Sean J Taylor
seanjtaylor.com
Sean J Taylor
@seanjtaylor.com
Data Science, Causal Inference, Economics, Statistics and Machine Learning

Two joys in life:
[1] Learning something interesting about the world
[2] Telling people about it

Data Science @OpenAI
Reposted by Sean J Taylor
I did update the figure a bit and added labels, but it probably works better as sequential slides in a presentation. If you want the code and the pptx just drop me a dm and I’ll send it all your way.
April 10, 2025 at 8:13 AM
I have an AI agent and the company has an AI agent, and they exchange a series of messages to converge on a contract. But they are very lazy, so the company's agent just reveals the marginal cost and my agent reveals my willingness to pay, and they take the average.
December 16, 2024 at 4:00 AM
I have a theory for why they are so useful. You can think of time as the universal cause of all outcomes. Effects may be mediated by other variables but the passage of time is the root cause of every change.
December 4, 2024 at 5:10 AM
I've heard it described as the methadone to Twitter's heroin.
November 15, 2024 at 4:51 PM
It’s so cool that the data is open. Creates a lot of unique opportunities.
November 14, 2024 at 4:46 AM
Reposted by Sean J Taylor
Other examples of things that the (rapidly growing) data geeks on Bluesky can help with: improving and accelerating graph analytics to improve the default Discover feed:

bsky.app/profile/dana...
danabra.mov dan @danabra.mov · Nov 13
@why.bsky.team updated cluster mappings yesterday and my Discover feels like it’s gotten a whole lot more relevant. not sure if real or imagined effect. it would be nice to figure out a way to run this thing every day
November 13, 2024 at 12:33 PM
The gang’s all here!
November 14, 2024 at 3:40 AM
He’s been very clear that Oompa Loompas are on at-will employment contracts and can leave at anytime.
November 11, 2024 at 9:31 PM