Thinking…

Generating…

Livestreaming…

https://openai.com/live/

But we can't do it alone; we’re also partnering with some of world's leading firms - like @Accenture, @Capgemini, CGI, @Cognizant, @Infosys, @PwC, and @TCS - to scale Codex worldwide.

The latest and more ⬇️ https://t.co/9vjsIWt7Tw (3/3)

We’re seeing meaningful growth as companies use Codex to automate workflows in engineering and beyond. That’s why we’re introducing Codex Labs - a new way to help enterprises put Codex to work on real problems. (2/3)

Codex just hit 4 million weekly active users! For those keeping track, that's an additional 1 million users in just two weeks. (1/3)

Anyone in the world can attend this conversation ⬇️ forum.openai.com/public/events/virtual-ev... (3/3)

- @gracexyzheng, Co-Founder and CEO, @perturbai_tx
- @xinjin PhD, Co-Founder, PerturbAI and Professor at @scrippsresearch and Howard Hughes Medical Institute (HHMI)
- @joyjiao12, Researcher at OpenAI (2/3)

Virtual Event: Decoding Biological Intelligence: Building AI Agents for the Brain Genome - Event | OpenAI Forum About the TalkGrace Zheng, co-founder of PerturbAI, and Xin Jin, co-founder of PerturbAI, and Professor of Scripps/HHMI, will show how we’re moving from mapping the brain to predicting how it works. In partnership with OpenAI, they’ve built one of the largest brain datasets ever created, capturing activity across 8 million cells, and are using AI to uncover the rules that drive biology.The result: a new era of biology and medicine that is more predictive, more precise, and far more effective.Zheng and Jin will also demonstrate how AI is transforming science itself, replacing complex, manual workflows with systems that let researchers explore data using everyday language, to help accelerate discovery and open the door to many more contributors.They’ll even introduce the idea of a “digital brain”, a simulation that could allow scientists to test treatments virtually, understand disease at its root, and bring life-saving therapies to patients faster than ever before.About the SpeakersXin Jin is a neuroscientist and molecular biologist whose work lies at the intersection of in vivo functional genomics and disease biology. She is an Associate Professor in the Department of Neuroscience and the Dorris Neuroscience Center at Scripps Research, and an HHMI Freeman Hrabowski Scholar. Her lab develops scalable in vivo genomic technologies to uncover how genetic programs shape brain circuits across development, homeostasis, and disease. She recently co-founded PerturbAI to help bring this vision into therapeutic discovery.Her work has pioneered in vivo Perturb-seq, a high-throughput approach that combines pooled CRISPR perturbation with single-cell readouts directly in living tissue. By integrating in vivo CRISPR screening with molecular, spatial, and whole-brain imaging approaches, her research aims to define how genes influence cell types, tissue architecture, homeostasis, and neural circuit function.Xin’s contributions have been recognized by honors including the HHMI Freeman Hrabowski Scholar Award, Sloan Research Fellowship, Pew Biomedical Scholar Award, McKnight Scholar Award, Peter Gruss Young Investigator Award from the Max Planck Society, and MIT Technology Review’s Innovators Under 35. Before joining Scripps Research, she was a Junior Fellow in the Harvard Society of Fellows. She received her PhD in Biology from The Rockefeller University and her BS in Chemistry from MIT.Grace Zheng is a genomics and machine learning leader with over 13 years of experience developing breakthrough platforms at the intersection of biology, data, and therapeutics.She is the Co-Founder and CEO of PerturbAI, an AI-native therapeutics company accelerating drug discovery through a systems-level understanding of disease. Her work focuses on integrating in vivo Perturb-seq with AI agents and models to map biological circuits directly inside intact organisms, enabling a new generation of biological models and therapeutics grounded in causal, in vivo biology.Previously, Grace served as Vice President of Computational Biology and Machine Learning at ArsenalBio, where she built and led interdisciplinary teams spanning computational biology, machine learning, and software engineering. She also led strategic collaborations with partners including Genentech and NVIDIA to generate large-scale functional genomics datasets and develop AI models for cell therapy discovery.Earlier in her career, she joined 10x Genomics as one of the company’s first employees, where she helped pioneer and launch several foundational single-cell genomics technologies.Her work centers on building enabling technologies, scaling interdisciplinary teams, and transforming biological data into predictive models that guide the next generation of medicines.Joy Jiao leads the life sciences team at OpenAI. The goal of the team is to accelerate basic research and drug discovery, by operating across the model training stack to improve model capabilities at all levels of biology, from molecules to organisms. At OpenAI, Joy previously worked on model safety, personalization, search, and representation learning. She holds a PhD in Systems Biology from Harvard, where she studied the in-patient evolution of cancer cells during immunotherapy as well as the evolution of antibiotic resistance.

