Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, and Bluesky):Prompt injection is a new attack vector made possible by AI browsers: https://www.theregister.com/2025/10/28/ai_browsers_prompt_injection/Using AI is changing how our brains use language: https://www.rnz.co.nz/life/wellbeing/how-generative-ai-could-change-how-we-think-and-speakAI is making the rich and powerful richer and more powerful: https://www.theregister.com/2025/10/30/meta_alphabet_q3_2025/Scientists must take responsibility for responsible AI: https://spectrum.ieee.org/responsible-aiYour AI search result is based on less popular pages. Does less popular mean less reliable though? https://arstechnica.com/ai/2025/10/ai-powered-search-engines-rely-on-less-popular-sources-researchers-find/I don’t think the kind of performative Christianity American oligarchs preach is going to be improved by AI: https://www.theguardian.com/technology/2025/oct/28/patrick-gelsinger-christian-ai-gloo-silicon-valleyMore corporate job losses because of AI: https://www.stuff.co.nz/world-news/360868458/amazon-cuts-14000-corporate-jobs-spending-artificial-intelligence-accelerates AI is making the rich richer, while distracting the rest of us with fake videos.The majority of businesses that lay people off in favour of AI end up regretting it, mostly because AI can’t deliver: https://www.theregister.com/2025/10/29/forrester_ai_rehiring/AI is good at discovering new algorithms: https://www.theregister.com/2025/10/25/openevolve_ai_better_algorithms/As AI hardware gets more powerful, AI models get more demanding. Just like every other piece of software: https://spectrum.ieee.org/mlperf-trendsAI will lead to neither utopia or disaster. But it will certainly make the rich richer: https://www.computerworld.com/article/4078090/whos-right-the-ai-zoomers-or-doomers.htmlSocialising with AI might make teenage boys feel less lonely, but how well prepared will they be for the real world afterwards? https://www.theguardian.com/technology/2025/oct/30/teenage-boys-using-personalised-ai-for-therapy-and-romance-survey-findsWill the AI bubble burst, or just slowly deflate? https://www.theregister.com/2025/10/28/forrester_ai_spending/RNZ used AI to clone a dead man’s voice, But only with the consent of his estate and family, and only to read words he had written in life. Ethical? https://www.rnz.co.nz/news/nark/577167/nark-how-and-why-we-used-ai-to-recreate-a-dead-man-s-voiceEnough people discuss self-harm with AI that the AI really need guardrails: https://arstechnica.com/ai/2025/10/openai-data-suggests-1-million-users-discuss-suicide-with-chatgpt-weekly/AI agents are now a vulnerability when it comes to phishing attacks: https://spectrum.ieee.org/ai-agent-phishingIt doesn’t matter if you are an AI company, you still have to follow the law: https://www.theregister.com/2025/10/28/noyb_criminal_charges_clearview/Energy is still the limiting factor in building more capacity in AI: https://www.theregister.com/2025/10/28/openai_100gw_power_demand/Every AI will have some biases inherited from its creators/owners. When that owner is a right-wing oligarch, those biases will be more obvious and more damaging: https://www.theguardian.com/technology/2025/oct/28/elon-musk-grokipediaAI might listen more than human doctors, but it will not give reliable health advice: https://www.theguardian.com/society/2025/oct/28/deepseek-is-humane-doctors-are-more-like-machines-my-mothers-worrying-reliance-on-ai-for-health-adviceDo not use AI to challenge a private investigator’s invoice: https://www.nzherald.co.nz/nz/private-investigator-cleared-after-client-used-ai-to-challenge-his-report/FHVEF32FBRAZHNCCUJKMJ2CR5U/A new use for generative AI – faking receipts for expense claims: https://arstechnica.com/ai/2025/10/ai-generated-receipts-make-submitting-fake-expenses-easier/Opt-out of LinkedIn is going to start using your data to train their AI: https://www.theregister.com/2025/10/28/openai_100gw_power_demand/AI companies rely on free access to data, but when taking that data makes existing biases worse, something needs to be done to protect it: https://www.rnz.co.nz/news/national/577327/iwi-leaders-unveil-ai-safeguards-to-protect-maori-data
