Artificial Intelligence
Google’s New AI Is Trying to Talk to Dolphins—SeriouslyIsaac Schultz | Gizmodo
“The model is DolphinGemma, a cutting-edge LLM trained to recognize, predict, and eventually generate dolphin vocalizations, in an effort to not only crack the code on how the cetaceans communicate with each other—but also how we might be able to communicate with them ourselves.”
Artificial Intelligence
Microsoft Researchers Say They’ve Developed a Hyper-Efficient AI Model That Can Run on CPUsKyle Wiggers | TechCrunch
“Microsoft researchers claim they’ve developed the largest-scale 1-bit AI model, also known as a ‘bitnet,’ to date. Called BitNet b1.58 2B4T, it’s openly available under an MIT license and can run on CPUs, including Apple’s M2.”
Artificial Intelligence
To Make Language Models Work Better, Researchers Sidestep LanguageAnil Ananthaswamy | Quanta Magazine
“We insist that large language models repeatedly translate their mathematical processes into words. There may be a better way. …In [two recent papers], researchers introduce deep neural networks that allow language models to continue thinking in mathematical spaces before producing any text. While still fairly basic, these models are more efficient and reason better than their standard alternatives.”
Future
Airbus Is Working on a Superconducting Electric AircraftGlenn Zorpette | IEEE Spectrum
“Glenn Llewellyn, Airbus’s vice president in charge of the ZEROe program, described the project in detail, indicating an effort of breathtaking technological ambition. The envisioned aircraft would seat at least 100 people and have a range of 1,000 nautical miles (1,850 kilometers). It would be powered by four fuel-cell ‘engines’ (two on each wing), each with a power output of 2 megawatts.”
Space
Skepticism Greets Claims of a Possible Biosignature on a Distant WorldJohn Timmer | Ars Technica
“So why are many astronomers unconvinced? To be compelling, a biosignature from an exoplanet has to clear several hurdles that can be broken down into three key questions: Is the planet what we think it is? Is the signal real? Are there other ways to produce that signal? At present, none of those questions can be answered with a definitive yes.”
Energy
Scientists Made a Stretchable Lithium Battery You Can Bend, Cut, or StabJacek Krywko | Ars Technica
“It’s hard to use [standard lithium-ion batteries] in soft robots or wearables, so a team of scientists at the University California, Berkeley built a flexible, non-toxic, jelly-like battery that could survive bending, twisting, and even cutting with a razor.”
Energy
These Four Charts Sum Up the State of AI and EnergyCasey Crownhart | MIT Technology Review
“Sure, you’ve probably read that AI will drive an increase in electricity demand. But how that fits into the context of the current and future grid can feel less clear from the headlines. …A new report from the International Energy Agency digs into the details of energy and AI, and I think it’s worth looking at some of the data to help clear things up.”
Future
What ‘Ex Machina’ Got Right (and Wrong) About AI, 10 Years Later Joe Berkowitz | Fast Company
“‘One day AI’s are gonna look back on us the way we look at fossils and skeletons in the plains of Africa,’ Bateman says at one point. ‘An upright ape living in dust, with crude language and tools, all set for extinction.’ …Has humanity officially entered its extinction era in the decade since Ex Machina won a Best Visual Effects Oscar and a Best Screenplay nomination for Garland?”
Space
Looking at the Universe’s Dark Ages From the Far Side of the MoonPaul Sutter | Ars Technica
“It will take humanity several generations, if not more, to develop the capabilities needed to finally build far-side observatories. But it will be worth it, as those facilities will open up the unseen Universe for our hungry eyes, allowing us to pierce the ancient fog of our Universe’s past, revealing the machinations of hydrogen in the dark ages, the birth of the first stars, and the emergence of the first galaxies.”
Artificial Intelligence
Researchers Claim Breakthrough in Fight Against AI’s Frustrating Security HoleBenj Edwards | Ars Technica
“In the AI world, a vulnerability called a ‘prompt injection’ has haunted developers since chatbots went mainstream in 2022. Despite numerous attempts to solve this fundamental vulnerability—the digital equivalent of whispering secret instructions to override a system’s intended behavior—no one has found a reliable solution. Until now, perhaps.”
Energy
Cosmic Robotics’ Robots Could Speed Up Solar Panel DeploymentsTim De Chant | TechCrunch
“Cosmic’s robot can place a panel within a few millimeters of where it needs to be. Workers spot the robot, ensuring everything looks right before fastening the panel to the rack. The goal is not just to lighten the load, but to speed things along, too. Emerick said that Cosmic’s robot could allow a standard crew to be split in two, doubling the amount of solar panels that can be installed in one day.”
Biotechnology
Jurassic Patent: How Colossal Biosciences Is Attempting to Own the ‘Woolly Mammoth’Antonio Regalado | MIT Technology Review
“Colossal Biosciences not only wants to bring back the woolly mammoth—it wants to patent it, too. MIT Technology Review has learned the Texas startup is seeking a patent that would give it exclusive legal rights to create and sell gene-edited elephants containing ancient mammoth DNA.”
The post This Week’s Awesome Tech Stories From Around the Web (Through April 19) appeared first on SingularityHub.
* This article was originally published at Singularity Hub
0 Comments