Everyone knows those first, dreadful signals that mean illness is coming. For me, it's the faintest tickle in the throat that lingers a moment too long. For my wife, it's a soreness in her legs. Whatever it is, we've all been there, and usually, by then it’s too late to do anything about it.

But what if we could sense illness even earlier, before those first warning signs? Researchers at McGill University have cracked open this possibility using wearable biosensors and AI specifically trained to recognize subtle signals our bodies emit way before we notice them.

The team monitored tiny physiological changes, collecting more than 2 billion personalized data points to train their model. Using this massive training set, they they built an AI system that can detect systemic inflammation, a key indicator of infection, with nearly 90% accuracy before clinical symptoms even appeared.

With earlier detection, we unlock earlier interventions: antivirals administered sooner, isolation protocols triggered faster, outbreaks prevented, and hospitalizations drastically reduced.

And it doesn't just mean less time in the waiting room. It means less time wondering what surprises your toddler brought home from day care. Less time doomscrolling symptoms. Less time worrying about illness at all. It means more time for the thing your body needs most: rest.

This could also signal a profound shift in healthcare. Moving from a reactive care system to proactive model. Building a future where strained healthcare systems get a breather, and so do we.

Think of the possibilities: your wearable quietly notices something amiss. It alerts your healthcare provider, who runs a quick diagnostic and prescribes the necessary antivirals immediately. And you haven’t even had a sniffle.

Goodbye flu season. Hello future.

Most people don't know what it's like to die five times before turning thirty. Dr. David Fajgenbaum isn't most people. Since being diagnosed with Castleman disease, a rare and often fatal condition that sits somewhere between cancer and autoimmune disorder, Dr Fajgenbaum has been racking up near-death experiences.

Fed up with his precarious relationship to life, he turned detective, poring over his own medical records until he found an unexpected answer: a decades-old transplant drug that put his disease into remission.

That discovery haunted him with a question: How many other cures are hiding in plain sight?

The answer is staggering. Of 7,000+ rare diseases, fewer than 400 have treatments. Meanwhile, 300 million people suffer while potential cures sit unused on pharmacy shelves.

Through his nonprofit Every Cure, Fajgenbaum built an AI system that analyzes every known drug against every known disease simultaneously. The results? Lives saved. When Joseph Coates was days from hospice with a rare blood disorder, the AI suggested an untested drug combination. Within a week, he stabilized. Today, he's in remission.

In 2024 U.S. government awarded Every Cure the largest investment ever in the drug repurposing world: $48.3 million.The AI has already identified potential treatments for sickle cell disease, ALS, and autism.

Sometimes the most powerful breakthroughs come not from creating something new, but from seeing what was always there, waiting patiently on the shelf.

For decades, we've trusted satellites floating in space to guide us safely home. GPS tech has helped us get places quicker, avoid traffic, and most importantly, keep us from getting lost on unlit backcountry roads on the way to the cottage. But what if it fails? What if we’re suddenly flying blind?

With growing threats to the reliability of GPS (signal jamming, spoofing, degradation), researchers have been working on a new way to find our way home. It turns out the Earth itself has been offering to help navigate all along.

Instead of looking up to satellites, researchers looked down to something that's been guiding us for millennia: Earth's magnetic field. It turns out, it’s more than just helping us find the North Pole. The earth’s magnetic field provides invisible signals left by ancient geology, mineral deposits, and tectonic shifts. This means that every location on Earth has its own magnetic personality, as distinct as our fingerprints.

Using quantum magnetometers, scientists can now read these planetary fingerprints with unprecedented precision. AI algorithms are essential to filter out interference from aircraft systems and vehicles and match the clean magnetic readings against detailed global maps. This marriage of quantum technology and AI brings us navigation that's literally grounded in Earth itself, refined by machine intelligence.

And it’s already being put to the test. Airbus subsidiary Acubed and quantum computing pioneer SandboxAQ have already logged over 150 flight hours with their MagNav technology. Astoundingly, their quantum navigation achieved FAA-level precision without ever reaching out to space.

This technology is new, but it opens up exciting new possibilities. Imagine a world where navigation can't be hacked, jammed, or spoofed. Where every flight, every autonomous car, every rescue mission has an unshakeable connection to Earth itself. This is the future of human mobility.

