Rosemary DeAragon

head of industry gtm (technology, industrials, and consumer) at openai

[Rosemary DeAragon @RosemaryHua]

Today was our first Data at the Frontier launch event, and perfect timing with this data-loving bunch.

We've launched a Data agent in ChatGPT Work / Codex. It allows you to:

❓Ask questions
📊 Make interactive dashboards and report
🔌 Connect to every major data warehouse, BI tool, and data source (yes that includes our previous home Snowflake)
📖 Build a context later using agents

We're now at the point where AI can pull your query history, how recently you've used this data, what's the relationship between data sets, how is data being used --> and convert that into human editable files that capture definition and data relationships.

This is automated context layer building... Natively within the Codex harness.

So much fun to be at OpenAI where we're moving at blistering speed. Let's go!!! 🚀🔥

[Pope Leo XIV @Pontifex]

In this era of artificial intelligence, it is becoming urgent to distinguish human art from what machines produce. There is an ontological difference, even before an aesthetic one, between art and what a machine can generate through statistical calculation based on millions of images created by others. Algorithms lack the spark of humanity. For this reason, the Church wishes to renew an alliance with artists and cultural institutions to safeguard our humanity.

[Rosemary DeAragon @RosemaryHua]

@Bob_Muglia @NathanLands 100% agree with Bob here.

[Rosemary DeAragon @RosemaryHua]

Honestly I find everything Matan is saying really hard to believe. I worked for @cwdegnan and have known him for 6.5 years. He is one of the most upstanding, authentic, and we'll intentioned leaders in enterprise technology. Silicon valley is a small world, and board members are often selected for their experience across tech companies and yes, that includes competing companies. It's not uncommon to have someone who has shared experience and connections across different rocketships. This reads as inexperience with how tech veterans operate.

[Rosemary DeAragon @RosemaryHua]

At our OpenAI executive lunch, @sama and @thsottiaux discussed everything from our dots launch to the new renaissance.. an explosion of creativity in solving the most challenging problems facing humanity.

Today is the last day of the quarter, and I feel the weight of not only the short-term pressure of closing everything out, but also the larger weight of the world on translating how AI can be used for good in society.

The past few weeks have been challenging - on the one hand, we are doing real work in using technology to better the lives of everyday people. From using our AI models for renewable energy, to reducing millions of pounds of food waste in cruise lines and restaurants.

On the other hand, there is a recurring social commentary around the evils of AI. Risks are serious, and we have invested heavily in trust, safety, privacy, cyber security, and more. With that being said, am I alone in feeling like there’s an imbalance in the doom & gloom commentary, versus what AI is doing for the good of society?

More to come on AI for Good, but for now I need to get back to work and close out another massive quarter! 🚀

#DevDay #AI #AIforGood

[Rosemary DeAragon @RosemaryHua]

Besides boba milk tea, @JensenHuang might be our top Taiwanese export. It was amazing to visit NVIDIA's jaw dropping headquarters!

We have an on-site codex lab where nvidians can meet, learn, and share best practices. Our strategic partnerships spans everything from codex to data centers.

GPT6 Astra is trained on 100,000+ NVIDIA Grace Blackwell NVLink72's. As Jensen put it, they helped us bring ChatGPT from o1 to Astra in 4 years.

So excited for what we're building together.

Separately, this week's focus has continued to be on security, safety, and alignment.

One thing our industry needs to remember is that offensive attacks are fast and random. Defense is all about politics. Humans agreeing with other humans to 1) invest in tech to defend, and 2) build the defensive mechanisms, and 3) hope that we don't get cancelled for building frontier defensive capabilities along the way.

Moving quickly across large enterprises and critical infrastructure is required. It's not as simple as slowing down development in the name of safety, given the reality of other geopolitical factors at play.

If you are at an enterprise and want to talk more about bringing in our cyber models, DM me!

#AI #Cyber #OpenAI

[Rosemary DeAragon @RosemaryHua]

Today is momentous. We believe we have solved a Millenium Problem, 1 of the 7 unsolved math problems. This is absolutely incredible, an unprecedented achievement in human history. We are at the point where humans have created intelligence that supersedes human's own intelligence.

For those unfamiliar with the Millenium Problems, these were chosen in 2000 as the deepest and most consequential challenges in mathematics.

The 1 our model solved answers whether the equations we use to describe fluid motion can develop a mathematical breakdown.

As a cardiac patient, this breakthrough could help model biological blood flow and the unpredictable nature of fluid friction in blood. As a frequent flier, this could help us learn how to resolve turbulence, which at its crux is a fluid dynamics issue.

It is an amazing time to be alive!

#Astra #OpenAI #AI and now ... #AGI

[Rosemary DeAragon @RosemaryHua]

100% agree with this decision. I always felt Cursor was in a precarious position and then once they got acquired by Elon / xAI, it truly no longer made sense.

[Tibo @thsottiaux]

We unfortunately have decided that we cannot continue providing access to our models through Cursor and are ending our partnership. It boils down to trust and we’ve asked that this takes effect on November 12 to give you some time to plan.

Many have used the GPT models through Cursor and here are options we know should work in the future:

- We will continue to allow using your own OpenAI API key and similarly will continue to provide access through our IDE extensions for Cursor.
- We will keep working with the broadest range of tools and harnesses, some of which are OSS, but also many many closed-source ones.

We are as committed as ever to continue supporting developers and the flourishing ecosystem of tools, harnesses and products. We will also continue to invest in our own open-source initiatives and believe in broad optionality for developers.

You can read more about our decision in the blog:
https://t.co/Oj76Bkc2KX

[Rosemary DeAragon @RosemaryHua]

Super happy that @AnthropicAI signed alongside us at @OpenAI. We compete, but this is more important than that.

We have a limited window to strengthen our defenses to continue protecting critical infrastructure (utilities, hospitals, food supply, etc) and society as a whole.

[Greg Brockman @gdb]

An open letter for a global surge in cyber defense, signed by over 100 organizations including Anthropic, AWS, Google, Microsoft, OpenAI, and Oracle. https://t.co/uKXPS8LdAU

[Rosemary DeAragon @RosemaryHua]

From presenting to 90+ Claude users to boarding a 6am flight to meet the largest brands in the world!

We just launched Jalapeño, our first chip designed specifically to handle LLMs efficiently rather than train them.

We built this out in ~9 months, true to the lightning speed here at @OpenAI

As a Taiwanese American and daughter of a semi-conductor software engineer, I jokingly call this our family business.

My science fair project at 12 years old was building a semi-conducting plaque etcher --> using semi-conducting materials to turn the magnetic touch of a nail on and off at extremely fast speeds to etch names onto various materials.

You can tell I was a 🤓

So excited to be at this research & development company, full circle in the "family business"!

#OpenAI #Jalapeño #AI #Tech #LLMs

[Greg Brockman @gdb]

inference numbers published for jalapeno, team did an amazing job https://t.co/WNJIA2dESm

[Rosemary DeAragon @RosemaryHua]

One of the more irritating habits VCs have is pattern matching. No 2 companies are alike, and just because another company is doing it, doesn't make it right.

An important reminder from Frank who would get lectured about other "role model companies" (who aren't around anymore)

[OpenAI @OpenAI]

As we continue to push the frontier of capabilities while improving efficiency, we're dropping API and credit pricing of GPT-5.6 Sol by over 20% for the next 3 months.