Announcing ARC-AGI-3
The only unsaturated agentic intelligence benchmark in the world
Humans score 100%, AI <1%
This human-AI gap demonstrates we do not yet have AGI
Most benchmarks test what models already know, ARC-AGI-3 tests how they learn
Bryan Yates retweeted
Introducing an upgraded Claude 3.5 Sonnet, and a new model, Claude 3.5 Haiku. We’re also introducing a new capability in beta: computer use.
Developers can now direct Claude to use computers the way people do—by looking at a screen, moving a cursor, clicking, and typing text.
Bryan Yates retweeted
Starting today our community can test Midjourney V5. It has much higher image quality, more diverse outputs, wider stylistic range, support for seamless textures, wider aspect ratios, better image prompting, wider dynamic range and more. Let’s explore!
Bryan Yates retweeted
Announcing GPT-4, a large multimodal model, with our best-ever results on capabilities and alignment: openai.com/product/gpt-4
Bryan Yates retweeted
The next wave of AI products will revolutionize the world.
Here are the top 20 AI companies you need to know about other than OpenAI:
(based on amount of $ raised)
Had a couple of people ask me if I received measure theory or difficult proof questions when I was applying for quant research roles.
When interviewing, I spoke about measure theory at a very high level if we were talking about statistical background, but was never asked in detail on the subject.
However, I’ve absolutely been asked to do proofs on the spot. When I was applying immediately out of school, a firm or two required a written math test. See that rarely today.
Very cool. As long as I don't need to wear a hat to work to protect my alpha models I’m okay with this.
okay so AI can literally read our minds now.
a team from osaka was able to reconstruct visual images from mri scan data using stable diffusion.
first row is the image presented to the test subject, second row is the reconstructed image from mri data.
wild.
2
7
2,032
While quant trading and ML research don’t directly require such sophisticated mathematical knowledge to be effective, I am always impressed if an interviewee has a grasp on these concepts (though I will not be asking about the Radon-Nikodym theorem so don’t worry).
3
10
1,767
I personally have the hard cover from Amazon since I like physical copies, but for those looking for the pdf I see Duke’s old math website still has the PDF, services.math.duke.edu/~rtd/…
7
1,465
Bryan Yates retweeted
ChatGPT and Whisper are now available through our API (plus developer policy updates). We ❤️ developers: openai.com/blog/introducing-…
Bryan Yates retweeted
Pairs trading is not dead, you just can’t run 40 year old models - those are dead. It’s very simply an optimisation problem to create a stable, and efficient to trade, mean reverting portfolio. You can do a lot better than ADF brute forcing all bivariate pairs.
















