: 🖼️ FLUX.1 Kontext image model debuts 📈 Mary Meeker’s 2025 AI trends 💸 Reasoning LLMs are pricy to test 🎥 STORM shrinks video tokens Read The Batch:

Curious about how reinforcement fine-tuning works in practice? The instructors of the "Reinforcement Fine-Tuning LLMs with GRPO" course are hosting a live AMA! Learn more 👇

Germany’s Black Forest Labs released three text-to-image models, FLUX.1 Kontext max, pro, and dev, that focus on maintaining the same character while changing backgrounds, poses, or styles. The models combine a convolutional image encoder-decoder with a transformer trained via

Columbia University researchers showed that LLM-based agents can be manipulated by placing malicious links on trusted websites like Reddit. By embedding harmful instructions within posts that appear thematically relevant, attackers can lure AI agents into visiting compromised

🥲 Original seen on Reddit: /ProgrammingMemes

Investment analyst Mary Meeker has released “Trends — Artificial Intelligence (May ’25),” her first tech market survey since 2019. 👉 The 340-page, data-rich report argues that AI’s breakneck adoption and escalating capital spending are fueling both record opportunities and

🧠 How did neural networks evolve to shape modern AI? From mimicking the brain in the 1950s to driving breakthroughs in deep learning today, neural networks have had a fascinating journey. But why did they fall out of favor—and how did they make a comeback? Learn more in the

Independent lab Artificial Analysis found that evaluating chain-of-thought reasoning models is rapidly becoming unaffordable for resource-constrained researchers. Benchmarking OpenAI o1 across seven popular reasoning tests consumed 44 million tokens and cost $2,767, while

New course with Meta! 🦙 Llama 4 introduced major upgrades for developers: a Mixture-of-Experts design that reduces serving costs, new multimodal capabilities like image grounding, and context windows that can handle entire books or codebases. To help you build with it, we've

Take your AI development skills to the next level with our latest course on Deeplearning.ai, "Building with Llama 4", taught by Andrew Ng and Amit Sangani, Director of Partner Engineering for Meta's AI team. In this comprehensive course, you'll learn how to harness the

This week, in The Batch, Andrew Ng discusses why making H1-B and F-1 visas more difficult to get harms U.S. competitiveness in AI and innovation worldwide. Plus: 🍎 Apple revamps models and developer API 🎬 Hollywood studios sue Midjourney over copyright use 🧠 OpenAI launches

Apple refreshed its Apple Foundation Models (AFM) with new versions for on-device and server use, aiming to improve performance in tasks like image understanding and multilingual reasoning. The company also released a Foundation Models API for developers to integrate the

Researchers introduced STORM, a text-video model that trims video input to one-eighth the usual size yet still yields state-of-the-art scores. STORM inserts mamba layers between a SigLIP vision encoder and a Qwen2-VL language model: the mamba layers aggregate information across

A reminder to check out our latest course with Meta: Building with Llama 4. In this course, you'll: 🦙 Get hands-on with Llama 4 family of models, understand its Mixture-of-Experts (MOE) architecture, and how to build applications with its official API. 🦙 Apply Llama 4’s

Disney and Universal filed a lawsuit against image generation company Midjourney, accusing it of training its models on their copyrighted content and reproducing it without permission. The studios claim Midjourney system generated unauthorized images of characters like