You’ve heard the hype. It’s time for results.

After two years of siloed experiments, proofs of concept that fail to scale, and disappointing ROI, most enterprises are stuck. AI isn't transforming their organizations — it’s adding complexity, friction, and frustration.

But Writer customers are seeing positive impact across their companies. Our end-to-end approach is delivering adoption and ROI at scale. Now, we’re applying that same platform and technology to build agentic AI that actually works for every enterprise.

This isn’t just another hype train that overpromises and underdelivers.
It’s the AI you’ve been waiting for — and it’s going to change the way enterprises operate. Be among the first to see end-to-end agentic AI in action. Join us for a live product release on April 10 at 2pm ET (11am PT).

Can't make it live? No worries — register anyway and we'll send you the recording!

Save your spot — register today

Today:

• AI Video Reaches New Heights

• Anthropic Launches Claude For Education

• OpenAI Releases PaperBench Benchmark Tool

• DeepMind Outlines AGI Safety Strategy

• Lindy Launches Agent Swarms Feature

AI Video is getting UNREAL... (GEN 4)

Runway has launched Gen-4, its most powerful AI video model yet. Unlike past versions, users say this one delivers: it creates consistent characters, objects, and environments across scenes—making real storytelling possible. 

Users are generating videos with strong visuals, smooth motion, and a surprising level of realism. While some clips still have glitches, early feedback is overwhelmingly positive, with Gen-4 praised as fast, fun, and creatively liberating.

WATCH THE VIDEO ON YOUTUBE

Anthropic Launches Claude For Education

Anthropic launched Claude for Education, an AI tool tailored for universities. It helps students think critically, supports faculty in personalized teaching, and aids administration with routine tasks. Features include Learning Mode, campus-wide access, and student programs. Partnerships with Northeastern, LSE, and Champlain College aim to make AI more accessible and responsible in education, while collaborations with Internet2 and Instructure ensure secure integration into existing academic systems.

Why this matters

1. Democratizes AI in education – Brings advanced AI tools to entire campuses, not just select users.
   

2. Promotes responsible AI use – Focuses on ethics, transparency, and critical thinking rather than automation alone.
   

3. Fosters future AI talent – Empowers students to build with AI and shape how it's used in society.

OpenAI Releases PaperBench Benchmark Tool

OpenAI released PaperBench, a benchmark testing whether AI can replicate top AI research papers from ICML 2024. The benchmark includes 8,316 detailed tasks based on 20 papers, with grading rubrics developed alongside the original authors. A custom AI judge evaluates performance. The best AI model, Claude 3.5 Sonnet, scored 21%, far below expert humans. PaperBench and its tools are open-sourced to advance research on AI’s real-world engineering capabilities.

Why this matters

1. Tests real-world AI skills – Goes beyond benchmarks to assess if AI can replicate complex, modern research workflows.
   

2. Reveals current limits – Shows even top AI models fall far short of human researchers, guiding expectations and development.
   

3. Boosts reproducibility – Encourages open science by evaluating and replicating AI research at scale with structured benchmarks.

DeepMind Outlines AGI Safety Strategy

Google DeepMind proposes a proactive approach to prevent severe harm from Artificial General Intelligence (AGI). The focus is on two key risks: misuse—by restricting access to dangerous capabilities—and misalignment—by designing models to follow human intent. Their strategy combines model training, oversight, and security systems. They emphasize early action before risks fully materialize, aiming to shape future safety standards while acknowledging the need for evolving research and broader governance.

Why this matters

1. Prioritizes human safety – Addresses AGI risks that could seriously impact humanity.
   

2. Focuses on prevention – Stresses securing AI before dangerous capabilities emerge.
   

3. Encourages standards – Pushes for shared safety practices to avoid a race to the bottom.

🧠RESEARCH

Any2Caption:Interpreting Any Condition to Caption for Controllable Video Generation

Any2Caption is a new system that turns almost any input—like text, images, or motion—into clear captions to guide video creation. It helps video generators better understand what users want. The team also built a huge dataset to train the model. Tests show it improves control and video quality.

JudgeLRM: Large Reasoning Models as a Judge

JudgeLRM is a new model designed to evaluate complex tasks better than current systems. Unlike standard methods, it uses rewards to learn how to judge answers more wisely. It beats top models like GPT-4 in reasoning-heavy tests, showing that smarter evaluation needs more than just fine-tuning.

CodeARC: Benchmarking Reasoning Capabilities of LLM Agents for Inductive Program Synthesis

CodeARC is a new benchmark that tests how well AI can write code by learning from examples. Unlike past tests, it lets models get feedback and improve their answers, simulating real-world coding tasks. The best model solved only half the problems, showing how tough inductive programming still is for AI.

🛠️TOP TOOLS

Deepfake Video Maker - Cloud-based online software designed to facilitate the creation of deepfake videos using artificial intelligence.

Image To Font Finder - AI-powered tool designed to help users identify fonts from any image.

Bai Chat - AI platform designed to simplify the integration of artificial intelligence into various workflows for professionals, developers, and businesses.

DiagramGPT - AI-powered tool developed by Fraser Xu that enables users to generate a variety of diagram types using natural language input.

JanitorAI - AI tool that integrate chatbot functionality into applications, leveraging technologies such as NLP, ML, and generative AI.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Adobe’s new Premiere Pro update adds AI tools to extend video clips, search footage using text, and translate captions. The highlight feature, Generative Extend, lets users lengthen scenes without reshooting for smoother edits.

• Lindy AI’s new 'Agent Swarms' feature lets users deploy hundreds of AI agents at once to handle many tasks in parallel. With over 5,000 integrations, it aims to be the most connected AI platform.

• China’s Genspark just unveiled Super Agent, an AI that handles entire tasks—from making recipe videos to deep research—without human help. It even beat top rivals Manus and OpenAI on the GAIA benchmark.

• Josh Woodward, creator of Google’s AI note app NotebookLM, is now leading the Gemini app as Sissie Hsaio steps down. Google says this change will help focus on the app’s next big evolution.

• MLCommons released two new tests to measure how fast AI systems run complex tasks. Nvidia's latest chips showed big speed gains, while AMD didn’t participate. One test uses Meta’s massive Llama 3.1 model.

• DeepMind’s Dreamer AI has learned to find diamonds in Minecraft entirely on its own. Unlike past systems, it wasn’t shown how to play—it imagined future outcomes using a built-in world model to guide decisions.