Economics @ OpenAI
Bull City, USA
Joined November 2021
- Tweets65
- Following74
- Followers1.2K
- Likes10.7K
Fun piece from @AnnieLowrey ft great quote from Martha (per usual).
Feels very spiritually aligned with my essay last month; great to see more perspectives around these and similar ideas: economicsofai.substack.com/p…
Great to speak with @AnnieLowrey about why the new jobs that emerge from AI may have absolutely nothing to do with AI and what the fact that our parents all think we spend our money indulgently can teach us. theatlantic.com/ideas/2026/1…
Harrison Satcher retweeted
I used GPT-6 Astra to break an unsolved cipher to one of Napoleon's generals that had gone unread for 217 years.
What makes this impressive isn't actually the codebreaking, but that Astra completed the entire multi-modal workflow in ~6 hours from a single image and goal.
1/6🧵
Harrison Satcher retweeted
One cool thing about ARC-AGI-3 is that it measures cost until success or timeout, instead of cost of first attempt.
This is arguably more aligned with real life, and shows how costly models at high efforts can counterintuitively be cheaper, by needing less iteration.
GPT-6.1 Sol from @OpenAI on ARC-AGI (Verified):
- ARC-AGI-3: 52.7%, $7.6K (standard harness), 96.4%, $4.4K (provider adapter harness)
- ARC-AGI-2: 94.2%, $0.25/task
- ARC-AGI-1: 98.5%, $0.06/task
Its 96.4% on v3 was comparable to GPT-6 Astra's 99.9% but at a 77% lower cost.
Harrison Satcher retweeted
Welcoming Ronnie-dot to the OpenAI Economic Research team. Now someone will finally read all the papers I forward.
Announcing dots. Dots work 24/7 for you, learn from your feedback, have their own computer, browser and can be connected to over 4k apps in our ecosystem.
They’re powered by Astra our best model yet. Included in your Pro plan, without drawing down on any of your usage. You can even call a dot while it’s working.
We’re getting you started with your primary dot today and soon you’ll be able to create entire teams of them.
Harrison Satcher retweeted
This is exactly why we release data publicly. Our Enterprise Signals data shows up all over this interesting analysis by @DavidGeorge83
Read more here: openai.com/signals/enterpris…
Great video from Ethan. These types of contingencies are part of why we don't know exactly what future technologies, or jobs, will look like; you don't just have to predict 1 idea; you have to predict an interlocking series
More in latest OAI Econ blog:
economicsofai.substack.com/p…
Harrison Satcher retweeted
ok like everyone i teased the model's early 6% insta-reading. but the way this thing works: the longer the nowcast doesn't dip, the more potentially legit it gets...
Q4 here we come
On September 25, the #GDPNow model nowcast of real GDP growth in Q3 2026 is 5.0%: bit.ly/32EYojR. #ATLFedResearch
Download our EconomyNow app or go to our website for the latest GDPNow nowcast: atlfed.org/46BOWkD.
Harrison Satcher retweeted
@HarrisonSatcher makes his return to Substack. I have waited with bated breath for this day. It would behoove you to read it.
We don’t have the nouns (yet) to describe the jobs of the future.
That doesn’t mean they don’t exist
We can take seriously near-term negative impacts without treating today’s work as the boundary of what’s possible
My latest in OpenAI’s Econ Substack: economicsofai.substack.com/p…
We don’t have the nouns (yet) to describe the jobs of the future.
That doesn’t mean they don’t exist
We can take seriously near-term negative impacts without treating today’s work as the boundary of what’s possible
My latest in OpenAI’s Econ Substack: economicsofai.substack.com/p…
Would love to see a breakout in the next wave on how people update their other forecasts wave over wave, conditional on their forecast here. @EzraKarger
In the second wave of LEAP @Research_FRI, we included a question on AI solving or substantially assisting in solving a Millennium Prize Problem. Median expert forecast was 10%, and 5.4% for Superforecasters
Harrison Satcher retweeted
GPT-6 Astra is here.
We hope it will begin to enable a new generation of entrepreneurship, scientific discovery, and building.
We believe it is the best model in the world for computer use, professional work, science, coding, cybersecurity, and more.
It took us some extra time to ensure that we could meet the safety and alignment standards required for this capability level, but we think you’ll find it worth the wait.
It scores 98% on FrontierMath Tier 4, 99.9% on ARC-AGI 3, and 100% on ExploitBench.
1/ New @OpenAI Econ Research on how AI adoption is diverging
There's an 8x token output gap btwn top firms & typical.
Includes:
- more context & work surfaces through plugins (21% vs 9% use)
- more codified workflows via skills (19 to 3%)
Integration, not shallow chatter 🧵🧵
5/ my read cont.:
maybe early signs of a smile-curve on adoption, with the smallest firms able to move nimbly and build AI-first, and the largest firms able to create space to experiment and invest in complements
6/ congrats to my teammates on an incredible paper & joint Enterprise Signals release: @RonnieChatterji, @daveholtz, @sonnytambe, Gawesha Weeratunga, @neelrakholia, Cassandra Duchan Solis, & @drew_m_johnston
Paper: cdn.openai.com/pdf/how-organ…
Overview: openai.com/index/how-enterpr…
A tangible example for people of why thinking exclusively about automation or augmentation is inherently limiting.
Especially on the consumer value side, there’s just a whole lot of tasks nobody would do 1yr ago but clearly have value
good prompt. found a $1200 United flight credit that expires in 2027 and several unused e-gift cards from as early as 2023 that i completely forgot about 🤝
Harrison Satcher retweeted
@OpenAI’s Econ team have been releasing excellent research on how people use AI throughout the economy. This piece on how AI is changing what people do at work is particularly important for understanding labor market transformation. @RonnieChatterji @Alex_M_Richmond and @caroline_m_chin should have way more followers.
We're seeing how AI may change who does what at work.
@Alex_M_Richmond and @caroline_m_chin's new @OpenAI Economic Research report studies “task crossover”: work associated with one occupation appearing in another worker’s AI use. openai.com/index/how-ai-is-e…
