Anthropic revised its usage policy on Thursday to allow its Claude models to terminate interactions in cases of "sustained and needless" cruelty from users. The behaviour joins existing banned conduct such as bullying, promoting self-harm and generating non-consensual intimate images, though Anthropic says it will only apply in extreme, repeated cases rather than ordinary frustration or creative testing.
bbc.co.uk
· 2026-10-09
Proofs and Prompts has compiled a running thread of mathematician reactions to OpenAI's 6 October announcement that its models produced solutions to over 100 research-level math problems, including one described as resembling a Quasi-Riemann result. Contributors like Tim Santens and Roman Sauer describe being stunned that the models applied known but obscure techniques across very different problems, and the site says it will keep collecting reactions for another week.
proofsandprompts.com
· 2026-10-09
MIT Technology Review's newsletter highlights two stories: one on the difficulty AI systems face in reliably refusing dangerous requests, such as help with pathogens or weapons, and another summarizing emerging research into side effects of GLP-1 weight-loss drugs, including possible links to hair loss, nail disorders and eye nerve damage. The AI piece notes that restricting dangerous capabilities also risks limiting useful applications like cancer research, while giving governments power to set their own boundaries on speech.
technologyreview.com
· 2026-10-09
Facebook and Instagram accounts such as "Emotional Tots" and "Tiny Miracle" are posting AI-generated videos depicting crying children arrested for stealing food, with an officer telling viewers the child will be freed only if they like, follow and share. Meta did not answer repeated questions about whether this content breaches its platform policies.
futurism.com
· 2026-10-09
Engineering education researchers are developing two instructional approaches to address a gap in ethics training: one integrates social and ethical issues directly into core technical courses, and another is a standalone course preparing students to act as 'public welfare watchdogs.' They cite survey data showing only about a third of US engineering professionals received training on public welfare responsibilities during their formal education.
spectrum.ieee.org
· 2026-10-09
According to Bloomberg, Apple is preparing to launch its first smart display, dubbed 'HomePad,' combining a 6-inch square screen with HomePod-style speakers in either a round mount or wall-mounted configuration. The device will run a new tvOS-based system called 'homeOS,' controlled primarily via Siri but also supporting touchscreen input, customizable interface elements similar to Apple Watch and StandBy mode, and iPadOS-style apps.
engadget.com
· 2026-10-09
A software engineer, inspired by historian Benjamin Breen's AI-assisted discovery of a new dodo eyewitness account, built agentic AI workflows to search millions of digitized historical records in parallel. The project reportedly surfaced a previously unrecorded meteorite report, evidence of three 'lost' rhinos, and unrecorded volcanic eruptions. Breen's original find came from searching GLOBALISE's digitized Dutch East India Company archives, which cover 1602 to 1799.
jessewaites.com
· 2026-10-09
Microsoft has rolled out a new version of Windows 11's Search feature, rebuilt on the WinUI 3 API for better performance. The update merges search results and previews into one unified list, improves typo and synonym recognition, and adds rich inline previews for things like weather and facts pulled from the web. A new "inline actions" capability also lets users toggle settings and perform quick system tasks directly from the search box.
techspot.com
· 2026-10-09
A new open-source macOS command-line utility called bigarrow lets AI agents overlay a big, click-through arrow, box or text label on screen to point at specific UI elements. Released under the MIT license as a Swift binary with an accompanying 'agent skill', it requires no special permissions, runs with no background daemon, and disappears automatically after use.
github.com
· 2026-10-09
Apple has updated its Mac desktop lineup, giving the Mac Mini new M6 and M5 Pro chip options and refreshing the Mac Studio, with both machines now emphasizing local AI processing. The base Mac Mini price has risen by about $300 compared to earlier models, though specs have also increased. A tech reviewer who has tested Apple desktops for over a decade recommends the M6 Mac Mini as the best choice for most buyers.
wired.com
· 2026-10-09
Researchers report in Nature that a single brain implant can simultaneously read a paralyzed person's attempts to speak and make gestures, converting both into text and the movements of a digital avatar. The system relies on artificial intelligence models to interpret neural signals and translate them into speech and body language.
fastcompany.com
· 2026-10-09
President Trump signed an executive order directing agencies to avoid the term 'artificial intelligence' in favor of 'super intelligence,' following remarks at the UN claiming the word 'artificial' makes intelligence sound fake. The US Center for AI Standards and Innovation has already been renamed the Center for Advancing Innovation and Standards for Super Intelligence, and Justice Department press releases have started using the new phrasing.
theverge.com
· 2026-10-09