OpenAI goes nuclear

PLUS: Google’s new AI handles phone calls

Good morning, AI enthusiasts. OpenAI has been making plenty of moves with the U.S. government, but it might’ve just been handed its most sensitive assignment yet — helping protect the nation's nuclear arsenal.

With the company’s models now deployed across National Labs for nuclear science, physics, and disease research, AI is quickly becoming a national security cornerstone.

Reminder: Our next workshop is today at 4 PM EST! Learn about how to automate tasks with OpenAI’s Operator and create your own free, personalized, offline assistant.

In today’s AI rundown:

  • OpenAI teams up with National Labs on nuclear security

  • Google’s new AI handles phone calls

  • Create AI visuals containing legible text

  • Riffusion launches free AI music platform

  • 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

OPENAI

Image source: Reuters

The Rundown: OpenAI just announced a new partnership with U.S. National Laboratories, giving thousands of government scientists access to its most advanced AI models for critical research, including nuclear weapons security.

The details:

  • 15k scientists will gain access to o1 models, with OpenAI supporting research including cybersecurity, power grid protection, disease treatment, and physics.

  • The company will also deploy an AI model on Los Alamos' Venado supercomputer in partnership with Microsoft.

  • OpenAI researchers with security clearances will consult on nuclear security projects focused on weapons safety and nuclear war risk reduction.

  • The partnership follows OpenAI's recent release of ChatGPT Gov, a specialized platform for government use across federal agencies.

Why it matters: AI systems are being integrated into the U.S.’s most sensitive national security infrastructure, and OpenAI continues to position itself as a crucial part of the country’s tech development. But if AI’s scientific capabilities reach even a fraction of the level many expect, it will become a global government necessity, not just a priority.

TOGETHER WITH HUBSPOT

The Rundown: HubSpot’s ‘AI For Business Builders’ free guide shows you how to join the savvy entrepreneurs quietly leveraging artificial intelligence for real business growth — with no P.H.D. needed.

This no-BS guide cuts through the hype, giving you:

  • A jargon-free breakdown of large language models (the engine behind AI's magic)

  • Concrete steps to leverage AI for immediate business growth

  • Real-world examples that turn "prompt engineering" from gibberish into a secret weapon

GOOGLE

Image source: Google Labs

The Rundown: Google just released two new Search Labs experimental features that handle phone calls on a user's behalf, including ‘Ask for Me’ for inquiring about local services and ‘Talk to a Live Representative’ for customer service calls.

The details:

  • ‘Ask for Me’ can contact local businesses to gather pricing and availability information for services like auto repairs and nail salons.

  • Users enter requirements through a search interface, with Google's AI handling the phone call and providing a summary via text or email within 30 minutes.

  • A separate ‘Talk to a Live Representative’ feature waits on hold with customer service lines and alerts users when a representative is available.

  • Both features utilize Google's advanced Duplex AI technology for natural-sounding voice interactions.

Why it matters: With much of the younger generation hating phone interactions, these new Google features could have major mass appeal. With much of customer service also transitioning to artificial intelligence, the future is (for better or worse) likely looking like a wave of AIs calling other AIs.

AI TRAINING

The Rundown: Ideogram turns text into eye-catching images for free, with a special ability to incorporate legible text into the generated visuals.

Step-by-step:

  1. Sign up for a free account at Ideogram (you'll get free daily generations).

  2. Click "What do you want to create?" on the dashboard and craft a detailed prompt.

  3. Adjust settings like aspect ratio and magic prompts and hit "Generate" to get four unique images.

Pro tip: After generating your images, use "Remix" for subtle tweaks, "Retry" for new variations, or "Upscale" for higher resolution.

PRESENTED BY SECTION

The Rundown: Join Jan Pedersen, Walmart’s VP of Search & Personalization, and Greg Shove, Section CEO for an in-depth look into how companies can implement AI-driven personalization at scale.

Join the free session on Feb. 5 to learn:

  • How AI is transforming the way businesses deliver personalized experiences

  • Insights from real personalization strategies at Walmart

  • Risks and opportunities in implementing AI-driven personalization across industries

RIFFUSION

Image source: Riffusion

The Rundown: San Francisco startup Riffusion just launched Fuzz, a new free generative AI music platform that allows users to create full-length songs and can adapt to a user’s preferences and tastes over time.

The details:

  • The platform allows users to create full-length original music through simple text prompts, audio snippets, or image inputs.

  • Fuzz also features an adaptive learning element, allowing the model to learn a user’s musical preferences through their generations and profiles.

  • The company raised $4M in 2023, with electronic music group The Chainsmokers acting as advisors and testers for the platform.

Why it matters: This has been a big week for free AI music, with the open-source YuE and now Fuzz hitting the scene. With increasing capabilities and lower costs, it's clear there is an AI music revolution coming — and it's likely that many songs in the wild are already leveraging the tech in some capacity without listeners even realizing it.

QUICK HITS

  • 🎥 Upscale to 4K - Luma Labs’ new feature to dramatically increase the resolution of a generated video

  • 🧠 Think Deeper - New reasoning capabilities available to all Copilot users

  • 📝 Stella - Automate tasks like meeting invites, emails, and note-taking with AI

  • 📣 Kiva - AI-powered SEO agent for agencies, SMEs & startups

  • 🤖 The Rundown - Writer (Robotics/Tech)

  • 🧠 OpenAI - Research Engineer/Scientist, Model Fusion

  • 🏠 Waymo - Real Estate Development Manager

  • 💼 Kumo - Account Executive (New York)

OpenAI is reportedly in talks to raise up to $40B at a staggering $340B valuation, potentially more than doubling its worth from late 2024.

Google is rolling out Gemini 2.0 Flash across its mobile and web apps, featuring faster responses, improved image generation through Imagen 3, and enhanced performance.

Krea AI teased Krea Chat, an upcoming tool powered by DeepSeek that provides a text interface for generating and editing images and videos on the platform.

Mistral released Small 3, a 24B-parameter open-source model that matches the performance of 70B models at 3x the speed and is deployable on consumer hardware.

Sakana AI unveiled TinySwallow-1.5B, a new tiny Japanese language model that can run offline on smartphones and achieves top performance among similar-sized models.

ElevenLabs officially announced a $180M Series C funding round, bringing the AI speech startup’s valuation to over $3B.

AI2 introduced Tülu 3 405B, the Seattle-based non-profit’s largest open-source model yet that surpasses DeepSeek V3 and GPT-4o on certain benchmarks.

COMMUNITY

Join our next workshop this Friday at 4 PM EST to learn about how to automate tasks with OpenAI’s Operator and create your own free personalized, offline assistant with Dr. Alvaro Cintas, The Rundown’s AI professor.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

We’ll always keep this newsletter 100% free. To support our work, consider sharing The Rundown with your friends, and we’ll send you more free goodies.

That's it for today!

Before you go we’d love to know what you thought of today's newsletter to help us improve The Rundown experience for you.

Login or Subscribe to participate in polls.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team

Reply

or to participate.