A neural network that keeps seeing art where we see mundane objects

When mundane objects such as cords, keys and cloths are fed into a live webcam, a machine-learning algorithm ‘sees’ brilliant colours and images such as seascapes and flowers instead. The London-based, Turkish-born visual artist Memo Akten applies algorithms to the webcam feed as a way to reflect on the technology and, by extension, on ourselves. Each instalment in his Learning to See series features a pre-trained deep-neural network ‘trying to make sense of what it sees, in context of what it’s seen before’. In Gloomy Sunday, the algorithm draws from tens of thousands of images scraped from the Google Arts Project, an extensive collection of super-high-resolution images of notable artworks. Set to the voice of the avant-garde singer Diamanda Galás, the resulting video has unexpected pathos, prompting reflection on how our minds construct images based on prior inputs, and not on precise recreations of the outside world.

When home is two sisters, a houseful of vulnerable men, and a lot of tough love

After she was widowed at age 32, Grażyna Sochacka founded the Panakeja Foundation – a social-assistance centre for homeless men on Sobieszewo Island in Gdańsk, Poland. Alongside her sister Wioletta Sienkiewicz, Sochacka has dedicated her life to caring for men living on the fringes of society, and often in need of intensive care due to alcoholism and other health problems. The Polish filmmaker Filip Jacobson’s observational short Home (Dom) traces the unending daily pressures the sisters face running the centre. These include changing bed sheets, providing medical assistance, keeping up with bills, and the ever-important business of doling out cigarettes. Imbued with a deep humanity and inflections of humour, the film explores the human need for a balance between structure, freedom and respect – as well as, from time to time, heavy doses of tough love.

‘I want to take the bullet and save my friends’ – the grim reality of safety drills in US schools

A generation ago, children in classrooms in the United States prepared for natural disasters such as fires and tornadoes. Today, active-shooter drills force them to confront the grim possibility that someone – perhaps a fellow student – might open fire in their school. In this StoryCorps animation, one such drill prompts a mother and her 10-year-old son in Texas to discuss a question no child should ever have to consider – whether he would sacrifice himself to try to save his schoolmates. An affecting and troubling short, The Drill gives an aching human voice to the psychological toll of school shootings and the culture of fear they’ve created for schoolchildren and their parents in the US.

If you think that modern cosmology leaves no room for ‘god’, start using your imagination

‘We’re not so much abandoning the idea of the gods, we’re just trying to pull them all the way into the Universe.’

From the possibility of infinite universes to the prospect of panpsychism, puzzles have arisen in physics that can take science to some very counterintuitive places. According to Mary-Jane Rubenstein, assistant professor of religion and feminist, gender and sexuality studies at Wesleyan University in Connecticut, new theories and breakthroughs at the forefront of cosmology need not – and moreover, should not – elbow out theology from the conversation about our place in the cosmos. Instead, as she argues in this wide-ranging interview recorded at the HowTheLightGetsIn Festival from the Institute of Arts and Ideas in 2019, science should encourage us to build more durable myths and theologies to suit our times.

What is it to be Bayesian? The (pretty simple) math modelling behind a Big Data buzzword

If you’ve ever tripped up over the term ‘Bayesian’ while reading up on data or tech, fear not. Strip away the jargon and notation, and even the mathematics-averse can make sense of the simple yet revolutionary concept at the core of both machine learning and behavioural economics. As this video from the YouTube channel 3Blue1Brown skilfully explains, at its most basic, Bayes’s theorem is a tool for assessing degrees of probability based on prior conditions. And there are ways to make it altogether more intuitive than the statistical formulas might suggest. Although the theorem dates back to its 18th-century namesake, the English statistician and philosopher Thomas Bayes, it has gained increasing relevance in the Big Data revolution.