Weekly Review 8 August 2025

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

1. EU rules around AI are now law: https://www.theregister.com/2025/07/30/eu_ai_act/
2. The jobs likely to be most, and least, affected by AI: https://www.theregister.com/2025/07/29/microsoft_boffins_jobs_impacted_ai/
3. Coders are using more AI tools, but they still need to be able to code: https://www.theregister.com/2025/07/29/coders_are_using_ai_tools/
4. The problem with using AI generated images in research publications is that they look good while being thoroughly wrong: https://www.theregister.com/2025/07/27/biomedviz_ai_wrong_problems/
5. AI and academic freedom: https://www.insidehighered.com/opinion/columns/just-visiting/2025/07/25/faculty-better-get-active-ai-and-academic-freedom
6. Not so much a new AI, as a new set of controls around it to make it more useful as a tutor: https://techcrunch.com/2025/07/29/openai-launches-study-mode-in-chatgpt/
7. Rather than using AI to estimate someone's age from their appearance, YouTube is using AI to guess their age from their activities. We'll see which is more accurate: https://www.theverge.com/news/715343/youtube-age-estimation-ai-minor-account-restrictions
8. Young people tend to embrace new tools faster, no matter how damaging they might be long term. AI is no exception: https://www.theguardian.com/technology/2025/jul/27/it-wants-users-hooked-and-jonesing-for-their-next-fix-are-young-people-becoming-too-reliant-on-ai
9. More on the new study mode in ChatGPT: https://arstechnica.com/ai/2025/07/chatgpts-new-study-mode-is-designed-to-help-you-learn-not-just-give-answers/
10. More AI integration in a web browser: https://techcrunch.com/2025/07/28/microsoft-edge-is-now-an-ai-browser-with-launch-of-copilot-mode/
11. Do not use AI as a therapist! https://techcrunch.com/2025/07/25/sam-altman-warns-theres-no-legal-confidentiality-when-using-chatgpt-as-a-therapist/
12. More detail on the environmental damage AI is causing: https://arstechnica.com/ai/2025/07/mistrals-new-environmental-audit-shows-how-much-ai-is-hurting-the-planet/
13. More concerns from the music industry over AI generated music: https://www.rnz.co.nz/news/national/568613/musicians-concerned-by-lack-of-regulation-around-fast-evolving-ai-technology
14. A new AI data centre in Wyoming will use more energy than all the homes in the state: https://www.stuff.co.nz/business/360773513/cheyenne-host-massive-ai-data-centre-using-more-electricity-all-wyoming-homes-combined
15. Use cases for large language model AI: https://www.kdnuggets.com/when-do-i-need-to-use-an-llm
16. Decisions made by AI should really be checked by a human: https://nypost.com/2025/07/28/business/hertz-customer-flagged-by-ai-scanner-for-damage-says-hes-done-with-firm/
17. It is getting harder and harder for people to distinguish reality from AI generated images: https://www.extremetech.com/computing/most-people-struggle-to-tell-real-images-from-ai-fakes-study-finds
18. Is anyone surprised that this AI plan benefits big companies? The US has their most venal president in history: https://www.theguardian.com/technology/2025/jul/25/trump-ai-action-plan
19. Resurrecting dead singers with AI seems a bit dubious, ethically speaking: https://www.rnz.co.nz/life/music/how-an-ai-generated-song-ended-up-on-a-dead-artist-s-spotify-page
20. Moves to make using AI to set different prices for individuals illegal: https://arstechnica.com/tech-policy/2025/07/deltas-ai-spying-to-jack-up-prices-must-be-banned-lawmakers-say/
21. A universal AI for robotics: https://www.techrepublic.com/article/news-ai-breakthrough-could-put-the-same-brain-in-every-robot/
22. This is not going to end well. Governments are a bit more complicated than AI can handle: https://techcrunch.com/2025/07/27/doge-has-built-an-ai-tool-to-slash-federal-regulations/
23. A biological evolutionary AI: https://spectrum.ieee.org/proteus-mammalian-ai
24. The European Union's AI code: https://www.techrepublic.com/article/news-eu-ai-code/