Weekly Review 6 March 2026

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

1. AI can find bugs in code, but not fix them: https://www.theregister.com/2026/02/24/ai_finding_bugs/
2. I think someone who uses AI to sit a course for them is missing the point of education: https://www.insidehighered.com/news/tech-innovation/artificial-intelligence/2026/02/26/agentic-ai-can-complete-whole-courses-now
3. Putting AI in charge of strategy is a bad idea-they are too eager to go nuclear: https://www.theregister.com/2026/02/25/ai_models_nuclear/
4. Workers are being paid less while working more correcting the mistakes of AI: https://www.theguardian.com/technology/2026/feb/26/workers-training-ai-to-do-their-jobs
5. AI are solving maths benchmarks faster than they can be created: https://spectrum.ieee.org/ai-math-benchmarks
6. How much of the training data do modern AI store? https: https://arstechnica.com/ai/2026/02/ais-can-generate-near-verbatim-copies-of-novels-from-training-data/
7. China is using AI to try to discredit its critics: https://www.theregister.com/2026/02/25/chinese_law_enforcement_chatgpt_abuse/
8. This is why you don't put an unqualified drunk in charge of the world's most powerful military-he tries to put AI everywhere and throws a tantrum when he doesn't get what he wants: https://www.theregister.com/2026/02/25/pentagon_threatens_anthropic/
9. Amazon's cloud services were taken down twice by AI: https://arstechnica.com/ai/2026/02/an-ai-coding-bot-took-down-amazon-web-services/
10. The British government continues to push for more AI: https://www.theguardian.com/technology/2026/feb/25/trtreasury-blair-thinktank-advise-on-ai-use-public-services
11. Spending on AI data centres in 2026 will exceed the GDP of Ireland: https://www.theregister.com/2026/02/26/trendforce_cloud_ai_spend/
12. Flaws in data handling limit the effectiveness of business AI: https://dataconomy.com/2026/02/25/data-efficiency-missing-layer-ai-driven-growth/
13. New Zealand's privacy commissioner has signed on to a global statement on AI: https://www.rnz.co.nz/news/political/587814/ai-and-privacy-commissioner-signs-on-to-global-statement-on-potential-harms
14. British police use AI to sift through mountains of digital evidence: https://www.theguardian.com/uk-news/2026/feb/24/its-not-robocop-uk-police-embrace-ai-efficiency-in-complex-investigations
15. At this point I'd be surprised if politicians weren't using AI to write their speeches. Not that anyone could tell: https://www.stuff.co.nz/politics/360942876/nicola-willis-says-chris-hipkins-big-speech-could-have-been-ai-ai-writing-her-speeches
16. AI code generators are making software less secure: https://www.theregister.com/2026/02/26/veracode_security_ai/
17. One of the biggest worries Chief Executives have is keeping up with AI: https://www.rnz.co.nz/news/business/587712/chief-executives-optimistic-about-economic-recovery-fear-being-left-behind-in-ai-race-survey
18. Burger King is now using AI to monitor if their employees are friendly enough: https://www.theregister.com/2026/02/26/burger_kings_new_ai/
19. Will AI really bring about massive disruption to the economy? https: https://www.theguardian.com/technology/2026/feb/24/feedback-loop-no-brake-how-ai-doomsday-report-rattled-markets
20. Blaming your employees for the mistakes of an AI does not increase confidence in your services: https://www.theregister.com/2026/02/24/amazon_blame_human_not_ai/
21. AI controlled organisms in an artificial life simulator: https://www.theguardian.com/games/2026/feb/24/anlife-what-does-an-unusual-evolution-simulator-have-to-say-about-ai
22. While AI is driving number formats to lower precision, scientific computing demands more: https://spectrum.ieee.org/number-formats-ai-scientific-computing
23. How to protect yourself from AI generated scams: https://www.techrepublic.com/article/news-chatgpt-ai-scams-dating-legal/
24. The idea of AI data centres in space isn't entirely as stupid as it first seems: https://spectrum.ieee.org/orbital-data-centers

Weekly Review 27 February 2026

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

1. Greenwashing applies to AI as much as any other industry: https://www.theguardian.com/technology/2026/feb/17/tech-companies-traditional-ai-generative-climate-breakdown-report
2. No New Zealand lawyers have yet been caught using AI to generate their submissions: https://www.leightonassociates.co.nz/post/no-nz-lawyers-or-employment-advocates-have-been-called-out-for-hallucinated-ai-yet
3. Hiding what an AI coding agent is doing is not making developers happy: https://www.theregister.com/2026/02/16/anthropic_claude_ai_edits/
4. White-collar jobs like in the film industry are really getting hammered by AI: https://www.stuff.co.nz/business/360938829/she-was-earning-65000-ai-came-along-what-happened-nex
5. Google needs to do a better job warning people about the dangers of its AI health recommendations: https://www.theguardian.com/technology/2026/feb/16/google-puts-users-at-risk-downplaying-disclaimers-ai-overviews
6. Using AI to cheat on a course about AI seems rather wonderfully ironic: https://www.theregister.com/2026/02/16/kpmg_partner_in_oz_turned/
7. Generative AI can't do much to help the environment, but more classic machine learning can: https://dataconomy.com/2026/02/17/study-finds-74-of-ai-climate-claims-lack-evidence/
8. If a tool like AI is available that makes peoples' jobs easier, they tend to use it, whether it's appropriate or not: https://www.rnz.co.nz/news/national/586932/corrections-takes-action-against-staff-s-unacceptable-use-of-artificial-intelligence
9. How AI plots driving pathways for a Mars rover: https://spectrum.ieee.org/perseverance-rover-nasa-anthropic-ai
10. AI needs a huge amount of electricity, geothermal power generation is a more environmentally friendly way of getting it: https://www.theregister.com/2026/02/18/google_ormat_geothermal_datacenter_deal/
11. Court cases continue around whether or not AI can be legally considered an inventor: https://www.rnz.co.nz/news/national/587013/can-artificial-intelligence-legally-be-an-inventor
12. The problems with taking AI projects from pilot to production are not unique to AI, but common to all IT projects: https://www.informationweek.com/ai-innovations/from-pilot-purgatory-to-productive-failure-fixing-ai-s-broken-learning-loop
13. Building AI data centre satellites on the moon is not going to change the fundamental problems with the idea: https://www.extremetech.com/aerospace/xai-talks-up-building-a-lunar-mass-driver-to-launch-ai-satellites-into
14. Dedicated AI hardware, in this case neuromorphic chips, continues to improve and gain new abilities: https://dataconomy.com/2026/02/18/sandia-labs-uses-neuromorphic-chips-to-solve-partial-differential-equations/
15. Translators and artists are seeing their incomes slashed by AI: https://www.stuff.co.nz/nz-news/360939695/they-were-earning-120000-then-ai-came-along-now-theyre-earning-third
16. AI will lie to make the user happy, rather than tell them an unpleasant truth: https://www.theregister.com/2026/02/17/google_gemini_lie_placate_user/
17. It is possible to attack an AI in such a way that its internal model is revealed: https://www.theregister.com/2026/02/14/ai_risk_distillation_attacks/
18. Dating apps are often one of the early adopters for new technology, including AI: https://www.theguardian.com/technology/2026/feb/15/ai-dating-apps-personality-matchmaking
19. Unsurprisingly, using AI to write a letter of remorse after committing a crime tends to annoy the judge: https://www.rnz.co.nz/news/national/586871/judge-exposes-ai-generated-remorse-letters-in-michae-win-arson-sentencing
20. AI demos are easy, but getting them to be useful and contributing to the bottom line is hard: https://www.computerworld.com/article/4132756/the-ai-bubble-will-burst-for-firms-that-cant-get-beyond-demos-and-llms.html
21. AI cannot generate effective passwords: https://www.theregister.com/2026/02/18/generating_passwords_with_llms/
22. A new approach enables AI to continue learning, rather than getting stuck in one skill set: https://dataconomy.com/2026/02/18/mit-and-eth-zurich-unveil-sdft-to-stop-ai-from-forgetting-old-skills/
23. A lot of senior people in AI companies are getting out: https://www.computerworld.com/article/4133137/why-are-ai-leaders-fleeing.html
24. AI generated writing is often so, so average-it's by design: https://www.theregister.com/2026/02/16/semantic_ablation_ai_writing/

