OpenAI’s ex-CTO launches rival lab

PLUS: New software benchmark tackles real-world freelance jobs

Author

Rowan Cheung
February 19, 2025

Good morning, AI enthusiasts. OpenAI's former CTO just unveiled her secretive new AI venture — and she's bringing some serious talent along.

With experts from top AI labs and a bold open-science vision for user-focused applications, will Mira Murati be the latest ex-OpenAI leader to rewrite the AI industry playbook?

P.S. We just launched two more newsletters! Join The Rundown Robotics for bi-weekly email updates on the Robotics industry and The Rundown Tech for bi-weekly email updates on everything happening in Tech — all in our signature rundown style.

In today’s AI rundown:

  • Mira Murati’s OpenAI rival ‘Thinking Machines Lab’

  • OpenAI’s new software engineering benchmark

  • Creating consistent visuals for presentations

  • Fiverr’s AI platform for gig workers

  • 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

MIRA MURATI

Image source: Getty Images

The Rundown: OpenAI’s former CTO Mira Murati officially brought Thinking Machines Lab, a new AI research company, out of stealth with the mission to make AI systems more “widely understood, customizable, and generally capable” through open science.

The details:

  • Thinking Machines plans to develop frontier models focused on science and programming with an emphasis on human-AI collaboration and multimodality.

  • Murati has hired a dream team for the company with OpenAI’s John Schulman and Barret Zoph as well as experts from DeepMind, Character AI, and Mistral.

  • The AI lab has also expressed commitment to open science and confirmed plans to regularly publish technical papers, code, datasets, and model specs.

  • Its introduction comes just six months after Murati abruptly left OpenAI “to create time and space for her own exploration”.

Why it matters: The move makes Murati the latest to go from OpenAI leadership to founding a rival lab, with Ilya Sutskever’s SSI also in talks to raise $1B+. While the stacked team may emerge as a major new player, its commitment to open science could be the big catalyst pushing the industry towards a more open-source mindset.

TOGETHER WITH FIREFLIES AI

The Rundown: Fireflies’s Notetaker is trusted by millions of users across platforms like Microsoft Teams and Google Meet — and now it offers real-time features that can extract instant, actionable insights from your daily meetings.

During live calls, Fireflies can:

  • Capture bullet-point notes summarizing the conversation in real-time

  • Curate action items, create comments, and bookmark key moments

  • Understand 100+ languages from different parts of the world

  • Answer questions you may have about the meeting

OPENAI

Image source: OpenAI

The Rundown: OpenAI just introduced SWE-Lancer, a new benchmark designed to measure AI’s coding performance against real-world freelance software engineering jobs — putting LLMs to the test with a total of $1M in actual task payouts.

The details:

  • SWE-Lancer features over 1,400 freelance software engineering tasks from Upwork, spanning from minor bug fixes to high-value feature implementations.

  • The benchmark evaluates both coding and technical management decisions of LLMs, challenging them to write code and select engineering proposals.

  • It introduces monetary metrics, with success measured by how much a model could theoretically "earn" by completing tasks correctly.

  • All top models struggled on the benchmark, with Claude 3.5 Sonnet performing best — solving nearly half of the tasks and earning $400k out of the $1M.

Why it matters: The benchmarks are increasing their difficulty to try and properly evaluate increasingly capable AI, but it’s hard to see any of these tests standing the test of time. Plus, while models “struggled” on the benchmark, $400k of value is no joke — and is a good example of the scale of displacement about to arrive in dev work.

AI TRAINING

The Rundown: Freepik lets you maintain a unified presentation style by creating custom AI styles that generate consistent visuals across all your slides.

Step-by-step guide to use it:

  1. Visit Freepik AI Suite and select the “Create” option in the main dashboard.

  2. Click the plus icon in the Style section.

  3. Upload 10-50 reference images and select Ultra/High quality for optimal results.

  4. Use targeted prompts for each slide type and generate visuals with your custom style at 100% strength for consistency.

Pro tip: Check out our exclusive workshop on creating custom styles with Freepik here. Additionally, The Rundown University members enjoy a free monthly premium Freepik plan.

PRESENTED BY HEROKU

The Rundown: Heroku’s new Managed Inference and Agents platform lets developers easily deploy, scale, and manage AI-augmented applications and agents.

Heroku’s platform offers:

  • Streamlined AI integration

  • Advanced AI tools

  • Scalable AI logic

Join Heroku’s pilot program for early access to powerful tools that can put AI to work for you.

FIVERR

Image source: Fiverr

The Rundown: Freelance service platform Fiverr just launched Fiverr Go, a new suite of AI tools that lets gig workers train models on their work and automate future jobs, while also announcing an equity program giving top performers shares in the company.

The details:

  • Freelancers can train personal AI Creation Models for $25/mo, allowing them to sell AI-generated versions of their work while retaining ownership rights.

  • A $29 monthly Personal AI Assistant helps manage client communications and handle routine tasks, using past interactions to provide customized responses.

  • Access is initially limited to "thousands" of vetted Level 2 and above freelancers in specific categories like voiceover, design, and copywriting.

  • The company is also launching an equity program that will give top-performing freelancers shares in Fiverr, though specific details haven't been disclosed.

Why it matters: AI is in the process of upending traditional gig work, and Fiverr is attempting to give freelancers a stake in automation instead of competing against it. While the platform could help some creators scale, it also will likely face some backlash from creatives who may feel that opting out of AI is becoming unavoidable.

QUICK HITS

  • 🤖 Grok-3 - xAI’s new SOTA next-gen reasoning model (slowly rolling out)

  • 🌎 Mistral Saba - Language model designed for Middle Eastern and South Asian cultures and linguistics

  • 🧠 DeepHermes 3 Preview - Nous Research’s new 8B model with the ability to balance reasoning and speed

  • 🤝 AndSend - AI Customer Relationship Agent that spots opportunities, suggests timely messages, and helps you reach your goals

OpenAI CEO Sam Altman posted a poll on X asking what project users would like to see open-sourced, with an “o3-mini” level model leading over a “phone-sized model.”

Elon Musk announced the launch of xAI’s Gaming Studio during its Grok-3 demo, with plans to build games with AI.

xAI also revealed that a Voice Mode for Grok will go live in ‘about a week’, with the company providing a brief teaser at the end of the recent demo.

HP acquired Humane’s AI software platform and team for $116M and will discontinue its AI Pin hardware, with plans to integrate AI capabilities across HP’s device portfolio.

Meta announced Llamacon, the company’s first dedicated generative AI developer conference, for April 29.

Google rolled out new AI features to Google Meet, including a scrollable caption history allowing users to review up to 30 minutes of live and translated captions.

COMMUNITY

Join our next workshop today at 3 PM EST and learn how to create custom AI soundtracks for your business content with Udio’s Andrew Sanchez, CEO/Co-founder, and Ian Braunstein, Head of Product.

RSVP here. Not a member yet? 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.