News Archive
Pick a date to read that day's stories.
Days with a blue dot have stories.
- 01 of 10LAUNCH
Meta releases Muse Wonder, its first flagship AI
Meta has launched Muse Wonder and a beefier Muse Wonder Full, its most capable AI models yet, built for tasks like reading a lease or coding an app. The models use what Meta calls Active Memory to remember details across chats, and pricing puts them roughly level with rivals from OpenAI and Anthropic.
OUR TAKEMeta's assistant finally remembers your plans, which is more than can be said for most humans.
Read more → - 02 of 10HEALTH
Google details AI tools aimed at disease research
Google has published an overview of how its AI systems are being used to speed up drug discovery, read genetic data and forecast disease outbreaks. The company says its research teams have published dozens of papers and are sharing some tools openly with scientists and doctors.
OUR TAKEA tech giant promising to cure disease is now a genre, but at least this instalment comes with footnotes.
Read more → - 03 of 10SOCIETY
Singapore PM says AI must serve people, not run them
Singapore's Prime Minister Lawrence Wong said the country wants AI to improve daily life rather than drive it, urging schools to protect basic skills like maths and employers to keep mentoring junior staff. He said the government will manage the transition over roughly a decade, with retraining and fairer support for workers who lose jobs.
OUR TAKEA rare case of a leader asking AI to serve the public rather than the other way round.
Read more → - 04 of 10COOPERATION
Experts urge South Korea to join global AI project
AI scientist Yann LeCun and other experts told a Seoul briefing that South Korea should build its own AI models while also joining international efforts such as Project Tapestry, which pools computing power across countries. They said training costs are climbing with diminishing returns, making shared development more sensible.
OUR TAKENothing says international harmony like several countries splitting one enormous electricity bill.
Read more → - 05 of 10GOVERNMENT
UAE reviews AI platform for crisis decisions
Dubai's Crown Prince Sheikh Hamdan reviewed TACTICA, an Abu Dhabi built AI platform that pulls together real time data to help officials respond faster during emergencies and security incidents. The system is already being applied across defence, crisis management and city operations.
OUR TAKEA system designed to make faster decisions, reviewed in person by a prince who presumably arrived on time.
Read more → - 06 of 10CRIME
Indian family loses savings to AI voice clone scam
Police in Nagpur, India are investigating a case in which a caller used an AI generated voice copying a well known local community figure to trick a family into handing over roughly 5 million rupees in cash. The victim believed he recognised the voice and sent the money through an acquaintance before realising the call was fake.
OUR TAKEProof that a familiar voice is no longer proof of anything, least of all good intentions.
Read more → - 07 of 10SECURITY
Britain adds AI defences against cyber threats
The UK government announced new measures to use frontier AI to strengthen its own cyber defences, after pilots already found and fixed more than a thousand vulnerabilities across government systems. Ministers also said they are preparing for serious incidents involving autonomous AI agents acting without close human supervision.
OUR TAKEBritain is teaching its computers to fight off the computers, which feels like the plot everyone warned us about.
Read more → - 08 of 10REGULATION
EU softens AI Act timeline in new Omnibus deal
The European Union formally adopted changes to its AI Act that push back several compliance deadlines, giving companies until December 2027 for many high risk systems instead of this year. The update keeps the law's core risk based structure but adds new bans on deepfake sexual images and AI generated abuse material.
OUR TAKEBrussels giving everyone more homework time while quietly adding new rules nobody will argue with.
Read more → - 09 of 10PRODUCT
Oppo's new phone software leans on AI helpers
Oppo is rolling out ColorOS 17 globally and in India, built on Android 17 with a suite of AI tools including a photo editor, meeting note taker and a feature called Mind Suggest that tries to guess what users need before they ask. The update also adds AI based warnings for scam calls and face swapped video.
OUR TAKEA phone that anticipates your needs is handy, right up until it anticipates the wrong ones.
Read more → - 10 of 10FUNDING
African investors want proof AI startups can last
A new report from TechCabal finds African venture capitalists are growing wary of AI startups that are little more than a wrapper around existing chatbot tools. Investors interviewed said they are now looking for founders who can show real technical depth and a business that would not collapse if a bigger company copied the idea.
OUR TAKETurns out "we just plugged in a chatbot" is not the pitch deck line it once was.
Read more →