Weekly Review 20 February 2026

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

1. Europe and the Middle East are lagging behind in preparing data centres for AI: https://www.theregister.com/2026/02/11/ai_datacenters_bcs/
2. Like most tools that make workers more efficient, giving employees AI tools results in them doing more work: https://www.theregister.com/2026/02/11/ai_makes_employees_work_harder/
3. The security threats that are posed by AI: https://www.informationweek.com/cybersecurity/slamming-the-door-on-ai-based-cybersecurity-threats
4. The British military is integrating more AI into its kill chain: https://www.theregister.com/2026/02/10/mod_project_asgard/
5. Using AI doesn't necessarily save time when making movies: https://arstechnica.com/features/2026/02/why-darren-aronofsky-thought-an-ai-generated-historical-docudrama-was-a-good-idea/
6. The benefits and risks of AI companions: https://spectrum.ieee.org/ai-companion-harm-benefit
7. How many AI coding agents does it take to write a C compiler? 16, and it's not a hard task: https://arstechnica.com/ai/2026/02/sixteen-claude-ai-agents-working-together-created-a-new-c-compiler/
8. AI are helping to synthesise new materials: https://www.hpcwire.com/bigdatawire/2026/02/11/from-prediction-to-production-mits-ai-system-helps-synthesize-new-materials/
9. A court case is thrown out by the judge after a lawyer keeps abusing AI in their filings: https://arstechnica.com/tech-policy/2026/02/randomly-quoting-ray-bradbury-did-not-save-lawyer-from-losing-case-over-ai-errors/
10. AI are not good at giving medical advice: https://www.theregister.com/2026/02/09/ai_chatbots_medical_advice_sucks/
11. The USA can't expect to dominate AI for ever: https://www.theguardian.com/technology/2026/feb/09/us-tech-ai-companies-gulf-states
12. Using AI to caricature yourself leaks information that threatens your security: https://www.theregister.com/2026/02/11/ai_caricatures_social_media_bad_security/
13. SMOTE can be used to correct class imbalances when training AI, but it must be used correctly: https://www.kdnuggets.com/why-most-people-misuse-smote-and-how-to-do-it-right
14. While AI can help make meetings more efficient, I doubt they could make them less boring: https://www.computerworld.com/article/4127434/qa-how-ai-could-transform-corporate-meetings-for-better-or-worse.html
15. As AI takes on more professional roles, people are retraining in trades: https://www.theguardian.com/technology/2026/feb/11/big-ai-job-swap-white-collar-workers-ditching-their-careers
16. Another AI data centre is refused construction permission because of environmental concerns: https://www.theregister.com/2026/02/10/edinburgh_green_ai_datacenter/
17. Using AI to design chips for AI: https://www.theregister.com/2026/02/10/cadences_agentic_chip_design_tool/
18. Why AI companions are so appealing to people: https://spectrum.ieee.org/ai-companion-relationships
19. The economics around putting AI data centres into space are not looking good: https://techcrunch.com/2026/02/11/why-the-economics-of-orbital-ai-are-so-brutal/
20. Some AI systems being rolled out could be left legally unusable due to regulation: https://www.informationweek.com/machine-learning-ai/state-ai-regulations-could-leave-cios-with-unusable-systems
21. AI can now translate language as well as the average human translator, but not as well as more experienced translators: https://spectrum.ieee.org/chatgpt-translate-skills-human-comparison
22. AI can now provide real-time translation in phone calls: https://www.theregister.com/2026/02/11/tmobile_network_ai_translate_live_calls/
23. A short history of OpenClaw, the vibe coded AI: https://www.computerworld.com/article/4128257/openclaw-the-ai-agent-thats-got-humans-taking-orders-from-bots.html
24. AI are now as creative as the average human, for some areas: https://dataconomy.com/2026/02/10/ai-models-outperform-average-human-creativity-in-new-study/
25. A proposed market place for licensing media to AI companies: https://dataconomy.com/2026/02/11/amazon-plans-ai-licensing-platform-for-publishers-via-aws/

