Meet Real-Time Interaction Model

Thinking Machines launched human-like real-time AI, Google merged Android and ChromeOS into AI-first Googlebook, and Anthropic introduced Claude for legal workflows.

This week in AI, the industry is rapidly moving beyond basic chatbots toward AI systems that feel more human, deeply integrate into everyday computing, and specialize in real-world professional workflows. The focus is no longer just smarter models it’s about making AI continuously interactive, context-aware, and operational across industries.

• Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, unveiled a new “Interaction Models” system designed to make AI conversations feel fluid and human-like. Unlike traditional turn-based assistants, the model can listen, see, reason, and respond simultaneously across audio, video, and text using a real-time “micro-turn” architecture that reacts within milliseconds.

• Meanwhile, Google introduced “Googlebook,” an AI-first laptop platform that merges Android and ChromeOS into a single Gemini-powered ecosystem. Features like the AI-driven “Magic Pointer” aim to transform the operating system itself into an intelligent assistant capable of proactively understanding on-screen context, suggesting actions, and integrating AI directly into everyday computing workflows.

• Anthropic also expanded AI deeper into professional services with a specialized version of Claude built specifically for the legal industry. The platform helps legal teams analyze contracts, summarize cases, draft documents, and automate research workflows while maintaining enterprise-grade compliance and security standards.

Together, these developments show how AI is evolving from standalone assistants into deeply embedded, real-time systems powering conversations, operating systems, and industry-specific professional work.

---

Mira Murati’s Startup Launches Human-Like AI Interaction System

Thinking Machines Lab has unveiled a new AI system called “Interaction Models,” designed to make conversations with AI feel more like talking to a real person instead of waiting through turn-based prompts and replies. Unlike traditional chatbots that process one input at a time, the new model can simultaneously listen, see, think, and respond in real time across audio, video, and text. The company’s research preview, called TML-Interaction-Small, uses a “micro-turn” architecture that processes interactions in 200ms slices, allowing the AI to interrupt naturally, react instantly, and maintain live context during conversations. The system also combines a fast interaction layer with a background reasoning model for deeper tasks like browsing, tool usage, and long-form reasoning while the conversation continues seamlessly. Former OpenAI CTO Mira Murati says the goal is to make AI collaboration feel fluid and human-like rather than delayed and robotic.

Google unveils Googlebook, merging Android and ChromeOS

Google has officially introduced “Googlebook,” a new generation of AI-powered laptops designed around Gemini Intelligence and built by combining the best parts of Android and ChromeOS into a single platform. The company says this marks a shift from traditional operating systems to what it calls an “intelligence system,” where AI becomes deeply integrated into everyday computing. One of the biggest features is “Magic Pointer,” an AI-powered cursor that can understand on-screen context and proactively suggest actions like scheduling meetings from emails or combining images instantly. Googlebook will also support Android apps, Chrome browsing, custom Gemini-generated widgets, and seamless syncing with Android phones. Major brands including Acer, Asus, Dell, HP, and Lenovo are expected to launch Googlebook devices later this year, positioning them as premium AI-native alternatives to traditional laptops and Chromebooks. Early reactions online praise the ambitious Android-ChromeOS merger, while others question how much AI users actually want embedded into every interaction.

Anthropic launches Claude for the legal industry

Anthropic has introduced a specialized version of Claude designed specifically for legal professionals, aiming to streamline research, contract analysis, drafting, and case preparation with enterprise-grade AI. The company says law firms and legal teams can now use Claude to quickly analyze massive legal documents, summarize complex cases, compare contracts, and generate structured legal insights while maintaining strong security and compliance standards. Anthropic highlighted partnerships with major legal technology platforms and firms already integrating Claude into daily workflows to reduce repetitive legal tasks and improve productivity. The launch reflects a growing trend of AI becoming deeply embedded in professional industries, especially law, where firms are racing to adopt generative AI tools for faster research and document handling while balancing concerns around accuracy, confidentiality, and legal accountability.

Hand Picked Video

OpenAI released their most comprehensive study ever, analyzing over 1 million conversations from 700 million users worldwide. The findings reveal surprising shifts in how we're actually using AI.

Top AI Products from this week

• Memoket Gem - We’re opening 50 free spots for our Founding Member Program for founders, SMB owners who want to try Memoket Gem early. Memoket Gem is an all-day AI wearable that captures meetings, calls, coffee chats, and decisions on the go.

• Latitude - Latitude is an observability and quality platform for AI agents. It helps developers find and fix failure modes before they reach production. Most tools give you logs. Latitude gives you issues: failure modes with states and evals attached.

• CraftBot - A self-hosted Proactive AI Assistant that lives inside your machine and works 24/7 for your daily execution. It autonomously interprets tasks, plans actions, and executes them to achieve your goals. It learns your objectives, proactively helping you plan and initiate tasks to achieve your life goals.

• Googlebook - A new category of laptops built from the ground up for Gemini intelligence. These devices feature the Magic Pointer for contextual suggestions and custom widgets to help you organize your tasks. Keep an eye on googlebook.com for more updates before the devices launch this fall.

• Blaze- Blaze 2.0 is the marketing solution for people who don't have time to do marketing. It learns your business, your audience, and your voice — then creates and manages your entire content strategy, automatically.

• apideck -Apideck is a unified API platform that lets B2B SaaS and Fintech companies connect to accounting, HRIS, and CRM software without building and maintaining each integration separately. One API, dozens of connectors, normalized data models out of the box.

This week in AI

• Krea 2 launches next-gen AI image model - Krea 2 is Krea AI’s first fully in-house image generation model, built for stronger aesthetics, real-time creativity, and advanced style transfer controls.

• AllenAI releases MolmoAct2 FAST Tokenizer - MolmoAct2 FAST Tokenizer converts robot actions into compact AI tokens, helping robots process movements faster for real-world reasoning and automation tasks.

• Isomorphic Labs secures $2.1B for AI drug discovery - Google DeepMind spinout Isomorphic Labs raised $2.1 billion in Series B funding to scale its AI drug design engine and accelerate AI-powered medicine development.

• Rivian brings AI assistant to its EVs - “Hey Rivian” AI assistant adds natural voice control, smart navigation, messaging, and real-time vehicle support across R1 vehicles via its 2026.15 update.

• DeepMind gives AI “eyes” for your cursor - Google DeepMind unveiled AI Pointer, a system that helps computers understand what users are pointing at on screen, enabling smarter AI interactions and more intuitive agent-based computing.

Paper Of the day

A new AI paper explores how specialist AI agents can independently improve machine learning systems through continuous experimentation and feedback loops. Instead of relying on humans to guide every step, the system autonomously writes code, tests ideas, evaluates results, and refines training recipes over hundreds of trials showing how AI could eventually automate parts of scientific and AI research itself. The research highlights multi-agent collaboration, where separate AI systems handle planning, coding, evaluation, and optimization together like a virtual research team. Early results show these agents can discover better model configurations and improve performance with minimal human intervention. Researchers believe this approach could accelerate scientific discovery, software engineering, and future AI development workflows.

Read this whole paper 👉 here