Greetings, AI Enthusiasts!

Welcome!

We're back to chronicle AI's relentless march into the future. 🤖

In this edition, an AI chipmaker pushes the boundaries on language model speeds, while Google elevates Workspace with advanced Gemini integration.

We've also packed in the latest on OpenAI's mega funding round, moves by Big Tech to counter election deepfakes, and more. Plus - a toolkit of leading AI resources, trending news bytes, and a glimpse of creative AI in action.

Let's crack this open and explore! The frontiers of AI await... ✨

^{Read time: 6 minutes}

📰 AI and Tech Today

^{News and insights}
Startup Groq Speeds Up Language Models

AI chipmaker Groq has developed specialized processors called Language Processing Units (LPUs) to accelerate large language models. In tests, LPUs achieved speeds up to 500 tokens per second on models like GPT-3.5 — much faster than today's GPUs.

The energy-efficient LPUs overcome bottlenecks in traditional systems, allowing effortless scaling. Their design works with common ML frameworks for easy integration.

For now, LPUs focus on inference rather than training. But their speed and scalability could provide an alternative to scarce, expensive Nvidia chips for running the latest natural language systems.

Groq was founded in 2016 by industry pioneer Jonathan Ross to commercialize his work on Google's TPU AI chips. With LPUs now rolling out, Ross aims to repeat that success in powering more capable language technologies.

Read the full story ➡

Google Introduces Gemini Business and Enterprise Plans for Workspace Users

• Google launching "Gemini Business" and "Gemini Enterprise" tiers for Workspace, providing access to advanced AI capabilities from Gemini 1.0 Ultra.

• Includes data protection measures tailored for enterprises, with employees' conversations not used to train models.

• Uncertain on exact pricing, but different tiers likely based on company size; timeline separate from existing Google One AI Premium rollout

Read the full story ➡

⚙️ AI Tools and Resources

• TableFlow: 📊🚀 Streamline data import & validation

• CalenAI: 🗓️ Automatically schedules appointments through AI-powered conversations.

• The YC Bot: 💡🚀 Discover AI-powered YC insights with training from Startup School videos and practice real interviews.

• Retell AI: 🤖📢 Connect an LLM with conversational speech and Twilio integration for seamless chatbots and voice assistants

• Trademark Owl: 🎯👷🏷️ A self-service, AI-powered trademark registration platform with transparent pricing and quality control.

🔥 Trending in AI

» Google has released Magika, an AI tool that quickly and accurately detects file types. It uses a custom deep learning model to identify binaries and text files within milliseconds. Internally, Google applies Magika at scale to route Gmail, Drive, and Safe Browsing files to the right security scanners.

» Deutsche Telekom has revealed a revolutionary smartphone concept powered by artificial intelligence. The phone features an AI assistant that can fulfill requests and tasks via voice commands, negating the need for apps. Created in partnership with Qualcomm and Brain.ai, this "magenta concierge" aims to simplify daily life activities.

» OpenAI has secured new funding from Thrive Capital that skyrockets the AI startup's valuation to over $80 billion—nearly triple what it was 10 months ago. Now standing as the 3rd most valuable private tech firm behind ByteDance and SpaceX, the deal lets OpenAI employees cash out shares and provides a major vote of confidence after recent leadership turmoil.

» Major tech companies including Microsoft, Meta, Google, and others signed a pledge at the Munich Security Conference to address election-related deepfakes. Joining forces with vendors like Adobe and IBM as well as AI startups like OpenAI, they committed to detecting, labeling, and swiftly responding to misleading deepfakes on their platforms while protecting legitimate expressions.