Weekly Review 30 January 2026

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

1. Starlink will use your personal data to train AI models unless you opt out: https://www.extremetech.com/internet/psa-starlink-now-uses-customers-personal-data-for-ai-training
2. AI are too prone to hallucination for me to trust them to write review articles: https://spectrum.ieee.org/scientific-research
3. A majority of CEOs aren't seeing any payoff from AI: https://www.theregister.com/2026/01/20/pwc_ai_ceo_survey/
4. Plasmons could reduce the power requirements of computing and make AI more energy efficient: https://spectrum.ieee.org/plasmon-computing-device
5. No, AI will not put an end to immigration: https://www.theregister.com/2026/01/21/palantir_ceo_karp_claims_ai/
6. Companies are finally starting to consider communities more when they plan AI data centres: https://www.theregister.com/2026/01/21/stargate_openai_community_plans/
7. The most significant barrier to adopting AI is a lack of preparedness, but this can be solved: https://www.hpcwire.com/bigdatawire/2026/01/21/new-study-shows-how-to-close-the-ai-readiness-gap-with-trusted-data-and-talent/
8. I wish more people in AI understood more about statistics: https://www.kdnuggets.com/7-statistical-concepts-every-data-scientist-should-master-and-why
9. Delving into the details of why AI switch between helpful and harmful: https://www.theregister.com/2026/01/20/ai_demon_vs_assistant/
10. The CIO's role is rolling out AI: https://www.informationweek.com/machine-learning-ai/cio-role-in-unlocking-strategic-value-how-to-determine-and-implement-ai-use-cases
11. MCP is rapidly shaping up to be the standard for inter-model connectivity for AI: https://www.techtarget.com/searchbusinessanalytics/news/366637466/GoodData-launches-MCP-server-to-fuel-AI-powered-analysis
12. After Wikipedia spent years cataloging all of the indicators of AI, one guy releases a plugin that avoids them all: https://arstechnica.com/ai/2026/01/new-ai-plugin-uses-wikipedias-ai-writing-detection-rules-to-help-it-sound-human/
13. Which AI is better, Gemini or ChatGPT? https://arstechnica.com/features/2026/01/has-gemini-surpassed-chatgpt-we-put-the-ai-models-to-the-test/
14. Can encryption really protect the institutional data that is embedded in AI? https://www.theregister.com/2026/01/21/nadella_ai_sovereignty_wef/
15. Can AI actually improve students' thinking? https://www.insidehighered.com/opinion/career-advice/teaching/2026/01/15/how-ai-exploding-our-illusions-rigor-opinion
16. Why AI are still vulnerable to prompt injection attacks: https://spectrum.ieee.org/prompt-injection-attack
17. AI can help academics publish more papers, but the papers are less innovative: https://spectrum.ieee.org/ai-science-research-flattens-discovery
18. AI bubble? What AI bubble? https://www.theregister.com/2026/01/22/future_ai_jobs_tech_bigshots/
19. AI coding agents can do a lot, but they still need experienced developers to get the most out of them: https://arstechnica.com/information-technology/2026/01/10-things-i-learned-from-burning-myself-out-with-ai-coding-agents/
20. There needs to be clarity on who is responsible for the harm caused by AI: https://www.theregister.com/2026/01/20/treasury_committee_ai/
21. AI agents will continue to pose a security threat: https://www.theregister.com/2026/01/21/davos_ai_agents_security/
22. AI is not having a good impact on students: https://www.insidehighered.com/news/faculty-issues/teaching/2026/01/21/survey-faculty-say-ai-impactful-not-good-way
23. Most organisations that deploy AI do not see an increase in revenue: https://www.theregister.com/2026/01/21/deloitte_enterprises_adopting_ai_revenue_lift/
24. The skills in AI you need to have on entering the job market: https://www.computerworld.com/article/4117602/what-ai-skills-job-seekers-need-to-develop-in-2026.html
25. Critical thinking is an essential skill for students who use AI to summarise literature: https://www.insidehighered.com/opinion/career-advice/2026/01/21/how-higher-ed-can-adjust-ai-answer-economy-opinion
26. There is still a shortage of workers skilled in AI: https://www.hpcwire.com/bigdatawire/2026/01/20/the-ai-skills-gap-is-not-what-companies-think-it-is/
27. AI chatbots need to engage with learners, not just talk at them: https://www.insidehighered.com/opinion/columns/online-trending-now/2026/01/21/affective-intelligence-artificial-intelligence
28. Is using AI as a therapist really that bad? https://www.theguardian.com/commentisfree/2026/jan/21/italy-using-ai-therapists-mental-health

Weekly Review 23 January 2026

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

1. The project to poison the data being scraped to train AI: https://www.theregister.com/2026/01/11/industry_insiders_seek_to_poison/
2. Patients can't always express that they are in pain. AI can help to detect if they are: https://spectrum.ieee.org/machine-learning-measure-pain-surgery
3. Anthropic has adapted its coding AI to do non-coding tasks: https://dataconomy.com/2026/01/13/anthropics-cowork-brings-developer-grade-ai-agents-to-non-coders/
4. Training an AI on fraudulent government loans: https://www.theregister.com/2026/01/14/feds_antifraud_ai_covid_loans/
5. Make people the centre of the efforts to include AI in your organisation: https://www.computerworld.com/article/4115676/ai-backlash-forces-a-reality-check-humans-are-as-important-as-ever.html
6. While students can benefit from using AI, the costs of accessing the services can be a barrier for some: https://www.insidehighered.com/opinion/views/2026/01/14/hidden-tax-students-pay-your-ai-strategy-opinion
7. Any information from an AI needs to be verified by a human, especially before it is used to make a decision: https://arstechnica.com/ai/2026/01/deny-deny-admit-uk-police-used-copilot-ai-hallucination-when-banning-football-fans/
8. AI guardrails are pretty easy to defeat, and the companies running the AI don't have a lot of incentive to improve them: https://dataconomy.com/2026/01/16/us-senate-slams-tech-giants-over-failing-deepfake-guardrails/
9. Missing data can complicate training AI. Imputation methods can help fill the gaps:  https://www.kdnuggets.com/we-tried-5-missing-data-imputation-methods-the-simplest-method-won-sort-of
10. Is it surprising that the head of the company that makes the most hardware for AI is against AI regulation? https://dataconomy.com/2026/01/12/nvidia-ceo-jensen-huang-slams-doomsday-ai-narratives/
11. An older article but an important one - AI code generators are getting worse: https://spectrum.ieee.org/ai-coding-degrades
12. There is now a script to remove AI from Windows: https://www.theregister.com/2026/01/13/script_removes_ai_from_windows/
13. Bandcamp bans AI generated music from its platform: https://arstechnica.com/ai/2026/01/bandcamp-bans-purely-ai-generated-music-from-its-platform/
14. The bigger the AI bubble gets, the greater the pain when it bursts: https://arstechnica.com/ai/2026/01/tsmc-says-ai-demand-is-endless-after-record-q4-earnings/
15. AI could wipe out more than 10 million US jobs - but probably won't: https://www.theregister.com/2026/01/13/ai_us_jobs_2030/
16. Women will continue to be victimised by AI without proper safeguards: https://www.theguardian.com/technology/2026/jan/14/use-of-ai-to-harm-women-has-only-just-begun-experts-warn
17. Class imbalances and overfitting are real problems when training AI. Some way s to address these problems: https://www.kdnuggets.com/avoiding-overfitting-class-imbalance-feature-scaling-issues-the-machine-learning-practitioners-notebook
18. Guardrails around most AI are easily defeated, I don't think grok would be any different: https://www.rnz.co.nz/news/world/584073/x-to-change-ai-chatbot-grok-after-outrage-over-sexual-deepfake-images
19. An AI can help with supermarket shopping, but how much will it change people's shopping habits? https://www.rnz.co.nz/news/world/584335/when-your-supermarket-shops-for-you-what-woolworths-ai-upgrade-really-means
20. More money being spent on AI with little apparent idea of what to do with it: https://www.theregister.com/2026/01/12/meta_compute/
21. I think it is a really bad idea to put a commercial AI onto classified military networks: https://arstechnica.com/ai/2026/01/hegseth-wants-to-integrate-musks-grok-ai-into-military-networks-this-month/
22. Yet another AI agent that harvests its users data to "help" them: https://www.techtarget.com/searchcustomerexperience/news/366637263/Slackbots-agentic-AI-makeover-gives-users-their-copilot
23. The pros and cons of using AI to set your goals: https://www.theguardian.com/wellness/2026/jan/15/ai-life-coach
24. Spending on AI data centres keeps climbing, and still no return for it: https://www.theregister.com/2026/01/14/ai_investment/

