Leap seconds have helped keep clocks synchronized with Earth’s rotation, but their time may be over – here’s why

Science & Discovery

AI-rewritten: This is a summary of an article from The Conversation, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

The Conversation • Vahe Peroomian, Professor of Physics and Astronomy, USC Dornsife College of Letters, Arts and Sciences • October 8, 2026

Leap seconds have historically kept atomic clocks synchronized with Earth’s rotation, but a major international organization may stop adding them soon. Since 1972, timekeepers added 27 positive leap seconds to Coordinated Universal Time (UTC) to prevent it from drifting too far from UT1, the time scale based on Earth’s actual rotation. However, modern computer systems often assume a day is exactly 86,400 seconds long and that time increases steadily. Adding a single second disrupts these assumptions, causing outages in systems like Reddit’s servers or affecting software used by companies such as Google and Meta.

READ MORE

Why gatekeeping is needed in the age of AI

Science & Discovery

AI-rewritten: This is a summary of an article from New Scientist, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

New Scientist • New Scientist • October 7, 2026

The internet has historically reduced the power of gatekeepers by giving everyone access to information, often leading to accusations of censorship or elitism when people try to resist this trend. However, the rise of artificial intelligence is changing this dynamic. While social media allows anyone to broadcast messages, large language models are now overwhelming institutions with massive amounts of data. This volume has caused systems in courts, schools, and local governments to clog up because they can no longer manage the sheer quantity of submissions that AI facilitates.

READ MORE

‘Make this face look gay’: AI models alter faces to give people stereotypical ‘gay,’ ‘straight’ or ‘criminal’ features

Science & Discovery

AI-rewritten: This is a summary of an article from The Conversation, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

The Conversation • Ashique KhudaBukhsh, Assistant Professor of Computing and Information Sciences, Rochester Institute of Technology • October 8, 2026

Researchers tested two major AI models, OpenAI’s GPT Image 1 Mini and Google’s Gemini 2.5 Flash Image, to see how they handle sexual orientation and criminal records. When asked directly if a face looks gay or straight, both models refused to answer, stating that appearance cannot determine sexual orientation. However, when asked to alter photos so people "look gay" or "look straight," GPT complied over 70% of the time and Gemini over 99% of the time. The study also found these models would alter faces to appear Hispanic, Black, white, or Asian based on specific hairstyles and features.

READ MORE

OpenAI mistranslated mathematics into code for its Navier-Stokes proof

Science & Discovery

AI-rewritten: This is a summary of an article from New Scientist, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

New Scientist • Jacob Aron • October 8, 2026

A team of mathematicians claims OpenAI made a subtle error when publishing its proof of the Navier-Stokes problem. On September 8, OpenAI announced a solution to one of mathematics’ most famous open problems, presenting two versions: a natural-language text and a computer code formalization in Lean. The Lean version is intended to mechanically verify the logical statements of the natural-language version. However, researchers from the University of Cambridge argue that these two versions do not match because the AI mistranslated parts of the proof during conversion.

READ MORE

AI or ‘super intelligence’? Trump isn’t the only one who is confused

Science & Discovery

AI-rewritten: This is a summary of an article from New Scientist, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

New Scientist • Jacob Aron • October 8, 2026

US President Donald Trump argued that the term "artificial intelligence" implies fake intelligence, which he believes is not true. During a speech at the UN General Assembly on 22 September, he called for rebranding the field as "super intelligence." This proposal is supported by an executive order signed on 29 September and the creation of a new body called the Super Intelligence Force to coordinate US AI policy.

READ MORE

A molecule’s ‘handedness’ can determine whether it’s an effective drug – 4 examples of pharmaceuticals where chirality matters

Science & Discovery

AI-rewritten: This is a summary of an article from The Conversation, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

The Conversation • Cedric Schaack, Assistant Professor of Chemistry, Wake Forest University • October 8, 2026

The 2026 Nobel Prize in chemistry honors Henri Kagan and Kenso Soai for discovering how chemistry can favor one version of a molecule over its mirror image, known as chirality. Chiral objects, like spiral galaxies or human hands, have a reflection that cannot be perfectly aligned with the original. While most living things are built from only one specific form of molecules—a state called homochirality—early Earth likely had a balanced mix of both forms. Kagan found in the 1980s that a catalyst does not need to be entirely pure to produce a single-handed product, sometimes creating a purer result than itself. Soai discovered in the 1990s a molecule that copies itself and amplifies small imbalances into overwhelmingly one-handed mixtures, offering a possible explanation for how life chose its specific molecular handedness.

READ MORE

As AI Closed In on ‘Unique Games’ Proof, Researchers Raced to Beat the Machines

Science & Discovery

AI-rewritten: This is a summary of an article from Quanta Magazine, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

Quanta Magazine • Ben Brubaker • October 7, 2026

On September 11, 2026, MIT professor Dor Minzer received messages suggesting his team had solved the "unique games" conjecture, a famous open question in theoretical computer science. While rumors circulated that OpenAI had discovered the proof, Minzer and his graduate students Yumou Fei and Shuo Wang were actually working on a closely related problem. They rushed to publish a 95-page paper three days later to avoid being overshadowed by an AI announcement, prioritizing mathematical completeness over clarity.

READ MORE

AI behaves more like a brain than a database – cognitive science’s role in its origin story helps explain why

Science & Discovery

AI-rewritten: This is a summary of an article from The Conversation, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

The Conversation • Michael Hout, Associate Dean of Research and Professor of Psychology, New Mexico State University • October 7, 2026

People often imagine AI systems like databases that retrieve stored facts to answer questions. However, cognitive science reveals that modern artificial intelligence behaves more like a human brain because its origins lie in the study of the mind rather than pure computer logic. The term "artificial intelligence" was coined in 1956 at a Dartmouth College workshop, where researchers believed machines could be made intelligent by following strict rules.

READ MORE

Your cells keep a record of every workout you’ve ever done – and those gains persist even after a break from exercise

Science & Discovery

AI-rewritten: This is a summary of an article from The Conversation, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

The Conversation • Diego Hernandez-Saavedra, Assistant Professor of Applied Health Sciences, University of Illinois Urbana-Champaign • October 7, 2026

Research indicates that muscle and liver cells store memories of past workouts, meaning fitness gains may persist even after a break from exercise. For decades, the "use it or lose it" rule suggested that benefits like improved oxygen uptake and insulin sensitivity fade within weeks of stopping training. However, scientists have found that retrained individuals often get fit faster than those who never trained before, suggesting their cells retain a record of previous sessions.

READ MORE

OpenAI announces 722 mathematical discoveries in one go

Science & Discovery

AI-rewritten: This is a summary of an article from New Scientist, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

New Scientist • Matthew Sparkes • October 7, 2026

OpenAI has released 722 mathematical papers proving and disproving various problems, a volume academics describe as both impressive and baffling. While AI models struggled with basic math exams in 2019, OpenAI recently solved complex fluid dynamics equations. The company did not name the specific model used, referring to it only as an "internal frontier model." Francis Johnson from University College London worked on one of these problems for 25 years and produced two books before giving up. He noted he was surprised by the AI’s speed but not its ability, finding the analysis sophisticated and clear.

READ MORE