Since the dawn of time the story of human progress has been measured by the materials we’ve been able to discover and harness. From the earliest cultures shaping entire civilizations with stone, bronze, and iron to the steel and glass skeletons of our modern cities, new materials have unlocked exciting new possibilities. But the pace and nature of that discovery are undergoing a transformation more profound than any that has come before, thanks to the twin revolutions of quantum physics and artificial intelligence. Quantum computers and machine learning join forces to design materials that have never existed. By modelling atomic behaviour at unprecedented scale, researchers are engineering tomorrow's superconductors, batteries, and space-age alloys at light speed. We are standing at the precipice a dramatic shift. Researchers at the USC Viterbi School of Engineering have announced a new AI model called Allegro-FM, shattering previous limitations.

Allegro-FM moves design and testing of new materials to the digital realm. Without the need of physical lab space and equipment, these virtual labs can rapidly sift through vast libraries of potential materials and predict their properties with quantum accuracy. Underpinning this model is an innovative approach to “interatomic potentials”, the forces that govern how atoms interact. Before integrating machine learning, calculating these forces required complex quantum mechanics formulas with separate, computationally intensive simulations needed for each element. Enter AI.

By leveraging its machine learning model, Allegro-FM can generate comprehensive training sets that allows it to predict the interaction functions of nearly the entire periodic table at once. The ability to generalize across the different elements dramatically reduces computational overhead and is already catalyzing innovation across a spectrum of industries.

In energy, this is being used to engineer next-gen batteries and create new materials capable of storing energy at a grid scale. It’s also improving carbon capture methods. Imagine a future where the concrete in sidewalks and the asphalt in roads actively absorb CO2 from the atmosphere!

This is just the beginning. This tech could give us materials that perfectly mimic the properties found in nature. Photosynthetic surfaces that absorb sunlight and convert it into energy just like a plant. Building materials that mimic termite mounds to passively regulate temperature and airflow, reducing our dependency on energy intensive HVAC systems. Metals that bleed and scab to quickly fix fractures on spacecrafts or submersibles. Polymers with a nervous system to detect stress or chemical degradation. Medical nanobots made from custom designed proteins to hunt for cancer cells, deliver medicine on a cellular level, or perform microscopic repairs within an individual cell.

The true paradigm shift with Allegro-FM is the move from discovering materials to designing them from first principles. We are on the cusp of tailoring the building blocks of our world to specifically meet our most pressing challenges.

What if the best tool for protecting your brain is already in your kitchen? Using AI, researchers from Fudan University analyzed the diets of 185,000 people and created a system that translated complex biochemistry into everyday choices.

It’s called MODERN. This model sifted through decades of data to find the subtle dietary patterns that lowered dementia risk. It wasn't about discovering a single miracle food, but about identifying the powerful combinations of things like leafy greens, berries, and balanced eating habits.

For people tired of conflicting and confusing dietary advice, MODERN provided a clear scoring system to guide their daily choices. The results speak for themselves. People who followed the AI-guided diet reduced their dementia risk by 36%, better than established brain-healthy diets that took years to develop.

Imagine a small robot perched on a child's shoulder with tiny animal ears and a gentle dome-shaped head. When the child asks a question, the robot responds not in English or French, but in fluent Anishinaabemowin, an Indigenous language spoken by just 36,000 members of the Anishinaabe nation of North America.

This is Skobot, the creation of 24-year-old Danielle Boyer, an Anishinaabe roboticist who grew up below the poverty line in Michigan. Inspired by talking Elmo toys, Boyer designed these colourful companions to help preserve endangered Indigenous languages through play. Each robot is motion-activated and uses children's voices to create conversations in languages that were once forbidden, now finding new life through silicon and code.

Around the world, 40% of humanity's 6,700 languages are at risk of extinction, with many spoken by Indigenous communities. But a quiet revolution is underway. Indigenous researchers are crafting AI tools to preserve the voices of the tradition-keeping elders and amplify their voices across generations. In Quebec, the FLAIR initiative developed portable hardware called "language in a box.” Computer scientist Jacqueline Brixey created Masheli, a chatbot that speaks Choctaw through animal stories.

These aren't typical tech solutions imposed from outside. They're created by community members who understand that "our languages are living things," as Boyer puts it. "Our languages need to have community relationships. Our languages need to be learned from people. It can't be pure technology."

It’s about more than just preserving languages. Research shows that connection with linguistic heritage corresponds to lower teen suicide rates and better physical health. When AI helps a Choctaw child hear ancestral stories or enables an Anishinaabe teenager to practice traditional phrases, it's nurturing cultural wounds and healing generational trauma.

In a world where artificial intelligence often feels like it's erasing human diversity, here's technology doing the opposite: ensuring that every voice echoes into the future.