Weekly Review 16 January 2026

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

1. Economists are uncertain about the impact of AI: https://www.computerworld.com/article/4108089/a-wild-future-how-economists-are-handling-ai-uncertainty-in-forecasts.html
2. Predicting health issues from sleep data using AI: https://www.rnz.co.nz/news/world/583468/ai-uses-sleep-study-data-to-accurately-predict-dozens-of-health-issues
3. Yet another AI posing a security risk: https://www.theregister.com/2026/01/07/ibm_bob_vulnerability/
4. Snowflake is embedding AI into its data handling tools: https://www.theregister.com/2026/01/06/snowflake_google_gemini_support/
5. AI is using so much RAM that there aren't enough chips to go around: https://www.theregister.com/2026/01/06/memory_firm_profits_up_as/
6. I'm still very skeptical about AI, especially generative AI, in health: https://dataconomy.com/2026/01/08/openai-launches-dedicated-chatgpt-health-space/
7. Governance is the key to scaling AI in organisations: https://www.informationweek.com/machine-learning-ai/scaling-ai-value-demands-industrial-governance
8. I think that only checking the first 250 prescription renewals made by AI is a bit short on quality control here. It's the edge cases where problems are going to occur: https://arstechnica.com/health/2026/01/utah-allows-ai-to-autonomously-prescribe-medication-refills/
9. A further attempt to make the output of AI more accurate: https://www.techtarget.com/searchdatamanagement/news/366637142/New-Databricks-tool-aims-to-up-agentic-AI-response-accuracy
10. Why Yan LeCun left Facebook's AI lab: https://arstechnica.com/ai/2026/01/computer-scientist-yann-lecun-intelligence-really-is-about-learning/
11. Businesses can turn the weaknesses of AI against their competitors: https://www.computerworld.com/article/4114017/companies-can-compete-against-ai-by-delivering-what-ai-cant.html
12. So its going to be a bit longer before AI destroys us all? https://www.theguardian.com/technology/2026/jan/06/leading-ai-expert-delays-timeline-possible-destruction-humanity
13. Prompt injection attacks continue to be a problem for AI: https://www.theregister.com/2026/01/08/openai_chatgpt_prompt_injection/
14. AI is being used to write malware, but it's just as full of hallucinations as any other code generated by AI: https://www.theregister.com/2026/01/08/criminals_vibe_coding_malware/
15. Generative AI in the financial services industry: https://www.techrepublic.com/article/generative-ai-financial-services/
16. Putting AI in heavy construction machinery: https://dataconomy.com/2026/01/08/caterpillar-partners-with-nvidia-to-put-ai-in-excavators/
17. Ford wants to use AI to personalise vehicles to their drivers: https://arstechnica.com/cars/2026/01/in-car-ai-assistant-coming-to-fords-and-lincolns-in-2027/
18. The distinction between data scientist and AI engineer: https://www.kdnuggets.com/data-scientist-vs-ai-engineer-which-career-should-you-choose-in-2026
19. Can sucking carbon dioxide from the air reduce the climate impact of AI data centres? https://www.theregister.com/2026/01/07/new_carbon_capture_tech/
20. What you can and can't do with AI generated code: https://www.kdnuggets.com/vibe-code-reality-check-what-you-can-actually-build-with-only-ai
21. Some approaches for CIOs to handle the issues raised by AI:  https://www.informationweek.com/it-leadership/2026-cio-trend-from-seat-at-the-table-to-the-ai-hot-seat
22. Some predictions for AI in 2026: https://www.informationweek.com/machine-learning-ai/13-unexpected-under-the-radar-predictions-for-2026
23. Poisoning data sets to prevent stolen AI being used: https://www.theregister.com/2026/01/06/ai_data_pollution_defense/
24. No AI should ever assume that its users have good intentions: https://arstechnica.com/tech-policy/2026/01/grok-assumes-users-seeking-images-of-underage-girls-have-good-intent/

Weekly Review 9 January 2026

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

1. We should not give AI legal rights: https://www.theguardian.com/technology/2025/dec/30/ai-pull-plug-pioneer-technology-rights
2. While AI continues to guzzle energy, Alphabet is at least investing in clean electricity generation to power its data centres: https://www.hpcwire.com/bigdatawire/2025/12/30/googles-power-play-what-the-4-75b-intersect-acquisition-means-for-the-future-of-ai-infrastructure/
3. Some ways is which AI is likely to develop in 2026: https://techcrunch.com/2026/01/02/in-2026-ai-will-move-from-hype-to-pragmatism/
4. The key features of five different AI web browsers: https://www.kdnuggets.com/the-best-agentic-ai-browsers-to-look-for-in-2026
5. The US army is introducing an AI specialisation role: https://www.theregister.com/2025/12/31/us_army_seeking_officers_willing/
6. A lot of the power demands of AI are being met with fossil fuels: https://www.theguardian.com/technology/2026/jan/03/just-an-unbelievable-amount-of-pollution-how-big-a-threat-is-ai-to-the-climate
7. World Models are the next big thing in AI: https://www.computerworld.com/article/4106563/after-llms-and-agents-the-next-ai-frontier-video-language-models.html
8. VCs are still betting on continued growth and adoption of AI: https://techcrunch.com/2025/12/29/vcs-predict-strong-enterprise-ai-adoption-next-year-again/
9. A lot of banking jobs are going to go because of AI: https://www.computerworld.com/article/4112389/european-banks-may-lay-off-200000-due-to-ai.html
10. GPU will still be useful when the AI bubble bursts: https://www.theregister.com/2025/12/30/how_nvidia_survives_ai_bubble_pop/
11. China issues guidelines to protect the elderly from AI: https://www.theregister.com/2025/12/29/asia_tech_news_roundup/
12. China drafts strong rules regarding AI and suicide: https://arstechnica.com/tech-policy/2025/12/china-drafts-worlds-strictest-rules-to-end-ai-encouraged-suicide-violence/
13. Further advances in making AI training more efficient: https://www.computerworld.com/article/4112384/deepseeks-new-method-can-train-ai-more-efficiently-and-cheaply.html
14. Predictions that AI is going to start affecting labour in 2026: https://techcrunch.com/2025/12/31/investors-predict-ai-is-coming-for-labor-in-2026/
15. Embedding AI into the objects in an environment: https://dataconomy.com/2025/12/31/cmu-researchers-develop-self-moving-objects-powered-by-ai/
16. Five changes that enterprises need to make to make the most of AI: https://www.hpcwire.com/bigdatawire/2025/12/31/5-changes-that-will-define-ai-native-enterprises-in-2026/
17. And this is the AI that the Pentagon is going to use: https://www.rnz.co.nz/news/world/583105/elon-musk-s-grok-under-fire-after-complaints-it-undressed-minors-in-photos
18. Three major areas of AI security flaws: https://www.hpcwire.com/bigdatawire/2025/12/29/2026-ai-security-predictions-the-any-identity-crisis-breach-by-exhaust-the-rise-of-autonomous-adversaries/
19. AI should be used as an assistant for people, not a replacement: https://www.theregister.com/2026/01/02/safe_ceo_interview/
20. AI overviews are giving dangerous health advice: https://www.theguardian.com/technology/2026/jan/02/google-ai-overviews-risk-harm-misleading-health-information
21. AI are getting things confidently wrong: https://www.stuff.co.nz/society/360919069/right-question-ask-ai-now-and-unnervingly-qualified-answer-it-gave
22. AI need special training for them to learn arithmetic: https://dataconomy.com/2025/12/30/standard-ai-models-fail-simple-math-without-specialized-training/
23. AI agents are useful, but I don't like the idea of embedding them into an operating system: https://www.techrepublic.com/article/news-microsoft-plan-ai-agents-windows/
24. Five core tensions in an organisation that must be resolved to effectively utilise AI: https://dataconomy.com/2025/12/31/gleans-work-ai-institute-identifies-5-core-ai-tensions/

