Hey Genesis Residents!

AI is blowing up this week! 🔥

From advancements in AI safety policies to groundbreaking nuclear energy partnerships and AI-driven video creation

Here’s everything you need to know.

Let’s get into it

📝 Today’s Menu

• 🧠 Nvidia Launches Nemotron 70B: Crushing GPT-4

• ☢️ Google’s Groundbreaking Deal for Nuclear SMRs

• 🐝 OpenAI Unveils Swarm Multi-Agent Framework

• 🎥 Adobe Unveils Firefly Video Model at MAX

• 🛡️ Anthropic Elevates AI Safety Measures

• 🚘 Tesla Unveils Autonomous ‘Robotaxi’ and ‘Robovan’

• 🔥 Quick AI Developments

• 🤖 Trending AI Tools

^{Read time: 15 minutes}

MAIN AI UPDATES

^_NVIDIA
🧠 Nvidia Launches Nemotron 70B: Crushing GPT-4

Nvidia, a giant in both AI and chipmaking, has quietly released a new large language model (LLM) called Llama-3.1-Nemotron-70B-Instruct.

This fine-tuned, open-source model is already outperforming other major AI systems like GPT-4o and Claude 3.5 Sonnet on various benchmarks.

Breakdown:

📈 Built on Meta’s Llama 3.1: Nemotron is based on the Llama 3.1 70B model but has been fine-tuned by Nvidia using techniques like Reinforcement Learning from Human Feedback (RLHF).

📈 Benchmark Success: The model achieved top scores on key alignment benchmarks, such as Arena Hard (85.0), AlpacaEval 2 LC (57.6), and GPT-4-Turbo MT-Bench (8.98).

📈 Outperforming Larger Models: Despite its smaller size (70 billion parameters), Nemotron is outperforming giants like GPT-4o and Claude 3.5 Sonnet on several key metrics.

Nvidia’s move into the LLM space is particularly exciting because it’s not just about the hardware anymore.

The release of Nemotron illustrates that smaller, more efficient models can compete with industry giants, challenging the belief that bigger always means better.

Nvidia’s focus on open-source models opens doors for wider accessibility and innovation within the AI community.

QUICK WORDS FROM TODAY’S SPONSOR

^{_{GOOGLE x NUCLEAR POWER}}
☢️ Google’s Groundbreaking Deal for Nuclear SMRs

In a groundbreaking deal, Google has become the first corporation to sign an agreement to purchase nuclear energy from a fleet of small modular reactors (SMRs).

The reactors, built by California-based Kairos Power, are expected to supply up to 500 MW of clean energy to Google’s U.S. operations by 2035.

Key Details:

♻️ SMRs as the Future of Clean Energy: Unlike traditional nuclear reactors, SMRs are smaller, factory-built, and offer quicker deployment, making them a more practical solution for powering AI data centers.

♻️ Reliable, Round-the-Clock Power: The deal is part of Google’s broader clean energy strategy, offering reliable 24/7 power that is essential for energy-hungry AI systems.

♻️ Scaling Nuclear Energy: Google’s bet on SMRs isn’t just about powering its own operations. These reactors could potentially serve entire communities, significantly reducing carbon emissions across industries.

As the demand for energy in AI continues to rise, finding sustainable, reliable energy sources is crucial.

Google’s move toward nuclear energy sets a precedent for the tech industry, as other giants like Microsoft and Amazon are also exploring nuclear options.

This development could help revolutionize clean energy adoption within the sector.

^{_{OPENAI NEW UPDATE}}
🐝 OpenAI Unveils Swarm Multi-Agent Framework

OpenAI has revealed a new experimental framework named Swarm, which aims to simplify the creation and management of multi-agent AI systems.

This open-source project will make it easier for developers to coordinate multiple AI agents to work together on complex tasks.

Framework Features:

⚡ Lightweight Coordination: Swarm is designed to be simple and controllable, with key features including agents (encapsulated with specific instructions and tools) and handoffs (allowing agents to pass tasks between each other).

⚡ Open-Source: The framework is available on GitHub, providing a resource for developers interested in building and experimenting with multi-agent AI systems.

Swarm represents a leap forward in multi-agent systems, which could enable users to oversee entire “teams” of AIs working in unison to complete multi-step, complex tasks.

The potential for these systems to revolutionize industries and workflows is immense.

^{_{ADOBE UPDATE}}
🎥 Adobe Unveils Firefly Video Model at MAX

Adobe has expanded its Firefly AI model to include new video generation capabilities, marking a major leap in creative AI tools.

The announcement came during Adobe’s MAX Conference, where a host of AI-powered updates to its creative suite was revealed.

What’s New?

