Weekly Review 1 May 2026

Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, and Bluesky):

1. AI generated music is becoming more common, and harder to detect by listeners: https://arstechnica.com/ai/2026/04/deezer-says-44-of-new-music-uploads-are-ai-generated-most-streams-are-fraudulent/
2. Michael Dell is funding research towards an AI-led hospital system: https://www.stuff.co.nz/business/360969141/michael-and-susan-dell-fund-ainative-medical-center-750
3. Out of date UK government web pages have been ingested by AI, which are now giving people inaccurate advice: https://www.theregister.com/2026/04/23/stale_govuk_pages_are_feeding/
4. Mythos AI is such a security threat that it won't be released to the public. So who is going to be using it? https://www.theguardian.com/technology/2026/apr/22/what-is-anthropic-mythos-ai-threat-global-cybersecurity
5. Google's structural AI advantage: https://www.theregister.com/2026/04/23/google_cloud_next_interview_ai_stack/
6. I can't stand micromanaging humans, I certainly wouldn't put up with a micromanaging AI: https://futurism.com/artificial-intelligence/nvidia-ceo-ai-micromanaging-boss
7. The storage crunch caused by AI is being exacerbated by disruptions to fuel supplies: https://www.theregister.com/2026/04/23/everpure_letter_covid/
8. More AI security tools, to fight AI security threats. At this point it's all just AI vs AI: https://www.theregister.com/2026/04/22/google_unleashes_even_more_ai/
9. Being able to judge what an AI tells you is the most important part of being AI-ready: https://www.computerworld.com/article/4149417/ai-ready-skills-are-not-what-you-think.html
10. Some AI will make psychosis worse in their users: https://futurism.com/artificial-intelligence/certain-chatbots-worse-ai-psychosis-study
11. I don't want any company monitoring my screen, let alone an AI company: https://www.theregister.com/2026/04/22/openai_chronicle_no_privacy_screenshot/
12. Another AI assisted system break-in: https://www.theregister.com/2026/04/21/vercel_ceo_points_to_aidriven/
13. After medical advice, legal advice is the worst use-case for AI: https://www.rnz.co.nz/news/business/592911/ai-tells-tenant-she-should-ask-for-40-000-tribunal-hands-her-80
14. More tech job cuts justified with AI: https://www.rnz.co.nz/news/world/593297/meta-to-cut-10-percent-of-staff-as-it-pours-billions-into-ai
15. If an AI helps someone to commit a crime, is the AI liable? https: https://www.stuff.co.nz/world-news/360968207/floridas-attorney-general-launches-criminal-probe-chatgpt-over-shooting
16. So Meta's employees won't even know that they're training their AI replacements, just by doing their work: https://www.computerworld.com/article/4161929/meta-to-track-employee-keystrokes-screen-activity-to-train-ai-agents.html
17. No, AI data centres in space are NOT a good idea: https://futurism.com/space/spacex-admits-ai-data-centers-terrible-idea
18. DO NOT use AI to develop a cancer treatment plan: https://futurism.com/artificial-intelligence/ai-chatbots-cancer-alternative-cures
19. Supply chain problems are going to continue to cause problems for developing AI data centres: https://futurism.com/artificial-intelligence/trump-data-center-fermi
20. Most AI generated bug reports continue to be rubbish: https://dataconomy.com/2026/04/22/ai-flood-drives-surge-in-bogus-crypto-bug-bounty-reports/
21. Latest ranking of AI companies shows that those with the best data lead: https://www.hpcwire.com/bigdatawire/2026/04/21/forbes-ai-50-list-shows-data-emerging-as-the-core-of-ai-value/
22. Don't add AI work to IT, it causes big blindspots: https://www.informationweek.com/machine-learning-ai/the-invisible-labor-crisis-inside-it-ai-work-the-org-chart-can-t-see
23. Some unusual ideas for ways to use AI: https://www.kdnuggets.com/7-specific-unconventional-things-to-do-with-language-models
24. Now AI can design improved thermoelectric generators: https://spectrum.ieee.org/ai-designed-thermoelectric-generator
25. AI controlled robots can now beat top table tennis players: https://www.theguardian.com/science/2026/apr/22/ai-powered-robot-beats-elite-table-tennis-players-milestone-robotics
26. I don't think an American AI company should be providing services to a British polic force. Especially not Palantir: https://www.theguardian.com/uk-news/2026/apr/22/met-police-talks-palantir-ai-tech-criminal-investigations-automate-intelligence
27. An Indian med student used AI to grift thousands out of MAGA using an AI-generated girl: https://arstechnica.com/tech-policy/2026/04/indian-med-student-rakes-in-thousands-with-ai-generated-maga-hottie/