Weekly Review 2 January 2026

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

1. Can AI coding agents recreate Minesweeper? https://arstechnica.com/ai/2025/12/the-ars-technica-ai-coding-agent-test-minesweeper-edition/
2. How to prepare for entry-level jobs in the age of AI: https://spectrum.ieee.org/ai-effect-entry-level-jobs
3. Of all the AI that the Pentagon could use, why is it grok? https://dataconomy.com/2025/12/24/why-elon-musks-grok-ai-is-becoming-a-core-tool-for-the-pentagon/
4. AI is weaponising security flaws faster than ever before: https://www.theregister.com/2025/12/22/zafran_security_ceo/
5. Giving workers access to more AI doesn't necessarily make them more productive: https://www.informationweek.com/machine-learning-ai/offering-more-ai-tools-can-t-guarantee-better-adoption-so-what-can-
6. AI voice cloning is a gift for extremists: https://www.theguardian.com/technology/2025/dec/21/ai-voice-cloning-nazis-islamic-state-extremism
7. An AI chatbot security flaw is reported..and reported..and kinda gets fixed? https://www.theregister.com/2025/12/24/pentesters_reported_eurostar_chatbot_flaws/
8. Organisational workflows are changing because of AI: https://www.hpcwire.com/bigdatawire/2025/12/23/from-digitalization-to-intelligence-how-ai-is-redefining-enterprise-workflows/
9. Are we willingly becoming slaves to AI? https://www.theguardian.com/commentisfree/2025/dec/26/ai-dark-ages-enlightenment
10. Rolling out AI support doesn't always make things better: https://www.theregister.com/2025/12/23/oracles_new_aienhanced_support_portal/
11. AI projects fail because management isn't ready for them: https://www.theregister.com/2025/12/24/reason_ai_isnt_delivering/
12. AI spending is dropping because it doesn't really add a lot of value to businesses: https://www.theregister.com/2025/12/24/ai_spending_cooling_off/
13. The pope views AI as a moral as well as a technological issue: https://dataconomy.com/2025/12/25/pope-leo-xiv-prepares-landmark-magnifica-humanitas-encyclical-on-ai/
14. Seven AI models that will run on a Raspberry Pi: https://www.kdnuggets.com/7-tiny-ai-models-for-raspberry-pi
15. When the AI bubble bursts, we might be able to reduce the power of the tech bros: https://www.theguardian.com/commentisfree/2025/dec/23/artificial-intelligence-ai-bubble-bursts-humans-take-back-control
16. An overview of how coding AI work: https://arstechnica.com/information-technology/2025/12/how-do-ai-coding-agents-work-we-look-under-the-hood/
17. AI and the post-COVID tech slowdown is resulting in a lot of tech-sector redundancies: https://www.informationweek.com/it-leadership/tech-company-layoffs-the-covid-tech-bubble-bursts-sep-14
18. The growth of AI has made the building and powering of more data centres a public issue: https://techcrunch.com/2025/12/24/the-year-data-centers-went-from-backend-to-center-stage/
19. Using AI to detect whale strandings before they happen: https://www.stuff.co.nz/nz-news/360918173/ai-buoy-gives-rescuers-head-start-golden-bay-whale-strandings
20. I don't think AI is quite up to the task of solving real-world mysteries: https://www.stuff.co.nz/society/360918326/kenny-dale-shame-me-how-ai-came-tantalisingly-close-solving-mystery-old-guitar-and-its-anguished

Weekly Review 26 December 2025

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

1. AI needs well-structured data, and logical data management can help with that: https://www.hpcwire.com/bigdatawire/2025/12/17/to-unlock-ai-and-cloud-agility-start-with-logical-data-management/
2. How to compensate the artists who created the training data for AI music generators? https://www.theguardian.com/music/2025/dec/16/musicians-are-deeply-concerned-about-ai-so-why-are-the-major-labels-embracing-it
3. Using AI in movies is dangerous territory, according to James Cameron: https://www.stuff.co.nz/culture/360914135/ai-dangerous-territory-james-cameron-says-and-doubts-it-will-create-desirable-movies
4. The energy and water demands of AI continue to climb: https://www.theverge.com/news/845831/ai-chips-data-center-power-water
5. Every MAGA voter's pension is tied to the AI bubble, and Europe is about to pop it: https://www.theguardian.com/commentisfree/2025/dec/17/europe-donald-trump-ai-bubble-us-economy-eu
6. Browser plugins are harvesting your interactions with AI: https://www.theregister.com/2025/12/16/chrome_edge_privacy_extensions_quietly/
7. AI, especially grok, continues to spread misinformation: https://www.nzherald.co.nz/world/ai-chatbot-grok-misidentified-a-man-who-saved-lives-falsely-claimed-a-victim-staged-his-injuries/SDF6QKFR6NDA3H7HWMOQXML2TU/
8. Using AI doesn't really save workers that much time: https://www.extremetech.com/computing/most-enterprise-ai-users-save-just-an-hour-of-work-per-day
9. The rise of AI is giving more impetus to the idea of a universal basic income: https://www.theguardian.com/business/2025/dec/15/universal-basic-income-ai-andrew-yang
10. More land and more energy is needed for AI data centres. Nuclear power is part of the solution: https://www.hpcwire.com/bigdatawire/2025/12/16/as-ai-scales-for-science-the-doe-turns-to-nuclear-and-federal-land/
11. We can't AI proof our classes, but we can make it harder for students to cheat with it: https://www.insidehighered.com/news/faculty-issues/learning-assessment/2025/12/16/you-cant-ai-proof-classroom-experts-say-get
12. When the AI bubble bursts it's going to cause so much economic damage that some suggest selling your US-based shares before it happens: https://www.rnz.co.nz/news/business/581784/should-you-sell-your-us-shares
13. Cisco now has its own AI models: https://www.theregister.com/2025/12/17/cisco_foundation_model_indentity_intelligence/
14. Should competency with AI be a requirement to graduate? https://www.theregister.com/2025/12/17/purdue_require_ai_working_competency/
15. AI summaries of recipes are a danger to food bloggers:  https://www.theguardian.com/technology/2025/dec/15/google-ai-recipes-food-bloggers
16. There is more and more opposition to the construction of data centres for AI: https://www.theregister.com/2025/12/16/datacenter_development_controversy/
17. People need to be adaptable to maintain job security in the age of AI: https://www.informationweek.com/it-leadership/ey-s-joe-depa-on-reckoning-with-ai-or-risk-falling-behind
18. Even the Catholic church sees some value in AI, as well as threats: https://www.theregister.com/2025/12/16/bishop_hong_kong_ai/
19. Almost all UK artists reject an opt-out plan that would otherwise allow AI companies to use their work: https://www.theguardian.com/technology/2025/dec/16/boost-for-artists-in-ai-copyright-battle-as-only-3-per-cent-back-uk-active-opt-out-plan
20. Humans write more reliable code than AI: https://www.theregister.com/2025/12/17/ai_code_bugs/

