Source: wikimedia

In a video that was widely shared last Friday, a representative from Stanford Medical Center spoke to residents protesting how the hospital chose to allocate its first shipment of COVID-19 vaccines. The hospital had around 5,000 initial doses to distribute (and expects to have tens of thousands more within the next few weeks), and came up with an allocation scheme in which only 7 of the approximately 1,300 residents were on the list. Many of these residents deal directly with patients who have COVID-19, whereas other more senior physicians, as well as other front line workers, such as nurses and…


In many areas of science, there is an increasingly urgent unmet need, a role that could be simultaneously fascinating, rewarding, and potential remunerative. It is a role that already exists in various forms, but which could be made into something much more potent, especially if forces converged to make it more prominent. I am talking, of course, about the professional science critic.

In the popular imagination, science operates something like a priesthood: scientists enter elite institutions as novices and emerge years later as full-fledged representatives of The Truth. Along the way, they are trained to be experts and professionals in…


The original plate of “View from the Window at Le Gras”, a heliograph made by Nicéphore Niépce around 1827.

Although it isn’t normally thought of in these terms, taking a photograph involves recording a four-dimensional block of space-time and projecting it down to a two-dimensional representation. With sufficiently sensitive material (and a fast enough shutter), one can produce images more or less instantaneously, but longer exposures reveal the inherent temporality of this process, showing us something that is clearly based on the world, yet quite different from our experience of it. Today, the ability to create images is so commonplace, of course, that we easily take it for granted, but early commentaries on photography reveal just how extraordinary it…


Anthropometric data sheet (both sides) of Alphonse Bertillon (1853–1914). Source

For anyone who has been paying attention, it will not have gone unnoticed that the past year has seen a dramatic expansion in the use of face recognition technology, including at schools, border crossing, and interactions with the police. Most recently, Delta announced that some passengers in Atlanta will be able to check in and go through security using only their face as identification. Most news coverage of this announcement emphasized the supposed convenience, efficiency, and technical novelty, while underplaying any potential hazards. …


In a new paper to appear at the ACM conference on Fairness, Accountability, and Transparency (FAT* 2019), we present a method called deep weighted averaging classifiers (DWACs), which can transform any deep learning architecture into one that is more transparent, interpretable, and robust to out-of-domain data.

Although deep learning models differ in their details, the vast majority of deep architectures for classification include a final softmax layer, which projects a hidden representation into a vector of probabilities. That is, the output is computed as:

where h=g(x) is the output vector from the preceding layers, such as a convolutional or recurrent…


Framing Internet Safety, by Nathan W. Fisk, is a fascinating study of the surveillance and governance of youth online.

According to Fisk, the rhetoric of “protecting the children” has been a key part of most attempts at regulating the internet, such as the Communications Decency Act, whether targeted at offensive content, online predators, or more recent concerns around “cyberbullying”. While not completely imaginary concerns, there tends to be a rather large mismatch between the threats emphasized in the media, and the things that worry the people directly concerned.

For parents, the largest worry seems to be that their children will…


Note: this post was written for a general audience, and assumes only passing familiarity with machine learning.

For anyone who’s been paying attention, it should be apparent that statistical machine learning systems are being widely deployed for automated decision making in all kinds of areas these days, including criminal justice, medicine, education, employment, policing, and so on. (For a great overview of the hazards of applying machine learning in these domains, check out Cathy O’Neil’s book Weapons of Math Destruction).

Particularly with the recently enacted GDPR — the new European regulation about data and privacy — there is growing interest…


Although it was not the largest of its kind, or the most invasive, or even particularly surprising, the recent Cambridge Analytica scandal produced a surprisingly large amount of outrage and commentary. If nothing else, it was yet another reminder that we have gradually slipped into a regime where certain aspects of our privacy that could once be taken for granted are now long gone. Are people concerned? Is this something we should be worried about? What exactly are the harms that come from this sort of loss of privacy.

Part of the difficulty with trying to answer these sorts of…


TLDR; Achieving calibration at the group level is easy. This does not necessarily imply, however, that any group is homogeneous, or that group-level probabilities apply to the individual in each group.

1. Predicting recidivism

By now, surely everyone who is interested in fairness and transparency in machine learning is familiar with the ProPublica story about algorithms used in criminal justice. To summarize very briefly: there is a widely-used tool called COMPAS that is used to score people who have been convicted of a crime in terms of their risk of recidivism. By pulling the records of around 10,000 people (before filtering), ProPublica found…


I was recently reminded of Christian Marclay’s The Clock — a 24 hour film made up of thousands of short film clips, carefully edited together such that (when properly synchronized) the time depicted in each moment of film corresponds to the current time in the world. I’ve never seen it, and sadly only a few copies exist, but I think it’s a fascinating example of a kind of archival art.

When you hear the description, you might think of it as just a kind of a gimmick, one which would not reward actual viewing. However, based on the (bootlegged) clips…

Dallas Card

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store