Google’s Gemini 2.0 holiday surprise

Welcome, AI enthusiasts.

Gemini 2.0 is here, and it’s a large breakthrough for AI agents — with latest multimodal capabilities, deep research tools, and seriously impressive projects.

The ‘12 Days of OpenAI’ has been hogging the vacation highlight, but Google just crashed the party in a BIG way. Let’s get into it…

In today’s AI rundown:

• Google releases Gemini 2.0 with AI agent capabilities

• ChatGPT involves Apple Intelligence

• Transform AI into your personal code tutor

• Apple Intelligence gets an enormous upgrade with iOS 18.2

• 5 latest AI tools & 5 latest AI jobs

• More AI & tech news

Read time: 4 minutes

LATEST DEVELOPMENTS

GOOGLE

✨ Google releases Gemini 2.0 with AI agent capabilities

The Rundown: Google just launched Gemini 2.0, marking a significant leap into AI agents with enhanced multimodal abilities, native tool integration, deep research generation, and a number of other experimental projects.

Gemini 2.0

• Gemini 2.0 Flash debuts as a faster, more capable model that outperforms the larger 1.5 Pro on several benchmarks while maintaining similar speeds.

• The model now generates images and multilingual audio directly and processes text, code, images, and video.

• Gemini 2.0 Stream Realtime is available without spending a dime (versus the $200/mo ChatGPT Pro) and allows for text, voice, video, or screen-sharing interactions.

Project Highlights (not publicly available yet):

• Project Astra brings multimodal conversation abilities with 10-minute memory, native integration with Google apps, and near-human response latency.

• Project Astra can be being tested on prototype glasses, and it plans to eventually be utilized in products just like the Gemini app.

• Project Mariner introduces browser-based agentic AI assistance through Chrome, achieving 83.5% accuracy on web navigation tasks.

• Jules, a brand new coding assistant, integrates directly with GitHub to assist developers plan and execute tasks under supervision.

• Recent gaming-focused agents can now analyze gameplay in real time and supply strategic advice across various game types.

Deep Research:

• Deep Research is a brand new agentic feature that acts as an AI research assistant, now available in Gemini Advanced ($20/mo) on desktop and mobile web.

• Abilities include creating multi-step research plans, analyzing info from across the online, and generating comprehensive reports with links to sources.

Why it matters: OpenAI had the vacation momentum, but Google stole the show. Gemini 2.0 brings some extremely powerful upgrades, including one in every of the most important steps towards useful, consumer-facing agentic AI that we’ve seen yet. Projects like Astra could also set a brand new standard for a way we interact with AI heading into 2025.

ICYMI: We sat down with Google DeepMind CEO Demis Hassabis to debate the huge day of launches. Watch the total discussion on YouTube.

TOGETHER WITH RABBIT

🐰 Unlimited AI agents in a single orange package

The Rundown: The rabbit r1 AI companion consolidates your digital experience right into a single device, featuring an modern operating system that learns from interactions and keeps improving through continuous updates and enhanced LAM technology.

Customize your AI agent experience with:

• LAM playground for creating agents that navigate the online, check info, and take actions

• Teach mode beta for no-code agent customization

• Generative UI for personalized interfaces in any style with natural language prompts

Benefit from the one discount of the 12 months and secure your r1 this holiday season.

OPENAI & APPLE

📱 ChatGPT involves Apple Intelligence

The Rundown: OpenAI just announced the combination of ChatGPT into Apple Intelligence during Day 5 of the corporate’s live stream event, allowing Siri to leverage the AI assistant for brand new writing, reasoning, and visual tasks.

The main points:

• ChatGPT now seamlessly integrates with Siri on iPhone 16 and 15 Pro, routinely triggering when queries would profit from advanced AI reasoning.

• Visual Intelligence on iPhone 16 models can use ChatGPT to research and supply insights on images, as demonstrated in a Christmas sweater contest.

• The mixing also extends to systemwide Writing Tools, allowing users to generate content and pictures with ChatGPT directly inside Apple apps

• Users can access ChatGPT’s capabilities without an account, with built-in privacy protections stopping data storage and IP tracking.

Why it matters: While this can be a powerhouse collaboration between Apple and OpenAI, the thrill was definitely louder in the summertime. But given Apple Intelligence’s lackluster start, this could possibly be a catalyst to drive excitement and latest use cases while giving Siri a needed boost to compete with other quickly advancing voice assistants.