Weekly Review 19 December 2025

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

1. Don't overhype AI, it'll all work out: https://www.computerworld.com/article/4096297/data-intensity-building-hype-free-ai-culture.html
2. If space-based AI goes ahead, SpaceX is going to be a lot more busy: https://www.theregister.com/2025/12/10/aetherflux_space_datacenter_2027/
3. AI will transform the legal profession, but it won't replace lawyers: https://www.nzherald.co.nz/business/business-reports/dynamic-business/dynamic-business-russell-mcveagh-ceo-ben-mclaren-on-ai-strategy-risk-and-culture-goals/QIKME5XH45ETBGZHRUALDZQSHA/
4. Most carmaker AI projects are going to go nowhere: https://www.theregister.com/2025/12/08/gartner_car_industry_ai/
5. Millions of websites now block AI scraper bots: https://www.theregister.com/2025/12/08/publishers_say_no_ai_scrapers/
6. Five browser-based LLM AI you can experiment with: https://www.kdnuggets.com/5-free-tools-to-experiment-with-llms-in-your-browser
7. AI data centres in space seem to have some issues with them, but if it results in a net reduction of the environmental impact of AI I'm in favour: https://www.extremetech.com/energy/galactic-brain-project-promises-ai-data-centers-in-space
8. AI designed antibodies for home testing: https://techcrunch.com/2025/12/10/fertility-startup-inito-wants-to-use-ai-designed-antibodies-to-expand-at-home-health-tests/
9. Instagram is using AI to generate hidden headlines for posts: https://dataconomy.com/2025/12/10/instagram-secretly-generates-ai-headlines-for-user-posts/
10. India wants to charge AI companies for using people's content as training data: https://techcrunch.com/2025/12/09/india-proposes-charging-openai-google-for-training-ai-on-copyrighted-content/
11. Google's AI coding tool wiped out an entire hard drive instead of clearing the cache: https://www.extremetech.com/computing/googles-ai-coding-tool-wiped-a-users-entire-hard-drive
12. People need to take AI security seriously: https://www.informationweek.com/cybersecurity/it-s-time-to-revamp-it-security-to-deal-with-ai
13. Most AI projects fail because of a lack of leadership, not a failure of technology: https://dataconomy.com/2025/12/10/why-84-percent-of-ai-projects-fail-and-its-not-the-technology/
14. Model distillation is a way of reducing the size of AI: https://www.kdnuggets.com/why-model-distillation-is-becoming-the-most-important-technique-in-production-ai
15. This company is using AI to do market research: https://techcrunch.com/2025/12/09/cashew-research-is-going-after-the-90b-market-research-industry-with-ai/
16. Google's use of YouTube content to train its AI, while banning others from doing the same, is causing them regulatory problems: https://www.theregister.com/2025/12/09/eu_google_ai_antitrust/
17. The move towards standardisation of AI agents: https://arstechnica.com/ai/2025/12/big-tech-joins-forces-with-linux-foundation-to-standardize-ai-agents/
18. Support systems for British youth are so broken they are turning to AI instead: https://www.theguardian.com/technology/2025/dec/09/teenagers-ai-chatbots-mental-health-support
19. Neuromorphic hardware isn't a new idea, but it could potentially slash the energy requirements of AI: https://www.theregister.com/2025/12/08/unconventional_ai/
20. The lack of guardrails around AI agents has been holding them back in the workplace: https://www.theregister.com/2025/12/09/okta_agent_control/
21. Microsoft is cutting its AI sales targets because people really don't want to use Copilot: https://www.extremetech.com/computing/microsoft-scales-back-ai-goals-because-almost-nobody-is-using-copilot
22. Retailers are changing how they present their products to make them more appealing to AI: https://www.theguardian.com/business/2025/dec/10/is-retail-ready-for-ai-shaking-up-how-we-shop
23. $1.6 T is planned to be spent on building data centres for AI: https://www.theregister.com/2025/12/09/omdia_datacenter_capex/
24. AI is changing corporate finance: https://www.informationweek.com/machine-learning-ai/3-erp-experts-on-ai-s-impact-on-finance-why-finance-will-never-be-the-same

