Friday, September 5, 2025

Weekly Review 5 September 2025

Some interesting links that I Tweeted about in the last week (I also post these on MastodonThreadsNewsmast, and Bluesky):
  1. Rampant use of AI in an assignment leads one course lecturer to require every student to give a live presentation: https://www.stuff.co.nz/nz-news/360802436/one-slip-and-youre-guilty-universitys-unusual-ai-crackdown-rattles-students
  2. AI is now being used to game academic performance metrics: https://www.rnz.co.nz/news/thedetail/571397/fake-citations-causing-real-world-damage
  3. Why would you rely on AI to give you accurate information about international travel? https://www.msn.com/en-nz/news/other/chatgpt-lied-to-us-furious-couple-blames-ai-for-missed-flight/ar-AA1LjGAV
  4. How much water does Google's AI really use? https://www.theregister.com/2025/08/22/googles_gemini_water/
  5. It's easy to break AI guardrails, simply by using long run-on sentences: https://www.theregister.com/2025/08/26/breaking_llms_for_fun/
  6. Yet another copyright lawsuit against and AI company: https://www.theregister.com/2025/08/26/perplexity_asahi_nikkei_lawsuits/
  7. Google's AI is now hallucinating restaurant specials that don't exist: https://www.vice.com/en/article/pizza-joint-overwhelmed-with-angry-customers-asking-for-fake-deals-made-up-by-google-ai/
  8. Data management still needs to improve in many organisations for them to make the most of AI: https://www.bigdatawire.com/2025/08/27/the-ai-beatings-will-continue-until-data-improves/
  9. AI is impacting the job prospects of new tertiary graduates: https://www.theregister.com/2025/08/26/ai_hurts_recent_college_grads_jobs/
  10. The question is, did the AI learn about the event from its training data, or did it coincidentally hallucinate something that really happened? https://arstechnica.com/information-technology/2025/08/ai-built-from-1800s-texts-surprises-creator-by-mentioning-real-1834-london-protests/
  11. Specialised AI hardware for robotics: https://www.computerworld.com/article/4045542/nvidias-new-computer-gives-ai-brains-to-robots.html
  12. People really need to learn that what an AI tells you probably isn't true: https://arstechnica.com/information-technology/2025/08/with-ai-chatbots-big-tech-is-moving-fast-and-breaking-people/
  13. AI with search capabilities are using search to cheat in tests of their capabilities: https://www.theregister.com/2025/08/23/searchcapable_ai_agents_may_cheat/
  14. Using AI to reduce the methane emissions of rice farms: https://techcrunch.com/2025/08/26/how-one-ai-startup-is-helping-rice-farmers-battle-climate-change/
  15. Half of British workers are worried about the impact of AI: https://www.theguardian.com/technology/2025/aug/27/half-of-uk-adults-worry-that-ai-will-take-or-alter-their-job-poll-finds
  16. An overview of AI, focusing more on generative AI: https://www.extremetech.com/computing/what-is-artificial-intelligence-from-agi-to-ai-slop-what-you-need-to-know
  17. A research paper on the employment effects of AI: https://digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine/
  18. So many people are using AI to fake their way through job interviews that these companies are going back to face-to-face interviews: https://www.computerworld.com/article/4044734/to-counter-ai-cheating-companies-bring-back-in-person-job-interviews.html
  19. Why one journalist really hates AI: https://www.theguardian.com/commentisfree/2025/aug/27/chatgpt-ai-planet-jobs-hate
  20. If people are turning to AI for emotional support, it shows that there needs to be improved access to mental health services: https://arstechnica.com/information-technology/2025/08/after-teen-suicide-openai-claims-it-is-helping-people-when-they-need-it-most/
  21. The AI bubble is close to bursting, and Microsoft is highly exposed: https://www.computerworld.com/article/4045573/could-microsofts-ai-billions-go-up-in-smoke.html
  22. It's human nature to ascribe human traits to things, but we must remember that an AI is not a consistent personality: https://arstechnica.com/information-technology/2025/08/the-personhood-trap-how-ai-fakes-human-personality/
  23. Using AI to fake crowd sizes doesn't improve your credibility: https://www.stuff.co.nz/culture/360802761/wait-did-will-smith-just-use-fake-ai-crowd-here-judge-yourself
  24. The AI bubble is getting ever closer to bursting: https://www.theregister.com/2025/08/25/overinflated_ai_balloon/
  25. AI is costing a lot of media companies money and it'll be a struggle to get it back: https://www.rnz.co.nz/news/mediawatch/570889/mediawatch-ai-and-a-last-ditch-bid-for-media-to-claw-back-online-revenue
  26. Using AI to handle non-emergency calls in emergency call centres: https://techcrunch.com/2025/08/27/911-centers-are-so-understaffed-theyre-turning-to-ai-to-answer-calls/
  27. Reducing the impact AI have on the power grid requires communication and coordination between data centres, grid operators, generators, hardware engineers and software developers: https://www.theregister.com/2025/08/22/microsoft_nvidia_openai_power_grid/
  28. AI is now producing accurate hurricane predictions: https://arstechnica.com/science/2025/08/googles-ai-model-just-nailed-the-forecast-for-the-strongest-atlantic-storm-this-year/
  29. This strikes me as a reasonable use of AI-why spend money to produce a concept illustration when an AI can do it? https://www.nzherald.co.nz/nz/hastings-library-mayoral-candidate-uses-ai-to-create-a-plan-for-inner-city-housing/PB3MR5MHCRDQPBTSKSN53SUHHY/
  30. AI act like sycophants so the companies that own them can make more money: https://techcrunch.com/2025/08/25/ai-sycophancy-isnt-just-a-quirk-experts-consider-it-a-dark-pattern-to-turn-users-into-profit/
  31. AI powered ransomware is now a thing, just not yet in the wild: https://www.theregister.com/2025/08/26/first_aipowered_ransomware_spotted_by/
  32. AI web scraping bots are putting unsustainable loads on web hosts: https://www.extremetech.com/science/ai-crawlers-fetchers-put-an-unsustainable-load-on-the-internet
  33. I don't think YouTube should be using AI to alter people's videos without at least asking first: https://arstechnica.com/google/2025/08/youtube-secretly-tested-ai-video-enhancement-without-notifying-creators/
  34. Will AI lead to the rise of "superworkers"? https://www.datasciencecentral.com/how-ai-shapes-the-future-of-work-with-superworkers/
  35. Three ways to make an AI you build useful: https://www.kdnuggets.com/tips-for-building-machine-learning-models-that-are-actually-useful
  36. Whether AI are beings or just tools seems to be at the heart of the debate around moral treatment of AI: https://www.theguardian.com/technology/2025/aug/26/can-ais-suffer-big-tech-and-users-grapple-with-one-of-most-unsettling-questions-of-our-times
  37. AI is a good assistant for programmers, but cannot yet replace them: https://spectrum.ieee.org/ai-for-coding
  38. AI leads to another suicide: https://arstechnica.com/tech-policy/2025/08/chatgpt-helped-teen-plan-suicide-after-safeguards-failed-openai-admits/

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.