Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, and Bluesky):AI find security vulnerabilities so quickly we need AI to find solutions: https://www.rnz.co.nz/news/science-and-technology/618615/fighting-ai-with-ai-the-only-way-to-keep-up-on-new-frenetic-frontier
AI coding costs will soon exceed the cost of human developers: https://www.theregister.com/ai-and-ml/2026/06/24/ai-coding-agents-could-soon-cost-more-than-the-developers-using-them/5260864
Loop engineering looks like a scheme to funnel more money to AI companies: https://www.theregister.com/ai-and-ml/2026/06/24/loop-engineering-latest-ai-buzzword-still-needs-humans-in-the-loop/5261735
People have no patience with AI service agents: https://futurism.com/artificial-intelligence/customers-fed-up-ai-service-agents
How to beomce an AI architect: https://www.kdnuggets.com/the-roadmap-to-becoming-an-ai-architect-in-2026
Microsoft used AI to analyse malware infrastructure, leading to it being taken down: https://www.theregister.com/security/2026/06/24/microsoft-uses-ai-to-link-two-malware-operations-in-racketeering-suit/5261656
What do mathematicians do now that AI can do maths?: https://spectrum.ieee.org/ai-in-mathematics
AI raises a lot of issues with disaster recovery planning: https://www.informationweek.com/incident-response/ai-disaster-recovery-planning-is-years-behind-ai-adoption
AI means that security incidents can quickly escalate: https://www.theregister.com/security/2026/06/23/five-eyes-spooks-warn-ai-means-infosec-incidents-can-become-major-operational-and-financial-crises/5259916
Games that used generative AI receive poorer reviews compared to human-only created games: https://www.extremetech.com/gaming/games-that-use-generative-ai-receive-53-poorer-firstmonth-reviews-study
Ideology should never be a reason to retract an article. Boycott the Journal of the New Zealand College of Clinical Psychologists. Do not submit to it, do not review for it, do not cite it: https://www.stuff.co.nz/nz-news/360997918/peer-reviewed-article-maori-author-removed-because-it-wasnt-line-organisations-values
Organisations need to start preparing now for the security risks posed by AI: https://www.rnz.co.nz/news/politics/616849/nz-organisations-told-to-prepare-for-significant-rise-in-risks-from-ai
There is a strong argument that the companies most responsible for replacing human workers with AI should be responsible for retraining those workers: https://www.theregister.com/ai-and-ml/2026/06/26/ai-giants-back-non-profit-to-retrain-workers-left-behind-by-ai/5262601
Medical AI can be tricked into disclosing confidential information: https://www.theregister.com/ai-and-ml/2026/06/24/medical-diagnosis-ais-can-be-tricked-into-telling-whose-data-trained-them/5261501
We need regulation of AI: https://www.rnz.co.nz/news/in-depth/618503/karen-hao-exposing-the-brute-force-of-ai-that-is-trying-to-make-humans-redundant
AI is starting to learn the nuances of human emotion: https://spectrum.ieee.org/emotion-ai-context
AI is driving up the price of everything: https://www.computerworld.com/article/4186923/the-ai-revolution-comes-with-a-hidden-tax.html
A project that uses AI to help screen for bowel cancer: https://www.rnz.co.nz/news/health/631840/ai-promises-breakthrough-in-bowel-cancer-detection
Organisations are rolling out AI faster than their infrastructure can cope with: https://www.theregister.com/devops/2026/06/24/companies-are-not-looking-before-theyre-leaping-into-the-ai-playpen/5261819
AI is the asbestos in the walls of our technological society says Cory Doctorow: https://arstechnica.com/gadgets/2026/06/how-to-burst-the-ai-bubble-strike-at-its-roots/
AI still can’t tell the difference between fact and fiction: https://futurism.com/artificial-intelligence/duckduckgo-ai-trump-rabies
Adjusting the clock speed of GPU can reduce energy use without losing computation speed: https://spectrum.ieee.org/llm-training-energy-saving-trick
Experience software engineers are spending a lot of time fixing AI generated code: https://futurism.com/artificial-intelligence/software-engineers-crisis-drown-ai-code
AI helps to recover some of the text on a Roman scroll that was burned in the eruption of Vesuvius: https://www.theguardian.com/technology/2026/jun/24/ai-read-papyrus-scroll-burnt-vesuvius-eruption
Competitors can duplicate an AI if they can query it enough: https://www.computerworld.com/article/4189347/anthropic-accuses-alibaba-of-using-25000-fake-accounts-to-scrape-claude-ai-2.html