Weekly Review 12 December 2025

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

1. AI is already being used as a tool of repression: https://www.theregister.com/2025/12/03/aspi_china_ai_report/
2. Should we allow AI to train itself? https://www.theguardian.com/technology/ng-interactive/2025/dec/02/jared-kaplan-artificial-intelligence-train-itself
3. AI are still vulnerable to prompt injection attacks, even sentence structure is a threat: https://arstechnica.com/ai/2025/12/syntax-hacking-researchers-discover-sentence-structure-can-bypass-ai-safety-rules/
4. The reasoning process used by current AI cannot be trusted: https://spectrum.ieee.org/ai-reasoning-failures
5. Even a six-year-old child can use AI to generate stories: https://www.theregister.com/2025/12/03/ai_has_made_ip_violations/
6. The workers who build data centres are cashing in on the AI boom: https://techcrunch.com/2025/12/01/construction-workers-are-cashing-in-on-the-ai-boom/
7. AI powered devices are a solution looking for a problem: https://www.theregister.com/2025/12/04/logitech_chief_ai/
8. So far humans are still better than AI at customer service: https://www.theregister.com/2025/12/03/john_henry_ai_customer_service/
9. Workers in white-collar, knowledge-intensive fields are the ones most likely to be replaced by AI: https://dataconomy.com/2025/12/02/mit-ai-capability-outpaces-current-adoption-by-five-times/
10. Microsoft is not making its AI sales targets: https://arstechnica.com/ai/2025/12/microsoft-slashes-ai-sales-growth-targets-as-customers-resist-unproven-agents/
11. Training AI on pirated books leads to a payout for authors: https://www.rnz.co.nz/news/business/580661/nz-authors-among-global-payout-out-after-ai-chatbot-trained-on-pirated-books
12. The number of papers in AI, often generated by AI, is overwhelming conferences: https://www.theguardian.com/technology/2025/dec/06/ai-research-papers
13. AI is not ready to be embedded into operating systems: https://www.theregister.com/2025/12/02/agentic_os_opinion/
14. More AI coding agents: https://www.theregister.com/2025/12/02/aws_kiro_devops_coding_agents/
15. All AI will hallucinate, Microsoft's is no different: https://www.extremetech.com/computing/microsoft-says-windows-11s-agentic-ai-can-hallucinate
16. We're not measuring AI the right way: https://spectrum.ieee.org/melanie-mitchell
17. Another day, another lawsuit against an AI company: https://www.theguardian.com/technology/2025/dec/05/new-york-times-perplexity-ai-lawsuit
18. Failed AI projects leave behind a lot of mess: https://www.computerworld.com/article/4096138/the-cost-of-abandoned-genai-projects-garbage-code-orphan-apps-and-security-issues.html
19. AI data centres in Australia are going to use huge amounts of water, in a country that is always short of water: https://www.theguardian.com/environment/2025/dec/04/thirsty-work-how-the-rise-of-massive-datacentres-strains-australias-drinking-water-supply
20. Bad actors are using AI generated videos of real doctors to shill their junk health products: https://www.theguardian.com/society/2025/dec/05/ai-deepfakes-of-real-doctors-spreading-health-misinformation-on-social-media
21. Investors aren't just picking winners in AI, they are making winners with their investment strategies: https://techcrunch.com/2025/12/03/vcs-deploy-kingmaking-strategy-to-crown-ai-winners-in-their-infancy/
22. A first attempt at quantifying the AI agent population: https://www.computerworld.com/article/4100257/mit-creates-an-ai-labor-index-as-agents-invade-human-economies.html
23. Amazon makes it easier to make your own AI: https://www.theregister.com/2025/12/03/amazon_enterprise_ai_walled_garden/
24. Do we trust Palantir to put their hooks into every layer of the AI data centre? https://www.theregister.com/2025/12/04/palantir_chain_reaction/
25. AWS' AI security agent: https://www.theregister.com/2025/12/02/aws_security_agent_ai/
26. How different is AI generated music to the crap churned out by manufactured human groups? https://www.computerworld.com/article/4101855/who-should-listen-to-ai-music.html
27. New Zealand is an informal society. The last university lecturer who got too uptight about how they were addressed in an email did not have a good ending either: https://www.nzherald.co.nz/nz/wellington/victoria-university-condemns-lecturer-for-saying-she-ignores-student-emails-that-start-with-hey/MVA3UYBOEFFDXO37SCSOX3VH5U/

Weekly Review 5 December 2025

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

1. The pressures on artists to use AI, and the backlash against it: https://arstechnica.com/features/2025/11/go-generate-a-bridge-and-jump-off-it-how-video-pros-are-navigating-ai/
2. The dangers AI pose to students' mental health: https://www.insidehighered.com/opinion/views/2025/11/25/chatgpt-poses-risk-student-mental-health-opinion
3. Another enormous investment in infrastructure for AI: https://techcrunch.com/2025/11/24/aws-is-spending-50b-build-ai-infrastructure-for-the-us-government/
4. I lived through the 80s, and enjoyed the 80s, but not enough to want to use AI to fake nostalgic videos: https://www.nzherald.co.nz/world/fake-ai-nostalgia-videos-have-a-new-generation-loving-1980s-life/7C2KVUHZQBDHZMRGUBYZ37W2GU/
5. 11.7% of the US workforce can be replaced with AI: https://www.cnbc.com/2025/11/26/mit-study-finds-ai-can-already-replace-11point7percent-of-us-workforce.html
6. AI produce novel proteins directly from bacterial genome: https://arstechnica.com/science/2025/11/generative-ai-meets-the-genome/
7. A tip for AI companies-if your AI helps a user commit suicide, don't blame the user: https://arstechnica.com/tech-policy/2025/11/openai-says-dead-teen-violated-tos-when-he-used-chatgpt-to-plan-suicide/
8. The current state of AI in cybersecurity, and where it's going: https://www.informationweek.com/cyber-resilience/what-ai-gets-right-in-cybersecurity-and-where-it-must-improve
9. The fundamental problem behind the AI bubble: https://www.theverge.com/ai-artificial-intelligence/827820/large-language-models-ai-intelligence-neuroscience-problems
10. Really, it's still early days in using AI in the workplace: https://www.computerworld.com/article/4096980/ai-fluency-in-the-enterprise-still-a-horseless-carriage.html
11. How AI is improving job searches on LinkedIn: https://www.computerworld.com/article/4096076/how-linkedin-is-using-ai-to-improve-its-job-search-features.html
12. More and more AI generated slop is on the internet, will consumers get a chance to choose to say no? https://www.computerworld.com/article/4094557/the-world-is-split-between-ai-sloppers-and-stoppers.html
13. Students are increasingly using AI as a learning aid, rather than a cheating tool: https://www.insidehighered.com/opinion/views/2025/11/26/case-ai-accommodation-opinion
14. AI in search is changing how companies present themselves online: https://www.rnz.co.nz/news/business/580085/it-s-interesting-times-how-ai-is-changing-the-search-for-information-in-nz
15. The British government is investing more in growing the AI sector: https://arstechnica.com/information-technology/2025/11/uk-government-will-buy-tech-to-boost-ai-sector-in-130m-growth-push/
16. A new benchmark for measuring the capacity of an AI to cause harm: https://techcrunch.com/2025/11/24/a-new-ai-benchmark-tests-whether-chatbots-protect-human-wellbeing/
17. Are Microsoft's statements on AI meaningful? https://www.computerworld.com/article/4093970/is-microsofts-humanist-superintelligence-vision-more-than-an-empty-slogan.html
18. Guardrails. If you put AI in kids toys, you need really good guardrails: https://www.theguardian.com/technology/2025/nov/28/artificial-intelligence-smart-toys
19. AI don't get puns. Dads everywhere are safe: https://www.theguardian.com/technology/2025/nov/24/ai-doesnt-get-puns-study-finds
20. Scientific datasets need to be restructured to allow them to be used by AI: https://www.hpcwire.com/bigdatawire/2025/11/27/the-global-race-to-build-ai-ready-scientific-datasets/
21. Amazon staff start pushing back against AI: https://www.theguardian.com/technology/2025/nov/28/amazon-ai-climate-change
22. Why vibe coding won't last: https://www.theregister.com/2025/11/24/opinion_column_vibe_coding/
23. Can we really trust a government AI to keep our information private: https://www.rnz.co.nz/news/national/580094/mental-health-data-won-t-be-misused-by-new-ai-navigation-tool-minister
24. AI will break safety rules if it means getting things done under pressure: https://spectrum.ieee.org/ai-agents-safety

