AI Pulse 🚨
A bite-sized curation of this week's most important AI news.
🗣️ Sesame launches AI voice model delivering the most human-like AI voice to date, with remarkably natural speech patterns and emotional expression that adjusts based on conversational flow. Try it out yourself or check out this demo!
💸 OpenAI plans premium AI agents - The company is reportedly developing specialized AI agents priced up to $20,000 per month, including a PhD-level research assistant and software developer tools.
🎉 ChatGPT 4.5 rolls out to Plus users - OpenAI releases GPT-4.5 to Plus subscribers with improved emotional intelligence and natural conversation.
🤖 ChatGPT 4.5 shows “emotional intelligence” — Users are reporting that the new model from OpenAI is much better at understanding context and emotional nuance, delivering more human-like responses.
🌟 Anthropic launches new platform for team collaboration - The company released an updated Console with shareable prompts and extended thinking controls, allowing cross-functional teams to work together on AI projects.
🧠 Gemini launches endless memory for all users - Google has rolled out its memory feature to free users, allowing Gemini to remember personal information and maintain context across conversations.
🧠 Amazon developing hybrid reasoning AI model - Amazon’s upcoming Nova-branded model will combine quick responses with extended thinking capabilities, aiming to offer more price-efficient reasoning than competitors. Launch expected in June.
🌊 Windsurf releases Wave 4 - Major update brings previews, tab-to-import, linter integration, and suggested actions. Includes improved MCP discoverability and drag & drop file support. Claude 3.7 Sonnet integration tweaked for better performance.
🔈 xAI’s Grok 3 “uncensored” voice mode - Now available to all users with singing, laughing, and bold personalities like “Unhinged 18+” and “Ara,” though Android users still await access.
👁️ Cohere releases multilingual vision model - Aya Vision brings capabilities to 23 languages, with versions outperforming rivals many times their size. It interprets images and translates visual content across diverse languages.
🏢 Salesforce launches first-ever marketplace for AI agents - AgentExchange lets companies find pre-built agents or create their own with templates.
🔊 Stability AI releases Stable Audio Open - This open-source text-to-audio model generates up to 47 seconds of audio and runs 30x faster on smartphones through Arm partnership.
💰 Anthropic raises $3.5B at $61.5B valuation - Anthropic secured Series E funding, tripling its valuation to accelerate AI development, expand compute capacity, and drive international growth for its Claude platform.
🔍 Google introduces Data Science Agent in Colab - This tool automates notebook creation from natural language prompts, generating code for preprocessing, visualization, and ML tasks without requiring boilerplate code.
🔋 TSMC signs major US chip deal - The Taiwanese company will invest $100B in the US over four years, including Arizona AI chip plants.
🎙️ Opera becomes first major browser to get an AI agent - Browser Operator can control a user’s computer to find flights, order tickets, and perform other tasks.
🧪 OpenAI launches $50M NextGenAI consortium - This academic initiative supports AI research at 15 institutions including Harvard and MIT with grants, compute resources, and API access.
📱 Perplexity and Deutsche Telekom announced AI Phone - A sub-$1000 device featuring Magenta AI assistant with Perplexity, Google Cloud, ElevenLabs, and Picsart integration. It aims to be “always listening” for proactive assistance and launches in Europe in 2026.
