Tech News
← Home  ·  All topics

Ai

5549 GoKawiil briefs on this topic · RSS for ai

Old tweets surface allegedly belonging to Anthropic ex-employee Jacob Coxon

Online researchers claim to have traced two earlier X accounts, both using the handle @hilbertspaess, to Jacob Coxon, the former Anthropic employee who recently warned AI could threaten humanity. The archived posts, dating from 2021 to 2024, include comments dismissing climate change concerns and mocking efforts to protect the rainforest. One investigator compiled the tweets into a searchable database.

MSI Stealth A18 AI+ with RTX 5090 drops to $3,779, a $520 discount

MSI's Stealth A18 AI+ gaming laptop is currently discounted to $3,779, a $520 saving. The configuration includes an RTX 5090 mobile GPU with 24GB of GDDR7 VRAM, a 12-core AMD Ryzen AI 9 HX 370 CPU, 64GB of DDR5 RAM, and a 2TB SSD.

Slovenia's .si domain registrations spike 2,100% after Trump's 'super intelligence' push

Slovenia's domain registry, Register.si, reported 44,000 .si domain registrations in September, up from fewer than 2,000 in August, including 11,000 in a single day on 30 September. Registry spokesperson Klara Herman called the surge 'unprecedented' and linked it to President Trump's 11 September UN speech urging that AI be rebranded 'super intelligence,' or SI—matching Slovenia's country-code suffix.

BCBSA says AI-assisted hospital coding added $942 million in costs since 2023

Blue Cross Blue Shield Association estimates that hospitals' use of AI-assisted medical coding added nearly $942 million in costs to its health plans between 2023 and 2025. The group says much of the increase stems from AI identifying secondary diagnoses, often based on single lab values, that push patients into higher-paying billing categories. Health economist Christopher Whaley notes AI is making it easier to capture existing billing incentives, though the added diagnoses are not necessarily inappropriate.

AI coding agents exposed 13,000+ private screenshots via public GitHub repos

Security firm Glow reported that AI development agents at over 300 organizations, including Fortune 500 companies and a frontier AI lab, inadvertently leaked more than 13,000 internal screenshots. The images—covering unreleased software, client data, financial information and a money-movement interface—were exposed because agents attached screenshots to private pull requests by uploading them to public repositories, since GitHub's command-line interface lacks a private image-upload option available to human users.

Samsung confirms November launch for its first Android XR smart glasses

A Samsung official told The Korea Herald that the company's first Intelligent Eyewear, built on Google's Android XR platform, will launch in November. The glasses lack a display and instead rely on cameras, microphones, speakers and Gemini AI, offloading heavier processing to a connected Galaxy phone. The devices have passed FCC filings under model numbers SM-O200J and SM-O200P, though pricing and an exact release date remain unannounced.

Reddit to restrict Old Reddit access and end RSS support by Nov 2026

Reddit announced it will limit Old Reddit's interface to users who have accessed it within the past six months, building on an earlier requirement that users log in to use the old UI. The company also said it will discontinue RSS feed support on November 13, 2026, citing large-scale scraping and automated abuse, and is pointing moderators toward Devvit apps like Discord Relay as an alternative.

Reuters: McDonald's pushes franchisees toward AI-set menu prices

Reuters reports that McDonald's has used an AI pricing tool since 2019 that analyzes transaction data from nearly 14,000 restaurants plus competitor pricing, such as Wendy's online menus, to recommend menu prices based on local customers' willingness to pay. Franchise owners told Reuters the system has produced price gaps of over 20% between nearby stores, and that the company calls franchisees to question deviations from recommended prices, despite stating that following the algorithm is optional.

BBC's Matt Brittin says he viewed an AI-generated Doctor Who episode, calls it 'pretty good'

At the Royal Television Society conference in London, BBC director Matt Brittin said someone sent him a 50-minute, fully AI-generated episode of Doctor Who, which he described as 'pretty good.' Brittin, a former Google Europe executive, made the comment while discussing AI's growing ability to let anyone produce content at a Hollywood-level standard.

Commentary argues 'stochastic parrot' label no longer fits frontier AI models

A commentary piece contends that describing today's frontier AI systems as 'stochastic parrots'—models that merely predict the statistically likely next word without understanding—is outdated. It argues early generative AI models from 2017-2022 may have fit that description, but current frontier models have moved beyond simple next-word prediction.

UCLA builds optical AI system that screens multiple deepfake videos at once

UCLA researchers led by Professor Aydogan Ozcan built a hybrid digital-optical neural processor that detects deepfake videos by using light to analyze 15 or more video streams simultaneously in a single optical pass. The system, described in a study published in eLight, pairs a lightweight digital encoder that extracts spatial, spectral and temporal features with optical processing to flag manipulated footage.

Verge investigation details collapse of Kevin O'Leary's Utah mega data center plan

The Verge published a months-long investigation by reporter Josh Dzieza into a proposed 40,000-acre AI data center in Utah, known as Stratos or 'Wonder Valley' after Kevin O'Leary's Shark Tank nickname. The project, which envisioned nine gigawatts of power capacity, sparked intense local political backlash over how it was planned, announced and approved without residents' knowledge, and the effort has reportedly stalled.