todd gureckis
@toddgureckis.bsky.social
computational cognitive science @ nyu. director NYU minds, brains, and machines initiative. https://gureckislab.org. Are you interested in research in my lab? https://intake.gureckislab.org/interest/
def able to conjure educational interfaces of my dreams
February 6, 2026 at 7:04 AM
def able to conjure educational interfaces of my dreams
effect size feels somewhat academic when you can readily convert even small behavioral effects into substantial $.
January 13, 2026 at 5:05 PM
effect size feels somewhat academic when you can readily convert even small behavioral effects into substantial $.
nice post. i don't know this data is public, or it's actual assessment, but my hunch is the effectiveness of optimized social media application algorithms for retaining human attention is 5 sigma on a chronological, friend only feeds of the early 2000s.
January 13, 2026 at 4:29 PM
nice post. i don't know this data is public, or it's actual assessment, but my hunch is the effectiveness of optimized social media application algorithms for retaining human attention is 5 sigma on a chronological, friend only feeds of the early 2000s.
yeah, i've been white hatting some of this and i'm a little surprised at how bad most of these things are (for now), esp for multi-modal real-time stuff.
December 16, 2025 at 5:26 AM
yeah, i've been white hatting some of this and i'm a little surprised at how bad most of these things are (for now), esp for multi-modal real-time stuff.
e.g., bsky.app/profile/josh...
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
December 16, 2025 at 5:18 AM
e.g., bsky.app/profile/josh...
nice! i think people should talk about the good data experiences like this, because i have run across reviewers and people on grant panels and stuff who are now completely spooked on online data quality to the point of basically flat reject it's validity, despite maybe a more nuanced set of issues
December 16, 2025 at 5:15 AM
nice! i think people should talk about the good data experiences like this, because i have run across reviewers and people on grant panels and stuff who are now completely spooked on online data quality to the point of basically flat reject it's validity, despite maybe a more nuanced set of issues
I'm exploring this issue a little and interested in where the failure points are. Please DM me.
December 15, 2025 at 9:30 PM
I'm exploring this issue a little and interested in where the failure points are. Please DM me.
it seems like not to many people that follow me are using it but here's a good summary of the upsides: towardsdatascience.com/why-im-makin...
Why I’m Making the Switch to marimo Notebooks | Towards Data Science
A fresh way to think about computational notebooks
towardsdatascience.com
December 2, 2025 at 3:11 PM
it seems like not to many people that follow me are using it but here's a good summary of the upsides: towardsdatascience.com/why-im-makin...
the winner is columbia which had a box that looked like the pdf upload link but was actually a request to upload your business card (if appropriate). good captcha!
December 2, 2025 at 2:37 AM
the winner is columbia which had a box that looked like the pdf upload link but was actually a request to upload your business card (if appropriate). good captcha!