🎬 Firefly Video Model: The model, now in limited public beta, allows users to generate video content from text prompts or images. These capabilities extend to generating cinematic footage, 2D and 3D animations, and integrating text graphics and screen effects.

🎬 Premiere Pro Enhancements: Adobe has introduced a Firefly-powered feature called “Generative Extend,” designed to streamline clip extensions, transition smoothing, and other editing tasks.

With Firefly Video Model, Adobe is setting the stage for more accessible, AI-powered content creation.

As the race to dominate AI video generation heats up, Adobe’s integration of AI into its popular Creative Cloud ecosystem gives it a strong competitive edge.

_{Tweet of the Day}

LATEST DEVELOPMENTS

^{_{AI SAFETY POLICY}}
🛡️ Anthropic Elevates AI Safety Measures

Anthropic has rolled out a substantial update to its Responsible Scaling Policy, introducing fresh safeguards and governance measures that aim to fortify the safety of advanced AI systems.

This move places Anthropic at the forefront of responsible AI development, continuing its commitment to safety in the rapidly evolving world of AI.

Key Highlights:

🔎 New Thresholds: Anthropic has introduced ‘Capability’ and ‘Required’ thresholds, which will trigger additional safety measures when AI models reach certain levels of risk, particularly in the areas of bioweapons and autonomous AI research.

🔎 Governance Role: The company plans to regularly evaluate its AI models, overseen by a newly established "Responsible Scaling Officer," ensuring that policy implementation remains compliant with safety standards.

🔎 Transparency Pledge: Anthropic will increase its transparency by publicly disclosing capability reports and integrating input from external experts.

As AI technologies grow more powerful, the risks associated with their use increase.

Anthropic’s comprehensive safety-first approach sets a high bar for the industry, hinting at a wave of AI regulation on the horizon.

It also suggests that Anthropic might be readying for some major AI advancements soon.

^_TESLA
🚘 Tesla Unveils Autonomous ‘Robotaxi’ and ‘Robovan’

Tesla continues to push the envelope in autonomous vehicle technology, revealing new self-driving vehicles designed to operate without traditional controls.

Highlights:

🚕 Robotaxi (Cybercab): Tesla's Cybercab will be priced under $30,000, lacks a steering wheel or pedals, and features gull-wing doors. With operating costs as low as 20 cents per mile, this vehicle is poised to revolutionize personal transportation.

🚕 Robovan: This vehicle, equipped to transport up to 20 passengers or large cargo loads, is part of Tesla’s broader vision for autonomous public transport solutions.

🚕 Optimus Personal Assistant Robots: Tesla is also working on Optimus, a robotic assistant capable of tasks such as lawn mowing, grocery fetching, and babysitting.

Tesla’s push into fully autonomous, cost-effective vehicles hints at a future where personal transportation could be radically transformed.

The introduction of the Robotaxi and Robovan could disrupt traditional transport industries, and Musk’s vision of autonomous robots performing everyday tasks could change the way we think about domestic work and personal convenience.

QUICK AI BITES

⚡ Google Japan has released the Gboard double-sided, also known as the “Infinity Keyboard,” designed for multiple users to type or code simultaneously.

 📉 Google's share of the search ad market may fall below 50% for the first time in a decade due to the rise of AI search engines

💹 Perplexity expands on finance-based queries. It now shows real-time stock prices, compares companies, deep dives into financials and more.

🤖 Mistral releases Ministral Edge LLMs optimized for laptops and phones. It generally outperforms Google’s Gemma 2 2B and 9B and Llama 3.2 3B and 3.1 8B on benchmarks.

🚀 Reddit has launched AI keyword-targeting with dynamic audience expansion, multi-placement optimization, AI keyword suggestions, and unified targeting flow.

👀 X (formerly Twitter)’s ToS Update Allows AI Training on Your Data. A royalty-free permission to use, reproduce, modify, and the right to use your content for AI training for Grok. 

🤖 Trending AI Tools

🖼️ Magnific Mystic V2 - Advanced AI generator that can output up to 4k resolution images

⏱️ Motion - Automatically plan your day based on your tasks and priorities

🦙 LlamaCodeEditor > Allows users to generate an entire app from a text prompt using Meta’s latest and Llama 3.1 405B

📽️ Glato AI - Create video ads with expressive AI creators from a product URL 

🗣️ Synthesia - Create videos with AI avatars and voiceovers in 130+ languages

Thanks for reading!

If you enjoyed this, please help spread the love by forwarding this Newsletter to a friend or colleague.

SPONSOR US

Get your product in front of over 4000+ AI enthusiasts

Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners around the world. 

Get in touch today!

FEEDBACK

I hope to see you in the next one!

AI Genesis