AI
ChatGPT exposes the hidden trends in 2025 prediction reports
"By creating trend reports stuffed with catchy buzzwords, humans teach AI to prioritise these same repetitive ideas."
Jasper Hamill
Existential Risk
OpenAI's apocalyptic worst-case existential risk scenarios
How could an AI model take over the world and destroy humanity? OpenAI found out...
AI
Anthropic observes AI faking its "alignment" to deceive humans in ominous world-first experiment
If relatively basic large language models (LLMs) are already giving us the run-around, what hope do we have against AGI superintelligence?
AI
Jailbreaking Sora: OpenAI reveals how new model was tricked into making x-rated videos
Video-to-text bot appears to be fond of making raunchy material in a fantasy, medical, or science fiction setting, AI firm reveals.
Internet
"Please stop!": Elon Musk shadowbans "ugly" hashtags on X
Billionaire says "the system doesn't need" the hashtags first made famous by Twitter...
Internet
AI doomer explodes: AISafetyMemes goes nuclear over Beff Jezos doxxing drama
Cold war goes hot as long-standing conflict between champions of AI Safety and effective accelerationism (e/acc) intensifies.
AI
Microsoft admits new Phi-4 small language model is "tedious"
Waffly model delights in giving "long and elaborate" answers that can be just a little bit dull.
Business
Bipartisan bill creates panel to help Trump make the US a global XR leader
Commerce department officials will advise the president on how to ensure the US plays a guiding role in this fast-growing market.
Cyberwar
China's secret surveillance tool has a strange limitation
Sinister spy software lets cops collect "extensive sensitive data" from people's phones and snap pictures from their camera
AI
OpenAI outage: ChatGPT went down and people lost their sh*t
Students grappled with the grim reality of having to write essays and do exam revision the old-school way after an OpenAI horror outage.
Security
Krispy Kreme gets Kracked: Doughnut giant has a cybersecurity hole
Cake fans left glazed and confused as mystery incident causes major operational disruption.
Internet
Humans exploit female AI models and distrust male bots, study finds
Experiment uncovers clear gender biases in the way people treat AI agents.