🔬 What if we could predict disease before it starts, tailor treatments to each person, and discover new medicines in a fraction of the time?

🧬 Join OpenAI Forum for a conversation on how AI is decoding biological intelligence with: (1/3)

https://openai.github.io/euphony/

Introducing Euphony, an open-source tool for visualizing chat data and Codex session logs.

Paste in a public URL or upload a local file, and Euphony turns the raw data into an easy-to-browse view. It supports translation, filtering, editing, and more.

Image from Twitter

Read more:
openai.com/index/scaling-codex-to-e...

openai.com/index/scaling-codex-to-e...

Codex is now used by more than 4 million people every week.

That’s a lot of work getting done.

Thanks for building with us.

We’re proud to partner with @NABTU and the Michigan State Building and Construction Trades Council to help ensure Stargate Michigan is built by the best-trained workforce in the country. (2/2)

Quote Tweet: https://twitter.com/i/status/2046362363641942016

Demand for compute is creating one of the biggest opportunities in generations to reindustrialize the United States. (1/2)

Image from Twitter

This is not a screenshot.

Quote Tweet: https://twitter.com/i/status/2044827705406062670

ICYMI: Codex has new capabilities and is more integrated with your Mac! Whether it's launching a website for your brand, creating automations to help you work, or even generating images—everyone can be builder.

How do you use Codex?

The US cities with the most interest? It's New York (with its two teams), followed by Pittsburgh (which is hosting the draft), and then Phoenix.

Internationally, the cities outside the US that are sending the most messages are London; Hamilton, Ontario; and Toronto. (2/2)

Image from Twitter

Hundreds of thousands of ChatGPT users are sending queries related to pro football's upcoming draft, according to data studied through a privacy-preserving automated pipeline. (1/2)

Chronicle – Codex | OpenAI Developers Build Codex memories from recent screen context.

We’re starting with Pro users on macOS, except in the EU, UK, and Switzerland, while we learn where it helps most and improve the experience.

https://developers.openai.com/codex/memories/chronicle

Chronicle runs background agents to build memories from screen captures. It uses rate limits quickly.

Screen captures are stored temporarily on device to generate memories—also stored on device. You can inspect and edit memories. Be aware that other apps may access these files.

Over time, it helps Codex learn how you work: the tools you use, the projects you return to, and the workflows you rely on. (2/2)

With Chronicle, Codex can better understand what you mean by “this” or “that.”

Like an error on screen, a doc you have open, or that “thing” you were working on two weeks ago. (1/2)

Video: twitter.com/openaidevs/status/204628... (2/2)

Last week, we released a preview of memories in Codex.

Today, we’re expanding the experiment with Chronicle, which improves memories using recent screen context.

Now, Codex can help with what you’ve been working on without you restating context. (1/2)

This week, our teams hosted events from Warsaw, Poland to Abilene, Texas to Cal State, Bakersfield. These events equip builders worldwide to better their lives and the lives of others!

Where should the Academy go next? (2/2)

Image from Twitter

Image from Twitter

Image from Twitter

The OpenAI Academy is a learning platform that helps people - more than 3M so far - build practical AI skills through live events, training resources, expert content, and industry-specific communities. (1/2)

Agents that run code need a controlled workspace ready when work starts.

@modal shares why scale matters for long-running agents built with the Agents SDK.

Quote Tweet: https://twitter.com/i/status/2045207510039322668

You can just build things.

We welcome independent benchmarks like TeenAegis’ AI Danger Index that help drive accountability and a race to the top on child safety across the industry.

See the ratings for yourself - https://teenaegis.com/intelligence/ai-danger-index (3/3)

Protecting young people requires continuous work – from age-appropriate protections, robust monitoring and reporting, and stronger safeguards to initiatives like our Child Safety Blueprint and teen safety policies for gpt-oss-safeguard. (2/3)