AI TRAINING

🎓 Transform AI into your personal code tutor

The Rundown: Google’s newly released Gemini 2.0 ‘Stream Realtime’ feature means that you can get real-time coding guidance like a private tutor looking over your shoulder.

Step-by-step:

1. Visit Google AI Studio or find ‘Stream Realtime’ within the sidebar menu.

2. Click “Share your screen” to let AI see your coding environment.

3. Start coding and ask questions naturally through voice, like talking to an actual tutor.

4. Get immediate feedback and guidance as you write code.

Pro tip: Start with easy programming concepts and step by step increase complexity. You too can provide System Instructions for the model.

PRESENTED BY TURING

📶 Optimize your LLM strategy

The Rundown: Turing provides comprehensive post-training services for leading AI labs, specializing in model evaluation, SFT, RLHF, and DPO to boost models’ reasoning, coding, and multimodal capabilities.

Use their free, 5-minute assessment to:

• Discover your position within the GenAI journey

• Get personalized recommendations for advancement

• Align your strategy with business objectives

Start your assessment now.

APPLE

🍎 Apple Intelligence gets an enormous upgrade with iOS 18.2

The Rundown: Apple just rolled out its biggest Apple Intelligence update yet, AI-powered emoji creation, image generation capabilities, Visual Intelligence with camera control, and more — alongside the broader integration of ChatGPT.

The main points:

• Genmoji is now live and allows users to create custom AI-generated emojis from text descriptions or photos with options so as to add accessories and themes.

• Image Playground adds AI image creation across the system, with dedicated app access and integration into apps like Messages and Keynote.

• Visual Intelligence debuts as an iPhone 16-exclusive feature, using Camera Control to research surroundings and supply info through Google or ChatGPT.

• Apple Intelligence also expands to latest regions with localized English support, including the UK, Australia, Canada, and others.

• As revealed within the Day 5 livestream, Siri gains ChatGPT integration, letting users tap OpenAI’s capabilities directly without switching apps.

Why it matters: Apple Intelligence has been underwhelming thus far, to say the least, however the ChatGPT integration brings the system closer to what users likely envisioned when upgrading their iPhones for the brand new AI tools. Nevertheless, we’ll need to wait until 2025 for agentic Siri capabilities that may handle more complex actions.

NEW TOOLS & JOBS

Trending AI Tools

• 🧊 AISmartCube – A low-code platform to construct, automate, and enhance AI tools and assistants

• 📣 Remention – Place your product in billions of online discussions with AI with product keyword tracking and natural replies

• ⚙️ SmythOS – Construct, debug, and deploy AI agents in minutes

• 🗣️ Shortcut – AI Assistant to ask questions, organize ideas, or roleplay conversations all through natural dialogue

• 📊 Bricks – AI spreadsheet that does the be just right for you via natural language prompts with no formulas or data cleanup

Recent AI Job Opportunities

• 🔎 Author – Director of GTM Recruiting

• ⚖️ Synthesia – Business Counsel (DACH)

• 🤖 Captions – Research Engineer, Machine Learning

• 📜 OpenAI – Paralegal

• 👥 Twelve Labs – People Operations Manager

QUICK HITS

Midjourney founder David Holz unveiled ‘Patchwork’’ a collaborative multiplayer worldbuilding tool, alongside plans for multiple model personalization and video generation capabilities coming in 2024.

Google Cloud launched its Trillium TPUs, which feature 4x faster AI training and 3x higher processing speed than the previous generation — powering Gemini 2.0 and now available for patrons.

Microsoft AI CEO Mustafa Suleyman is putting together a brand new London-based consumer health division, recruiting key ex-DeepMind health experts including former unit head Dominic King.

Apple is reportedly developing its own custom AI server chip with Broadcom, hoping to cut back its dependency on Nvidia like other tech giants.

Russia announced the formation of a BRICS AI Alliance Network to challenge Western dominance in artificial intelligence development, with initial partners including Brazil, China, India, and South Africa.

Former Snap AI lead Alan Bekker launched eSelf, a brand new platform for creating video-based AI agents with sub-2-second response times, emerging from stealth with $4.5M in seed funding.

THAT’S A WRAP

See you soon,

Rowan, Joey, Zach, and Alvaro—aka The Rundown Team