Data & Analytics @OpenAI | ex-Public Policy Associate | tech, policy & coffee
sf
Joined August 2008
- Tweets1.5K
- Following417
- Followers6.5K
- Likes744
Something like Screen Time that connects your personal data across apps and spots patterns you wouldn't notice yourself, helping you manage your time, stay on top of priorities, and figure out where your days actually go. Feels like there's a lot of potential there.
👀
We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory.
openai.com/index/building-sc…
Jeff Joseph retweeted
6.1 Terra is here! It is good at a lot of things, but we are especially excited for this new era of computer use.
Jeff Joseph retweeted
During this red-teaming exercise, we also found a vulnerability that allows for unauthenticated remote code execution in Brotli, a widely-deployed compression library included in many pieces of software such as Node.js, Bun, and Envoy.
The flaw could let an attacker run their own code on affected systems by sending or replying with malicious compressed data, without any need to create a user account. That is, an unpatched server that accepts compressed input could be taken over by a malicious client, and an unpatched client that accepts compressed input (such as “curl --compressed”) could be taken over by a malicious server.
This root cause issue was actually fixed in the Brotli source code in February, but it wasn’t known at the time that it gave rise to a security vulnerability and there was no more recent release of Brotli. Since Brotli is built into other software, any software which includes this library needs to take the update too, followed by the users of that software. We disclosed the vulnerability to Brotli and other downstream projects such as Node.js, which released or are currently releasing security updates, and Brotli cut new release versions and published a CVE.
More details on the patch and mitigations below:
patchtheplanet.org/brotli
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
Really interesting work. Independent researchers evaluating how we monitor internal agents for potential misalignment. From a data and analytics perspective, the challenge is making these evaluations measurable, identifying gaps in our monitoring, and ensuring the signals we track actually capture the behaviors that matter.
Jeff Joseph retweeted
You can just build things with Codex
We built a site for you to explore hardware projects from the community and share your own.
Jeff Joseph retweeted
Codex is one of the most important untold stories in the history of technology.
A year ago, it was a scrappy effort in a quiet corner of @OpenAI. Its cloud product hadn’t found PMF, and the team was starting over from scratch.
Less than 12 months later:
- Codex swallowed the flagship ChatGPT desktop app from the inside (it literally runs as com.openai.codex on your Mac)
- Codex and ChatGPT for Work hit 40M+ weekly users across engineering and knowledge work
- Its leader @thsottiaux was elevated to run all of OpenAI’s core products, including ChatGPT, Codex, and the API
How did they pull it off?
They trained their own dedicated model line, separate from ChatGPT, with researchers and product builders sitting on the same team
They built four different products in a single year (cloud → terminal → IDE sidebar → a desktop app with no code editor at all) as every model leap made a new form factor possible
They used Codex to build Codex—so a handful of people could prove a new product in weeks instead of arguing for it in meetings
I had a front-row seat to the whole arc—from their failed Cloud launch, to OpenAI’s offices the night before GPT-5 dropped, to the day Codex took over ChatGPT.
And I spent months interviewing @gdb, @thsottiaux, @embirico, @mia_glaese, @aambrosino, and @edbayes to get inside how it all went down.
Here is the full, behind-closed-doors story of the team that remade OpenAI:
every.to/p/how-codex-happene…
We’re pushing the boundaries of how stories like this are told.
This piece is completely agent-native—designed to be read alongside your agent inside ChatGPT’s in-app browser:
- Every interview transcript is included in full: Click any quote and you (or your agent) can see the exact conversation it came from.
- Built for your agent to explore: Ask your agent to pull out product lessons for your team, cross-examine the transcripts, or find insights I missed when writing the piece.
It’s a story about the future of software—built to be read with an agent:
every.to/p/how-codex-happene…
We’ve had the chance to use this at OpenAI already, and tbh, it’s been a real time saver on the long run. Being able to pull context from multiple tabs without jumping back and forth makes a bigger difference than you’d think. Now everyone gets to experience it 🙌
Also, a little help managing my 47 open tabs is always welcome 😂
A few big upgrades for Chrome users today! We’re bringing the power of computer use to anyone who uses Chrome.
First up: ChatGPT can now use your tabs in Chrome to make answers more useful and get things done for you.
If you’re in Chat, it can look across your open tabs and gather info.
If you’re in Work, ChatGPT can control your Chrome tabs to get things, like sending messages, filling out forms, or downloading files. It can even help you organize and group that pile of tabs you have open.
Jeff Joseph retweeted
A few big upgrades for Chrome users today! We’re bringing the power of computer use to anyone who uses Chrome.
First up: ChatGPT can now use your tabs in Chrome to make answers more useful and get things done for you.
If you’re in Chat, it can look across your open tabs and gather info.
If you’re in Work, ChatGPT can control your Chrome tabs to get things, like sending messages, filling out forms, or downloading files. It can even help you organize and group that pile of tabs you have open.
Turning complex biological data into actionable scientific insight could be one of AI’s most exciting opportunities. I think GPT-Rosalind brings that vision closer empowering researchers to connect evidence, test hypotheses, and accelerate discovery responsibly 🔥
GPT-Rosalind is built to support that work: helping researchers reason across scientific literature and biological data, with scientists bringing the expertise and judgment to evaluate the evidence.
Eligible labs can receive up to 12 months of sponsored access to OpenAI tools, with GPT-Rosalind access subject to trusted-access approval.
Research data shared through the Rosalind Academic Plan is not used to train our models by default.
Fascinating to see how well it anticipates intent. The gap between having an idea and bringing it to life keeps getting smaller.
Jeff Joseph retweeted
As AI continues to advance, I think it’s even more important for humans to come together, connect, & build a shared vision for the future.
Six weeks into my role on OpenAI's DX community team, I’ve seen that happen through our Codex Ambassador events...
Really interesting to see how much room there is to optimize inference costs without sacrificing quality. Feels like context allocation is an underrated part of building efficient AI systems.
Jeff Joseph retweeted
ChatGPT now ships with Generative UI
A few hundred million people are now going to reset their expectations: AI answers with live interfaces not just paragraphs.
This is the future we have been building towards. If you want this in your own agent, that's exactly what OpenUI is for