Light-powered chip makes AI 100 times more efficient. Advances at the intersection of optics and silicon have yielded a prototype photonic chip that performs convolution operations—core to image recognition and neural networks—using laser light and Fresnel lenses etched directly onto a silicon wafer. Tests demonstrated 98 percent accuracy on handwritten digit classification while slashing energy use by two orders of magnitude.
“Performing a key machine learning computation at near zero energy is a leap forward for future AI systems,” said Volker J. Sorger, study leader and Rhines Endowed Professor in Semiconductor Photonics at the University of Florida.
This breakthrough promises to reshape the sustainability and scalability of AI, offering a glimpse of a future where energy-hungry models coexist with stringent carbon goals.Can we design machines to make ethical decisions?. This 2,800-word essay by Tom Chatfield probes the trolley problem and its many variations to explore whether algorithmic systems can—or should—be tasked with life-and-death judgments. Drawing on thinkers from Paul Virilio to Judith Jarvis Thomson, it examines the philosophical underpinnings of consequentialism and the limits of automating moral choice.
“Accidents mark the spots where anticipation met reality and came off worse. Yet each is also a spark of secular revelation…”
Readers will appreciate the synthesis of ethical theory and real-world dilemmas, together with reflections on the place of human fallibility in a world leaning ever more on autonomous systems.AI Is Taking Water From the Desert. In a dispatch from Goodyear, Arizona, this narrative feature reveals the hidden hydrological cost of hyperscale data centers powering generative AI. The article describes vast facilities drawing millions of gallons annually to cool racks of GPUs, set against the backdrop of drought-prone landscapes.
“The air crackled with a latent energy, and some kind of pulsating sound was emanating from the electric wires above my head…”
Through vivid scene-setting and on-the-ground reporting, it raises urgent questions about resource scarcity, corporate responsibility, and the true price of digital convenience.‘Mistakes are romantic’: the revival of point-and-shoot cameras. This feature charts Gen Z’s embrace of analogue photography, from disposable Hello Kitty cameras to refurbished Pentax 17s. It delves into the tactile allure of film—the slow pace, the suspense of development, the serendipity of light leaks and grain.
“Even the mistakes are romantic – the light leaks on the first few frames of a new roll, red-eye and grain.”
Blending market data with cultural insight, it illuminates how a generation disillusioned with instant gratification seeks meaning in the unpredictable beauty of imperfection.Designing the Digital Self. Robert J. Kett’s 2018 essay for SFMOMA examines Ant Farm’s seminal “Beyond Things Past” slideshow as a lens on our evolving relationship with technology. Through analyses of “plugging-in” and “plugging-out,” it traces the tension between utopian promise and critical skepticism that has marked digital art since the counterculture.
“Ant Farm’s project reflects a deep ambivalence about technology prevalent in the 1960s and 1970s, particularly within the counterculture in which the group were famous participants.”
By weaving art history, social commentary, and design theory, the piece fosters a richer understanding of how digital tools shape—and are shaped by—human identity and imagination.