Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

2. Master AI tools, tutorials, and news in just 3 minutes a day

3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

Today:

• Google Releases Free Firebase Studio

• Murati’s AI Startup Targets $2B

• OpenAI Teases New Model Features

• Canva Adds Code And Sheets

• OpenAI Cuts AI Testing Time

Google UNLEASHED Firebase Studio for AI app development (FREE)

Google launched Firebase Studio, a free AI-powered coding tool that lets anyone prototype and publish web apps with ease. It runs fully in the browser, mimics tools like Cursor, and offers both beginner and advanced modes. 

Users can chat with an AI to generate code, draw interface changes, and deploy apps fast. While still rough in places, its simplicity, power, and Google backing hint at major future impact.

WATCH THE VIDEO ON YOUTUBE

Murati’s AI Startup Targets $2B

Mira Murati, former CTO of OpenAI, is raising $2 billion for her new AI startup, Thinking Machines Lab—doubling her initial goal. The company, still under a year old and without a product, has attracted top AI talent and intense investor interest. Murati aims to make AI easier to understand and more adaptable. If successful, this could mark one of the largest early-stage fundraising rounds in tech history.

Why this matters

1. Leadership Migration: Top OpenAI leaders joining signals a major shift in AI talent.
   

2. Massive Investment: A $2B seed round shows rising confidence and urgency in AI innovation.
   

3. AI Accessibility Mission: The startup aims to make AI simpler and more usable, pushing the field toward broader adoption.

OpenAI Teases New Model Features

OpenAI is preparing to launch GPT-4.1, an updated version of its multimodal model GPT-4o, which processes audio, visuals, and text. The rollout may include smaller "mini" and "nano" versions, plus new models like o3 and o4 mini. Launches could happen next week, though delays are possible due to server capacity limits. CEO Sam Altman has hinted at new features but warned of potential slowdowns and instability during deployment.

Why this matters

1. Multimodal Advances: GPT-4.1 improves real-time reasoning across speech, images, and text.
   

2. Scalable AI: Mini and nano versions hint at more efficient, accessible AI for varied devices.
   

3. Model Pipeline Maturity: The pace of updates shows OpenAI’s growing ability to iterate and expand its model lineup.

Canva Adds Code And Sheets

Canva’s new Visual Suite 2.0 brings design, AI, and productivity tools into one platform, adding spreadsheets (Canva Sheets), coding tools (Canva Code), and a creative AI assistant. It lets users build documents, websites, and visuals collaboratively in real time. Features like “Magic Insights” and “Magic Charts” turn data into visuals, while AI handles tasks from photo edits to code generation—aiming to compete with Microsoft, Google, and Adobe in the workspace space.

Why this matters

1. AI Democratization: Canva brings advanced AI features to non-technical users through an easy interface.
   

2. Workspace Integration: Combines creativity and productivity tools into one unified, AI-powered system.
   

3. No-Code Coding: Canva Code expands access to software creation with AI-generated, editable code from simple prompts.

🧠RESEARCH

DDT: Decoupled Diffusion Transformer

DDT (Decoupled Diffusion Transformer) improves image generation by separating how a model understands overall meaning and fine details. This new setup solves conflicts in earlier models and speeds up training and inference. DDT sets new performance records on key benchmarks while running faster and scaling better with model size.

OLMoTrace: Tracing Language Model Outputs Back to Trillions of Training Tokens

OLMoTrace is a tool that tracks exactly where a language model’s words came from in its massive training data. It finds exact text matches quickly, helping users see how much a model copies, invents, or distorts. It's open-source and useful for checking facts, spotting errors, and studying model behavior.

Kimi-VL Technical Report

Kimi-VL is a lightweight, open-source AI model that understands both images and text with impressive skill. Despite using only 2.8B active parameters, it matches or beats larger models on complex tasks like reading diagrams, solving math, and following long instructions. It’s fast, smart, and freely available for developers to use.

VCR-Bench: A Comprehensive Evaluation Framework for Video Chain-of-Thought Reasoning

VCR-Bench is a new benchmark designed to test how well AI models reason through videos step by step. It includes over 850 videos and detailed question-answer sets with labeled reasoning paths. Results show that even top models struggle, especially with understanding time and space in video, exposing key weaknesses in current AI.

🛠️TOP TOOLS

Rewind - Browse, search, and ask Rewind about anything you’ve seen on your phone. Rewind is a truly personalized AI in your pocket.

Lovable - Superhuman full stack engineer.

Wondershare Virbo - AI-Powered One-Stop Video Marketing Solution to Maximize Your Conversions

HumanLayer - API and SDK that enables AI Agents to contact humans for feedback, input, and approvals.

RemoteBase - Quality-assured AI trainers, pre-vetted by experts to elevate your AI’s potential with speed and skill.

📲SOCIAL MEDIA

🗞️MORE NEWS

• OpenAI is speeding up the release of its new AI models by cutting safety testing time, raising concerns among insiders that public risks are being ignored in a rush to stay ahead of competitors.

• Writer launched “AI HQ,” a platform for building AI agents that automate complex business tasks. Unlike simple chatbots, these agents make decisions and learn from mistakes. Writer aims to make AI actually useful for enterprises.

• The EU plans to spend €20bn building massive AI supercomputer hubs to compete with the US and China. These “gigafactories” aim to power big scientific and industrial breakthroughs but raise environmental and regulatory concerns.

• A coalition has petitioned California’s Attorney General to investigate OpenAI, accusing it of abandoning its nonprofit mission. The group claims OpenAI now prioritizes corporate profits—especially Microsoft’s—over serving the public good.

• OpenAI has filed a countersuit against Elon Musk, accusing him of bad-faith attacks and a fake takeover bid aimed at disrupting the company. The suit claims Musk wants control over AI progress for personal gain.

• Ilya Sutskever’s new AI startup, Safe Superintelligence (SSI), is partnering with Google Cloud to power its research using advanced TPU chips. Backed by $1 billion, SSI is solely focused on building safe, superintelligent AI.

• An MIT study finds that AI models don’t hold real values or beliefs. Instead, they imitate human input inconsistently and unpredictably, challenging the idea that AI systems can be truly “aligned” with human intentions.