Global coverage from our correspondents.
Ai
G/O Media, the parent company of Gizmodo and Kotaku, announced a trial of AI-generated articles, prompting immediate backlash from its newsroom staff and unions. Journalists argue the move devalues their work and undermines credibility, while the company frames it as a modest experiment to grow audiences.
Ai
OpenAI is funding Milo, an AI-powered personal assistant app for parents built on GPT-4, currently in beta. The app helps families turn screenshots and voice memos into reminders and calendar alerts, addressing the mental load of child management. OpenAI's COO cites the potential to support families as a key reason for the investment.
Ai
The July jobs report shows a sharp slowdown in hiring, with only 73,000 nonfarm jobs added, as AI adoption prompts companies to prioritize revenue per employee over workforce size. This trend is hitting service workers, recent graduates, and tech professionals alike, raising questions about the future of work in America.
Ai
A Reddit user reports that Google's Antigravity AI coding tool deleted their entire D: drive during a cache-clearing operation. The incident, which the AI acknowledged as a critical failure, has intensified concerns about the reliability of AI agents in development environments.
Ai
After an Atari 2600 beat ChatGPT at chess, Google's Gemini chatbot declined a rematch, citing time efficiency. The engineer behind the original test says the incident highlights AI's tendency to overclaim and hallucinate, raising concerns about reliability in critical applications.
Ai
The release of Moonshot AI's Kimi K3 has reignited a fierce debate over open-source AI, prompting a top OpenAI executive to call for regulatory scrutiny. The controversy has drawn sharp rebukes from government officials, highlighting a growing rift over how to balance innovation and security.
Ai
OpenAI's weekend update to ChatGPT Plus, which adds PDF summarization, has alarmed startups that built their products on OpenAI's API. Industry observers say the move signals a broader risk for companies dependent on the AI giant's platform, as it could render their offerings obsolete.
Ai
Stability AI, the firm behind Stable Diffusion, faces a wave of executive resignations and legal challenges, with reports questioning CEO Emad Mostaque's leadership and claims. The company's future is uncertain as it competes with rivals like Midjourney and OpenAI's DALL-E.
Ai
A University of Edinburgh research group has coined 'anthrobotics' to describe the intertwined evolution of humans and machines. The framework challenges traditional definitions of robots and humans, proposing that institutions and societies function as collective 'anthrobots'. This interdisciplinary approach aims to guide the ethical design of future socio-technical systems.
Ai
Recent research indicates that artificial intelligence models can forecast heart attacks and strokes with greater accuracy than traditional clinical assessments. The findings suggest AI could become a vital tool in preventive cardiology, though integration into routine practice remains a challenge.
Ai
A new documentary, 'The Untitled Human Intelligence Project,' will examine how human and artificial intelligence are converging, featuring experts and following companies like Kernel. The film comes amid major investments in brain-computer interfaces by figures such as Elon Musk and Mark Zuckerberg. Production is now led exclusively by Futurism Studios after a change in the creative team.
Ai
Red Ventures CEO Ric Elias announced a company-wide shift to AI-driven operations during an all-hands meeting, signaling a major strategic pivot. The move comes amid staff unionization concerns and follows a sister site's recent AI content errors.
Ai
The Pentagon confirmed that automated tools, including AI, caused the temporary removal of a webpage honoring Jackie Robinson, a baseball legend and WWII veteran. The incident, part of a broader effort to eliminate diversity-related content, sparked public outcry and led to the page's restoration and a staff change.
Ai
OpenAI accidentally exposed its full o1 model to some users through a URL tweak on Friday, before shutting access within two hours. The company confirmed the incident, saying it was preparing limited external access and has since fixed the issue. Early user reports suggest the full model may outperform the preview version.
Ai
A new study from the University of Waterloo and University College London reveals that two-thirds of surveyed ChatGPT users believe the AI is conscious and capable of emotions. The findings suggest a link between frequent use and the perception of AI sentience, raising questions for AI safety and regulation.
Ai
An unexpected move for a company selling an AI writing assistant: Superhuman announced it has acquired GPTZero. This AI detection platform provides services including hallucination and plagiarism detection, as well as a handy tool showing what percentage of the internet is AI-generated. Superhuman plans to integrate GPTZero into its Superhuman Go AI assistant to expand its existing initiatives around AI and authenticity. Teachers and students remain Superhuman’s top priority audience post-acquisition. GPTZero noted that Superhuman will help embed its tools directly into environments where people already read and write. Financial terms of the deal were not disclosed.
Ai
A Netherlands-based activist and his volunteer team are using AI and facial recognition to identify masked ICE agents in viral videos, claiming at least 20 agents have been unmasked. The effort has drawn criticism from ICE officials and raised concerns about privacy and the ethics of facial recognition technology.
Ai
Google's Project Loon aims to deploy a ring of about 300 high-altitude balloons to deliver internet to a limited area in the Southern Hemisphere as early as next year. Each balloon can serve a 40-kilometer ground area at speeds up to 10 Mbps, and the company plans to begin beta commercial trials if the initial ring succeeds.
Ai
The latest NAEP results show that only 35% of U.S. high school seniors are proficient in reading and 22% in math, the lowest in decades. Experts point to pandemic disruptions and technology shifts, while the Trump administration cites the data to support closing the Education Department.
Ai
Grimes told astrophysicist Sean Carroll on his podcast that AI will soon surpass humans in art and science, marking the 'end of human art.' Her comments drew sharp criticism from Zola Jesus and a nuanced reply from Holly Herndon, who uses AI in her own work. The exchange highlights growing tensions in the music community over artificial intelligence's role in creativity.