Gemini AI updates:

—Gemini 2.5 Pro Deep Think, a new mode that uses parallel thinking to solve math and coding problems
—Gemini 2.5 Flash, upgraded with improved performance across benchmarks

Both also include native audio outputs across languages

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Jules: An asynchronous coding agent that can plan steps, modify files, and do other tasks (like fixing bugs) in codebases, in mere minutes

Now in public beta

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Stitch: A Google Labs product to generate designs and UIs from text prompts and download them into Figma

Coming from Google's acquisition of Galileo AI

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Try it on: An AI-powered Google Search Labs feature that will let users try on clothes virtually while shopping on Google

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

AI Mode: Now rolling out to all Google Search users in the U.S., providing direct answers to questions like ChatGPT and Perplexity

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Two early AI-powered Android XR (extended reality) devices demoed on stage I/O 2025:

—Samsung-built Project Moohan headset taking on Apple Vision Pro
—Unnamed AI smart glasses taking on Meta Ray Ban Glasses

Project Moohan will go on sale later this year

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Azure AI Foundry updated with new models including Grok 3, Flux Pro 1.1, and Sora (coming soon)

Also includes new model router, multi-agent workflows, observability features, and Foundry local for creating localized AI apps on Windows and Mac

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Windows enhanced with new AI-focused capabilities, including:

—Windows AI Foundry with Windows ML, Foundry Local, ready-to-use AI APIs for vision and language tasks
—Native MCP support
—App Actions
—Open-sourced Windows Subsystem for Linux

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Microsoft Discovery: A new agentic platform that enables researchers to collaborate with specialized AI agents to drive scientific research and outcomes

The agents generate ideas, simulate results, and learn over time

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

OpenAI launched Codex agent for software dev

—Built on codex-1, an o3 variant fine-tuned for software engineering
—Runs in isolated clouds to write features, fix bugs, answer code questions, and run tests
—Available to ChatGPT Pro, Enterprise, Team users

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

At COMPUTEX 2025, Nvidia CEO Jensen Huang announced:

—Isaac GR00T N1.5, an upgraded version of its open AI for humanoid reasoning
—GR00T-Dreams blueprint to generate synthetic motion data from single images for training (used to train N1.5)

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

YouTube and Netflix are both working to launch AI ads

YT is experimenting with Peak Points, where it will show ads right after emotionally charged moments

Netflix is developing AI ads that will place products over backgrounds inspired by programs

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Salesforce acquired Convergence, the UK-based startup building Operator's rival Proxy

The deal will see Salesforce tap Convergence's team and agentic technology to further enhance its Agentforce platform

Financial details of the deal remain undisclosed

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Google quietly launched Edge AI Gallery, a library of open AI models to access on smartphones

Currently on Android, the gallery lets users pick models for their task and run using the available device hardware, entirely locally

iOS version coming soon!

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

ElevenLabs launched Conversational AI 2.0, its next-gen platform for creating voice agents with:

—Real-time turn-taking model
—Built-in language detection
—Integrated RAG
—Multimodality
—Scalable batch calling

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Resemble AI dropped Chatterbox, a SOTA text-to-speech AI that beats ElevenLabs and offers:

—Zero-shot voice cloning from 5s audio
—Style + emotion intensity control
—Built-in voice conversion & watermarking
—Real-time voice synthesis
—Open weights

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

DeepSeek updated its R1 model, instantly propelling it to the #3 spot on the Artificial Analysis leaderboard with:

—Enhanced front-end capabilities
—Reduced hallucinations
—JSON output & function calling support

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Hume launched EVI 3, a new AI that can speak in any voice or personality via speech-to-speech interaction

It uses a voice-to-voice architecture and comes with a deeper understanding of tune, rhythm, timbre, and speaking style

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Japan's Sakana AI launched a self-improving agent that can modify its code to boost performance

On SWE-bench, the Darwin Gödel Machine improved its performance from 20.0% to 50.0%, while on Polyglot, it increased its success rate from 14.2% to 30.7%

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Tencent launched HunyuanVideo-Avatar, a new AI video model

It turns a single reference image and audio into a realistic short video with lifelike speech/singing

Suitable for domains like ecommerce and advertising

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots

Anthropic is adding a voice mode to Claude mobile, providing English-speaking users with:

—Five voice personalities to interact with Claude Sonnet 4
—Real-time transcription of chats
—20-30 voice messages for free users
—Higher usage for paid users

❤️ @chatgpt5 | 🤖ChatGptBot | 🤖 Our Bots