AI Can Now Create a Replica of Your Personality
James O’Donnell | MIT Technology Review
“Imagine sitting down with an AI model for a spoken two-hour interview. A friendly voice guides you thru a conversation that ranges out of your childhood, your formative memories, and your profession to your thoughts on immigration policy. Not long after, a virtual replica of you is in a position to embody your values and preferences with stunning accuracy. That’s now possible, in line with a brand new paper from a team including researchers from Stanford and Google DeepMind, which has been published on arXiv and has not yet been peer-reviewed.”
This AI Taught Itself to Do Surgery by Watching Videos—and It’s Able to Operate on Humans
Jesus Diaz | Fast Company
“For the primary time in history, Kim and his colleagues managed to show a synthetic intelligence to make use of a robotic surgery machine to perform precise surgical tasks by making it watch 1000’s of hours of actual procedures happening in real surgical theaters. …In accordance with their recently published paper, the researchers say the AI managed to realize a performance level comparable to human surgeons without prior explicit programming.”
Recent Fastest Supercomputer Will Simulate Nuke Testing
Dina Genkina | IEEE Spectrum
“El Capitan was announced yesterday on the SC Conference for supercomputing in Atlanta, Georgia, and it debuted at #1 within the newest Top500 list, a twice-yearly rating of the world’s highest performing supercomputers. …[The supercomputer], housed at Lawrence Livermore National Laboratory in Livermore, Calif., can perform over 2700 quadrillion operations per second at its peak. The previous record holder, Frontier, could just do over 2000 quadrillion peak operations per second.”
A Chinese Lab Has Released a ‘Reasoning’ AI Model to Rival OpenAI’s o1
Kyle Wiggers | TechCrunch
“On Wednesday, DeepSeek, an AI research company funded by quantitative traders, released a preview of DeepSeek-R1, which the firm claims is a reasoning model competitive with o1. …Just like o1, DeepSeek-R1 reasons through tasks, planning ahead, and performing a series of actions that help the model arrive at a solution. This will take some time. Like o1, depending on the complexity of the query, DeepSeek-R1 might ‘think’ for tens of seconds before answering.”
AI Could Cause ‘Social Ruptures’ Between People Who Disagree on Its Sentience
Robert Booth | The Guardian
“Significant ‘social ruptures’ between individuals who think artificial intelligence systems are conscious and people who insist the technology feels nothing are looming, a number one philosopher has said. …Last week, a transatlantic group of academics predicted that the dawn of consciousness in AI systems is probably going by 2035 and one has now said this might lead to ‘subcultures that view one another as making huge mistakes’ about whether computer programs are owed similar welfare rights as humans or animals.”
Get in, Loser—We’re Chasing a Waymo Into the Future
Wired Staff | Wired
“To supply essentially the most useful dispatch from the longer term…we realized we wanted a strategy to make self-driving cars feel strange again. A strategy to scare up the less superficial lessons of our city’s years with Waymo. …Our idea: We’ll pile a number of of us into an old-fashioned, human-piloted hired automobile, then follow a single Waymo robotaxi wherever it goes for an entire workday. We’ll study its movements, its relationship to life on the streets, its whole self-driving gestalt. We’ll interview as a lot of its passengers as will speak to us, and observe it through the eyes of the form of human driver it’s designed to exchange.”
Microsoft and Atom Computing Mix for Quantum Error Correction Demo
John Timmer | Ars Technica
“The 2 corporations [released] a draft manuscript describing their work on error correction [this week]. The paper serves as each a superb summary of where things currently stand on the earth of error correction, in addition to a superb take a look at a few of the distinct features of computation using neutral atoms.”
OpenAI Considers Taking up Google With Browser
Erin Woo, Sahil Patel, and Amir Efrati | The Information
“OpenAI is preparing to launch a frontal assault on Google. The ChatGPT owner recently considered developing an online browser that it could mix with its chatbot, and it has individually discussed or struck deals to power search features for travel, food, real estate and retail web sites, in line with individuals who have seen prototypes or designs of the products.”
INTERNET
Bluesky Says It Won’t Screw Things Up
Steven Levy | Wired
“In little greater than every week, its numbers soared from 14 million to twenty million and were growing at a pace of 1,000,000 a day. …After I spoke this week to Bluesky CEO Jay Graber, she was gratified by the brand new users. ‘It’s been a wild week,’ she says. But she noted that this spike was certainly one of several over the past few months. Bluesky, she says, is in it for the long haul. The thought isn’t to recreate classic Twitter, she says, but to reshape social media on the principle of openness and user control.”
All Life on Earth Today Descended From a Single Cell. Meet LUCA.
Jonathan Lambert | Quanta
“The [new analysis] sketched a surprisingly complex picture(opens a brand new tab) of the cell. LUCA lived off hydrogen gas and carbon dioxide, boasted a genome as large as that of some modern bacteria, and already had a rudimentary immune system, in line with the study. Its genomic complexity, the authors argue, suggests that LUCA was certainly one of many lineages — the remainder now extinct—living about 4.2 billion years ago, a turbulent time relatively early in Earth’s history and long thought too harsh for all times to flourish.”
