Stop the use of AI in war until laws can be agreed www.nature.com/articles/d41...
Stop the use of AI in war until laws can be agreed www.nature.com/articles/d41...
ChatGPT, Gemini, and other chatbots helped teens plan shootings, bombings, and political violence, study shows www.theverge.com/ai-artificia...
Toby Walsh: Can Australia meet the AI moment? www.innovationaus.com/toby-walsh-c...
The UNβs high commissioner for human rights said in October that the drone strikes were disproportionate and likely unlawful. www.aljazeera.com/news/2026/3/...
Grammarly is using our identities without permission www.theverge.com/ai-artificia...
Join a fascinating summer school on algorithmic game theory projects.croucher.org.hk/summer-cours...
Human Fracking, AI and the Capture of Attention with D. Graham Burnett events.humanitix.com/human-fracki...
Toby Walsh on the Six Ideas You Need to Understand AI www.events.unsw.edu.au/article/toby...
App stores, search engines in eSafetyβs sights over AI age rules www.innovationaus.com/app-stores-s...
Perhaps Anthropic's fight with Department of War is just to distract from them backtracking on safety?
Anthropic no longer requires new models to have adequate safety measures in place if it judges that it is not leading the global AI race.
Of course, we should never use large language models to control lethal autonomous weapons. And the US military have promised to keep appropriate levels of human judgment over the use of force. LLMs are not that! www.theguardian.com/us-news/2026...
I somehow doubt that commercially licensing Alan Turing's voice generated by AI (as ElevenLabs is doing) honours his remarkable legacy. Seems all a bit grubby to me. elevenlabs.io/iconic-marke...
Shouldn't all platforms do this? techcrunch.com/2026/02/26/i... via @techcrunch
I suspect this won't end well.
βinviting in foxes to consult on the future of the henhouseβ.
www.theguardian.com/technology/2...
Fills you with confidence that META's head of AI safety makes this rookie mistake .... www.businessinsider.com/meta-ai-alig...
How AI slop is causing a crisis in computer science www.nature.com/articles/d41...
You would have thought once would be more than enough?
Testimony from one former top safety executive at Meta, alongside internal documentation, claimed Meta once had a β17xβ strike policy for accounts engaged in the βtrafficking of humans for sexβ.
No it doesn't.
It costs only 0.02 GWh to train a human (energy in 20 years of food) compared to around 50 GWh to train GPT5.
Altman is once again talking out of his posterior ...
AI boomers and doomers alike mistake intelligence for power.
The most intelligent person in the room is frequently not the most powerful.
www.thefp.com/p/you-are-no...
Usual service from tech giants www.theguardian.com/technology/2...
The Race for AI: Boom or Doom?
Please join me for a provocative address to the National Press Club on Feb 25th. npc.org.au/events/scien...