-
1 week ago | 
technewstube.com | Stephen Ornes
Tech News Tube is a real time news feed of the latest technology news headlines.Follow all of the top tech sites in one place, on the web or your mobile device.
-
1 week ago | 
wired.com | Stephen Ornes
The original version of this story appeared in Quanta Magazine. Large language models work well because they’re so large. The latest models from OpenAI, Meta, and DeepSeek use hundreds of billions of “parameters”—the adjustable knobs that determine connections among data and get tweaked during the training process. With more parameters, the models are better able to identify patterns and connections, which in turn makes them more powerful and accurate. But this power comes at a cost.
-
1 week ago | 
flipboard.com | Stephen Ornes
I paid $20 for an AI marketing agent - and here's what it got meThink AI agents are a shortcut? Here's where the real magic comes from. I'm not one to believe everything I see online, but I've been seeing a lot of …
-
3 weeks ago | 
snexplores.org | Carolyn Wilke |Stephen Ornes |Lisa Grossman |Sarah Zielinski
Smartphones are helping chart an erratic region of Earth’s atmosphere that can mess with navigation systems and other tech. That region is the ionosphere — a sea of charged particles on the edge of outer space. There, the sun’s radiation cooks gases in the air. This rips electrons off neutral atoms and molecules, creating free-roaming charged particles. Levels of these charged particles shift with changing sunlight and solar activity.
-
4 weeks ago | 
snexplores.org | Aaron Tremper |Nikk Ogasa |Stephen Ornes |Rachel Ehrenberg
Visit your local natural history museum gift shop and you’ll probably find a container of dinosaur toys for sale. Dump out its contents, though, and you’ll likely find some imposters, says Nathan Smith. He should know. He’s director and curator of the Dinosaur Institute in California. It’s part of the Natural History Museum of Los Angeles. Winged pterosaurs are a common fraud. So is the sail-backed carnivore, Dimetrodon. The word “dinosaur” isn’t a catch-all term for just any scaly, prehistoric giant.
-
1 month ago | 
sjawards.aaas.org | Stephen Ornes
Children’s book author Ainissa Ramirez, who trained as a materials scientist, first learned about Jim West, the co-inventor of the electret microphone, two decades ago. She couldn’t believe she didn’t already know his story. Ramirez, like West, is a Black scientist; also, like West, she spent years working at Bell Labs in Murray Hill, New Jersey. The type of microphone that West designed, with German scientist Gerhard Sessler, changed the world.
-
1 month ago | 
quantamagazine.org | Stephen Ornes
For computer scientists, solving problems is a bit like mountaineering. First they must choose a problem to solve — akin to identifying a peak to climb — and then they must develop a strategy to solve it. Classical and quantum researchers compete using different strategies, with a healthy rivalry between the two. Quantum researchers report a fast way to solve a problem — often by scaling a peak that no one thought worth climbing — then classical teams race to see if they can find a better way.
-
1 month ago | 
flipboard.com | Stephen Ornes
3 hours agoDecolonizing My Love Life: What I Learned When I Stopped Dating White Men"My 'type' wasn’t just a preference: It was an algorithm shaped by media and colonial history." I received my first lesson in desirability politics via an unexpected text.
-
1 month ago | 
cancertodaymag.org | Stephen Ornes
FIVE MILES INTO a 30-mile bike ride on a sunny Sunday morning in August 2019, Craig Schumpert ran into trouble. He hadn’t thought twice about getting on the bike for that ride. Schumpert, then 51, had previously completed long rides of 100 miles. But halfway up a steep hill in Rock Creek Park near his home in Washington, D.C., Schumpert lost his breath and had to turn around. Before then, he had noticed a cough when he accompanied his wife on long runs, but he’d always been quick to recover.
-
1 month ago | 
quantamagazine.org | Stephen Ornes
Large language models work well because they’re so large. The latest models from OpenAI, Meta and DeepSeek use hundreds of billions of “parameters” — the adjustable knobs that determine connections among data and get tweaked during the training process. With more parameters, the models are better able to identify patterns and connections, which in turn makes them more powerful and accurate. But this power comes at a cost.
