Scientists at the University of Maryland have developed an 'invisibility cloak' for AI using adversarial patterns on a sweater, making the wearer nearly undetectable to standard object detection methods.

Missing data is not only a challenge for data scientists. The brain also has to handle it, as much of the information needed for action is missing. One way that the brain does this is smart imputation, the mechanisms of which we explore here
scientificamerican.com/podcast/episod…

The curse of dimensionality is itself cursed (ie highly dimensional)

earthquake?

This is incredible.

From Jamie Lane where AirBnBs are fully booked next week.

For unknown reasons, chatGPT uses 'delve' 500x more than human writers.

This should be interesting. A large part of the brain is dedicated to uncertainty management. Which makes sense in a world full inherently irreducible uncertainty.

How not to get stuck in a local minimum (momentum)

And here is a free big 5 personality measure: sapa-project.org/survey/start.p…

Something I taught 11 yo: The less competent the judges in a contest, the less it matters to win it.

How did this get through peer review? Who is trolling here?

Firefighters and MBA students were asked how they'd feel if they bought a new car, showed their friend, and next day the friend buys the same kind of car

% who report they'd feel:

Upset

Firefighters: 3%
MBAs: 47%

Betrayed

Firefighters: 13%
MBAs: 47% psycnet.apa.org/record/2007-15…

Ten months ago, we launched the Vesuvius Challenge to solve the ancient problem of the Herculaneum Papyri, a library of scrolls that were flash-fried by the eruption of Mount Vesuvius in 79 AD.

Today we are overjoyed to announce that our crazy project has succeeded. After 2000…

Is the horse walking towards us or away from us? 😏

The Wired video is 1 year old today
youtube.com/watch?v=CNeaQl…

CDS MS students Tayyibah Khanam & Harsh Asrani's study, presented at the Society for Neuroscience, analyzed real-time music perception and identified 4 unique types of music listeners:

1. Drifters
2. Jumpers
3. Gymnasts
4. Sitters.

Dive in:

nyudatascience.medium.com/pioneering-the…

Missed the CDS Homecoming event last month? Well, you're in luck! We just published the recordings of the event's #research Lightning Talks, which included lectures from Pascal Wallisch on the ' #datascience of Music', Falaah Arif Khan on #responsibleAI + more! youtube.com/playlist?list=…

If it is moving fast, you can see it. Even in outer space.

Irreplacable, devastating loss

Pearson correlations are only asymptotically (i.e., as n -> infinity) unbiased estimators of the population correlation. In finite sample sizes, a Pearson correlation will tend to under-estimate the population correlation. The bias is greater when n is smaller and r is around .6