Category: Uncategorized
Twitter Can Now Censor the President
Last week Twitter announced that it would no longer exempt most tweets by government officials from its rules regarding what…
Towards Data Science: Press Coverage of the Early 2020 Primary
A look at media coverage of the 2020 race to date. Read The Full Article.
FiveThirtyEight: Media Outlets Paid Way More Attention To Kamala Harris After The Debate
FiveThirtyEight's latest weekly roundup of the 2020 Democratic race. Read The Full Article.
A Reminder That Our Data-Driven Results Depend On What Data We Look At
The fact that the same analysis can yield wildly different findings simply by changing which of the supposedly interchangeable datasets…
CJR: MSNBC Public Editor: In (and Out) Of The Tank For Joe Biden
Columbia Journalism Review's Public Editor for MSNBC Maria Bustillos uses the TV Explorer to report on MSNBC's coverage of the…
As The World Becomes More Data-Driven Humans Are Overruling Our Algorithms
Companies are increasingly manually adjusting their algorithms either to address certain kinds of bias or to make outputs more in…
Could Buying A Driverless Car In The Future Mean Selecting Its Ethical Values Too?
As fully autonomous driverless cars eventually reach our roadways, could future car buyers have to weigh the ethical design and…
Algorithms And Data Construct 'Truth,' Not Discover It
The data-driven revolution is prefaced upon the idea that data and algorithms can lead companies away from biased human judgement…
The Mythical False Narrative Of Data-Driven Analysis
As companies confront biased or uncomfortable outcomes from their data-driven analyses, they increasingly seek to manually intervene, overriding their data-driven…
Dynamic Sampling Of Video To Imagery For Deep Learning
Dynamic sampling can help optimize the conversion of videos into sampled still image sequences for optimized analysis and even non-consumptive…
Using Machine Learning To Move Fact Checking From Articles To Claims
Could natural language processing help transform the way in which fact checking results are applied, flagging the individual sentences within…
Is There Such A Thing As Objective Truth In Data Or Is It All In The Eye Of The Beholder?
As data science increasingly reinserts human judgement over algorithmic purity and as the digital ecosystem reverses from its brief objective…
FiveThirtyEight: Which Democratic Presidential Candidate Was Mentioned Most In The News Last Week?
FiveThirtyEight's latest weekly roundup of the 2020 Democratic race. Read The Full Article.
For Mobile Users Ad-Supported 'Free' Sites Cost Them Real Money
We think of today’s ad-supported Web as “free” in which we barter our privacy in exchange for access to valuable…
What If Web Browsers Told Us How Much The 'Free' Web Was Costing Us?
What might it look like if our browsers and mobile devices actually tallied the privacy and monetary cost of online…
Why Don't We Have Online Fitness Trackers For Our Web Usage?
Would the equivalent of online fitness trackers help reshape how we think about the “free” Web? Read The Full Article.
We Must Recognize Just How Brittle And Unpredictable Today's Correlative Deep Learning AI Is
For AI to succeed companies must learn to distinguish hype from reality and understand the ways in which AI’s brittleness…
As AI Comes For Management Perhaps We Should Look Forward To Machines Taking Our Jobs
From hiring to performance reviews to real-time monitoring, the data-driven workplace is here to stay, driven in large part by…
Using AI To Analyze Video As Imagery: The Impact Of Sampling Rate
Sampling video into sequences of still imagery to analyze using deep learning can save storage and computational costs and creates…
Will Hypersonic Weapons Finally Push Us Towards AI-Powered Missile Defense?
The rise of hypersonic weapons, with just minutes between launch and impact, may be the development that pushes countries to…
Automated Speech Recognition Yields More Consistent Results Than Human Captioning
Instead of comparing machines against an idealized benchmark that simply doesn’t exist, we should recognize that machines are increasingly reaching…
We Must Stop Comparing Deep Learning's Real Accuracy To Nonexistent Human Perfection
If we really wish to understand the usability of AI today we should be comparing it against the human workflows…
Could Automated Claim Extraction And Tracking Combat Digital Falsehoods?
One of the greatest challenges facing fact checkers today is that humans are simply no match for real-time tsunami of…
Combatting Digital Falsehoods Through Coverage Maps
An intriguing approach to combating digital falsehoods is the creation of coverage maps that capture where a claim first appeared,…