Welcome back to AI Hungry. Our latest news update dives into Google's unveiling of the next-gen AI model, Gemini 1.5 Pro, marking a significant leap forward in cloud-based AI capabilities. As you explore the advancements in context window size and cross-modal analysis, you'll also gain insights into how to safeguard yourself against the increasingly sophisticated AI scammers using voice cloning technologies.
This edition not only highlights the cutting-edge developments in AI but also equips you with practical tips to stay one step ahead of the latest scams. Discover how you can fortify your defenses against the fake voice call frauds that have been fooling people with unprecedented realism.
Google Unveils Next-Gen AI Model Gemini 1.5 Pro on Cloud Platform
Google has announced the public preview of its advanced AI model, Gemini 1.5 Pro, on the Vertex AI Platform. This model offers the largest context window for developers, enhancing the ability to analyze extensive data and improve AI-driven services like customer support and tutoring.
Google Cloud's recent updates also include improvements to Imagen 2.0, CodeGemma, and expanded MLOps capabilities, reinforcing its AI and cloud computing offerings. Gemini 1.5 Pro's new audio processing feature enables comprehensive cross-modal analysis, broadening the scope of AI applications and solidifying Google's leadership in AI innovation.
Outsmarting AI Scammers: Tips to Protect Yourself from Fake Voice Call Frauds
As artificial intelligence (AI) advances, scammers are now using AI voice cloning technology to create convincing fake audio, posing as family members or friends in distress, to trick people into sending money. These voice clones are sophisticated enough to mimic almost anyone's voice in various languages, making it challenging to distinguish them from real human speech. However, there are strategies to protect yourself from these scams. Experts recommend hanging up and calling back using a verified number, creating a secret safe word with loved ones, or asking a personal question that only the real person would know the answer to. It's essential to be aware that anyone's voice can be cloned with just a few seconds of recorded speech, so it's crucial to stay vigilant and not give in to emotional appeals during unexpected calls asking for money or personal information.
🔍 OpenAI Unveils GPT-4 Turbo with Vision Capabilities for Developers. OpenAI has made GPT-4 Turbo with Vision generally available, offering improved speed, larger input context, and new features like JSON mode and function calling for image analysis and automation in apps. (Link)
🤖 California Startup Unveils AI Avatars to Address Workforce Shortages. SapientX introduces Vivian, an AI-powered avatar designed to assist businesses with customer interactions, showcased at Embedded World in Germany. Vivian operates on minimal computing power and can understand emotions, providing contextually relevant responses. (Link)
🎶 Spotify Introduces AI-Powered Playlist Creation Tool for Premium Users. Spotify launches AI Playlist, a feature for Premium users in the UK and Australia to create personalized playlists with natural language prompts. Users can refine their music selections and maintain control over the playlist. (Link)
🌐 Microsoft to Launch AI Hub in London, Bolstering UK's Tech Ecosystem. Microsoft is setting up a new AI hub in London to house its consumer AI research. Led by AI scientist Jordan Hoffman, the hub will collaborate with OpenAI and hire local AI talent, reinforcing the UK's position as a global AI leader. (Link)
🤖 Nvidia and Google Cloud Team Up to Boost AI Startups. Nvidia and Google Cloud are collaborating to offer startups up to $350,000 in credits for AI development, along with access to Nvidia's hardware, software, and training. This initiative aims to lower the costs and accelerate innovation in generative AI. (Link)
🚀 Intel Launches Gaudi 3 AI Accelerator, Promising Major Performance Gains. Intel's new Gaudi 3 AI accelerator aims to halve training times for large language models and increase inference throughput over Nvidia's GPUs. The 5nm chip features enhanced processing power, memory, and bandwidth, and will be available in Q2 2024. (Link)
