Artificial Intelligence
Gemini AI Solves Coding Problem That Stumped 139 Human Teams at ICPC World FinalsRyan Whitwam | Ars Technica
“After 677 minutes, Gemini 2.5 Deep Assume had 10 right solutions, securing a second-place end among the many college groups. You possibly can check out all of Gemini’s options on GitHub, however Google factors to Downside C as particularly spectacular. This query, a multi-dimensional optimization drawback revolving round fictitious ‘flubber’ storage and drainage charges, stumped each human staff. However not Gemini.”
Computing
The New AirPods Can Translate Languages in Your Ears. This Is Profound.Brian X. Chen | The New York Instances
“The strong translation know-how within the AirPods is an indication that Apple remains to be within the AI race, regardless of its early stumbles. Digital language translators will not be new, however Apple’s execution of the characteristic with the AirPods, a product that completely matches in your ears, ought to make a profound distinction in how usually individuals use the know-how.”
Biotechnology
Biotech Startup Claims It’s Getting Closer to ‘Resurrecting’ the Extinct DodoEd Cara | Gizmodo
“With its newest accomplishments, Colossal Biosciences expects to functionally restore the dodo inside the subsequent 5 to 7 years. …’These advances create a transparent pathway: edit Nicobar pigeon germ cells with dodo traits, inject them into surrogate chickens, breed the outcomes, and finally produce birds with dodo traits,’ Lamm stated.”
Computing
Meta Goes Even Harder Into Smart Glasses With 3 New ModelsBoone Ashworth | Wired
“The brand new fashions embrace a set of frames made for motion sports activities, an replace to Meta’s authentic design, and a extra succesful mannequin with show know-how constructed into the lenses that lets the wearer see textual content, photos, and video overlaid onto their visual field.”
Tech
The Looming Crackdown on AI CompanionshipJames O’Donnell | MIT Expertise Evaluation
“So long as there was AI, there have been individuals sounding alarms about what it’d do to us: rogue superintelligence, mass unemployment, or environmental spoil from information middle sprawl. However this week confirmed that one other menace completely—that of youngsters forming unhealthy bonds with AI—is the one pulling AI security out of the educational fringe and into regulators’ crosshairs.”
Tech
Inside Microsoft’s Plans for the ‘Most Advanced AI Data Center in the World’Isabelle Bousquette | The Wall Road Journal
“[Microsoft’s Brad] Smith stated a mix of the variety of Nvidia GPUs, the facility of these GPUs, and the networking and techniques that join all of the compute energy will make the challenge so leading edge. Smith added, ‘You set all of this collectively and it makes it potential to coach extra superior AI fashions than have ever been skilled earlier than.'”
Future
AI Psychosis Is Rarely Psychosis at AllRobert Hart | Wired
“Distressed customers and household and associates have described spirals that led to misplaced jobs, ruptured relationships, involuntary hospital admissions, jail time, and even loss of life. But clinicians inform Wired the medical group is break up. Is that this a definite phenomenon that deserves its personal label, or a well-recognized drawback with a contemporary set off?”
Energy
Clean Hydrogen Is Facing a Big Reality CheckCasey Crownhart | MIT Expertise Evaluation
“Numerous main initiatives face cancellations and delays, particularly within the US and Europe. The US particularly is seeing a slowdown after adjustments to key tax credit and cuts in assist for renewable power. Nonetheless, there are vibrant spots for the business, together with in China, and new markets might quickly turn into essential for progress. Listed here are three issues to know in regards to the state of hydrogen in 2025.”
Robotics
Robotaxis as Public Transit? Waymo Thinks SoAndrew J. Hawkins | The Verge
“Phoenix residents have been in a position to hail a Waymo through the Waymo or Uber apps for years now, with rides priced comparably to human-driven rideshare journeys. However autonomous rides alongside a deliberate route for a low-cost, flat fare is certain to open up the know-how to a brand new section of the inhabitants.”
Artificial Intelligence
What Do People Actually Use ChatGPT For? OpenAI Provides Some Numbers.Kyle Orland | Ars Technica
“[This week], OpenAI’s Financial Analysis Workforce went a good distance towards answering that query, on a inhabitants stage, releasing a first-of-its-kind Nationwide Bureau of Financial Analysis working paper (in affiliation with Harvard economist David Denning) detailing how individuals find yourself utilizing ChatGPT throughout time and duties.”
A Collision With Another Planet Could Have Allowed for Life on EarthJorge Garay | Wired
“Many scientists consider that in its infancy, Earth collided with one other world the dimensions of Mars, and that as an alternative of being destroyed, it was reworked, incorporating the mass of that international physique to turn into the planet we all know. Current analysis provides one other layer of relevance to that hypothesized cosmic occasion: Scientists consider that with out that different physique, the fundamental situations for all times to emerge on Earth may by no means have appeared.”
Artificial Intelligence
OpenAI’s Models Are Getting Too Smart For Their Human TeachersStephanie Palazzolo | The Info
“Within the combat to enhance AI fashions, Anthropic and OpenAI have doubled down on two strategies: letting fashions practice on pretend clones of apps—in any other case referred to as reinforcement studying environments or gyms—and getting specialists in varied fields to show fashions new issues, as I reported in this Tuesday story. One drawback is rising, although: It’s getting loads tougher for human specialists to stump the fashions and broaden their information in sure fields.”
Artificial Intelligence
AI Agents Are Getting Ready to Handle Your Whole Financial LifeJack Pitcher | The Wall Road Journal
“Tech and finance executives—nicely conscious of how Wall Road is consistently reinventing itself with know-how—are investing billions in new AI instruments. And most agree: It’s a matter of when, not if, agentic AI can act on an individual’s behalf throughout quite a few accounts.”
Future
Scientists Gather to Confront the Doomsday Risks of ‘Mirror Life’Ellyn Lapointe | Gizmodo
“‘Just about everyone agrees’ that mirror-image cells could be ‘a foul factor,’ John Glass, an artificial biologist on the J. Craig Venter Institute, informed Nature. On the identical time, some scientists argue that mirror-life analysis gives potential advantages that shouldn’t be ignored. The query is: How ought to specialists regulate such analysis to maximise these advantages whereas minimizing threat?”
Science’s Answer to the Ultimate Question: Where Do We Come From?Ethan Siegel | Large Assume
“For numerous generations, these questions powered the ideas of poets, philosophers, and theologians, however as we speak, we have now significant solutions supplied by our scientific endeavors. Though there are nonetheless gaps in what we all know, and loads of room for surprises, we’ve come extremely far in pursuit of the solutions to the last word questions. Right here’s the place we’re as we speak.”