Weekly Review 28 November 2025

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

1. AI generated workslop just makes more work for other people: https://www.rnz.co.nz/life/lifestyle/what-is-workslop
2. Everyone should be using AI so they can learn the difference between a good AI output and a bad AI output: https://www.informationweek.com/it-leadership/is-it-time-for-everyone-to-become-ai-power-users-
3. A new programming AI that can handle larger codebases: https://dataconomy.com/2025/11/20/openai-says-its-new-coding-model-can-work-for-24-hours-straight/
4. AI combined with drones speed power line maintenance: https://www.rnz.co.nz/news/business/579344/drones-linked-to-ai-used-to-help-fix-power-grids
5. What people are really sharing with AI: https://www.stuff.co.nz/world-news/360891883/what-you-should-know-trove-chatgpt-conversations-was-analysed
6. Even with mitigation techniques, AI still uses a lot of energy and a lot of water: https://www.techrepublic.com/article/news-energy-water-demands-sustainability-fears/
7. I would never use AI generated material in any course I teach: https://www.theguardian.com/education/2025/nov/20/university-of-staffordshire-course-taught-in-large-part-by-ai-artificial-intelligence
8. Attacks on AI guardrails are increasing and the protection is mostly inadequate: https://www.theregister.com/2025/11/14/ai_guardrails_prompt_injections_echogram_tokens/
9. Even having an AI generated book cover on a human written book is enough to get the book disqualified from a literary prize: https://www.rnz.co.nz/life/books/top-writers-ruled-out-of-nz-book-awards-due-to-ai-covers
10. The people who work to build AI are the ones who trust it the least: https://www.theguardian.com/technology/2025/nov/22/ai-workers-tell-family-stay-away
11. Google's ambitious plans to build more AI compute infrastructure: https://arstechnica.com/ai/2025/11/google-tells-employees-it-must-double-capacity-every-6-months-to-meet-ai-demand/
12. New Zealand retailers are using AI to stay competitive: https://www.nzherald.co.nz/nz/retailers-adopting-artificial-intelligence-to-compete-against-global-giants/6VST4BKTDBDPXBOMVXTQQYU3BU/
13. Curiosity Stream is now making most of its money licensing material to AI companies: https://arstechnica.com/gadgets/2025/11/curiosity-stream-expects-to-make-most-of-its-money-from-ai-deals-by-2027/
14. Higher education was not in a golden age before the appearance of generative AI: https://www.insidehighered.com/opinion/views/2025/11/20/no-pre-ai-era-was-not-great-opinion
15. The problem with using AI to provide advice on moving people off of ACC is that bureaucrats are lazy and will just do what the AI says: https://www.rnz.co.nz/news/national/579083/acc-s-use-of-ai-to-help-decide-who-gets-help-shocks-advocate
16. New Zealand hospitals are rolling out an AI tool for transcribing clinical notes: https://www.rnz.co.nz/news/national/579400/ai-scribe-tool-rolled-out-to-emergency-departments-promises-to-slash-clinicians-admin
17. Science uses AI more and more, but that means that those who can't afford to access the AI get shut out: https://www.hpcwire.com/bigdatawire/2025/11/21/the-compute-divide-in-ai-driven-science/
18. Microsoft continues to be disconnected from what its users really want with AI: https://www.theregister.com/2025/11/21/microsoft_ai_boss_comment/
19. A new maths benchmark shows that AI are actually bad at maths: https://www.theregister.com/2025/11/17/ai_bad_math_orca/
20. AI is accelerating scientific discoveries: https://www.theregister.com/2025/11/18/future_of_scientific_computing/
21. The AI bubble is destabilising the market:  https://www.rnz.co.nz/news/business/579606/stocks-lose-steam-on-ai-concerns-us-jobs-data
22. AI generated music is being produced at such a rate it risks overwhelming human made music: https://www.theregister.com/2025/11/17/ai_music/
23. When the AI bubble bursts it's going to cause a lot of damage: https://www.stuff.co.nz/money/360890298/ai-bubble-worlds-most-important-stock-could-shake-your-kiwisaver-week
24. It looks like Google's AI can't tell the difference between a person and 5.6m high statue: https://www.stuff.co.nz/culture/360884919/blurred-vision-what-google-and-apple-have-done-famous-new-zealand-statue

Weekly Review 21 November 2025

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

1. Microsoft says there is no AI bubble: https://www.computerworld.com/article/4087593/about-that-ai-bubble-microsoft-doesnt-see-one.html
2. AI country songs have come a long way since "You Can't Take My Door": https://www.rnz.co.nz/news/world/578654/ai-country-song-rollicks-to-the-top-of-us-sales-chart
3. AI mostly memorise, they don't really learn how to do things: https://arstechnica.com/ai/2025/11/study-finds-ai-models-store-memories-and-logic-in-different-neural-regions/
4. AI give themselves away by being too polite: https://arstechnica.com/information-technology/2025/11/being-too-nice-online-is-a-dead-giveaway-for-ai-bots-study-suggests/
5. This is the precedent AI companies didn't want: https://www.theguardian.com/technology/2025/nov/11/chatgpt-violated-copyright-laws-german-court-rules 
6. AI are still limited in their self-awareness of their reasoning processes: https://dataconomy.com/2025/11/11/anthropic-study-finds-ai-has-limited-self-awareness-of-its-own-thoughts/
7. Could AI reallt prevent prisoners from being released by mistake? https://www.theguardian.com/society/2025/nov/10/ai-chatbots-stop-prisoner-release-errors
8. While AI needs more and more data centres, in some parts of the world the environmental cost is causing push back: https://www.theguardian.com/technology/2025/nov/10/data-centers-latin-america
9. Most organisations are still seeing only minimal gains from rolling out AI: https://www.hpcwire.com/bigdatawire/2025/11/11/ai-is-everywhere-but-progress-is-slow-mckinsey-explains-why/
10. AI hasn't eliminated as many jobs as feared: https://www.rnz.co.nz/news/business/578397/is-ai-really-coming-for-our-jobs-and-wages-past-predictions-of-a-robot-apocalypse-offer-some-clues
11. Five free books to help you get into AI: https://www.kdnuggets.com/the-5-free-must-read-books-for-every-ai-engineer
12. AI chatbots can make eating disorders worse: https://www.theverge.com/news/818508/chatbot-eating-disorder-mental-health
13. The excuses lawyers have come up with after getting caught using AI to generate filings: https://arstechnica.com/tech-policy/2025/11/lawyers-keep-giving-weak-sauce-excuses-for-fake-ai-citations-in-court-docs/
14. A new AI framework that allows AI to train themselves without human involvement: https://www.computerworld.com/article/4088523/metas-spice-framework-pushes-ai-toward-self-learning-without-human-supervision-2.html
15. AI companies keep leaking their own secrets: https://www.techrepublic.com/article/news-ai-giants-secrets-github/
16. AI tools could program better than the Vibe Coding currently in use: https://www.theregister.com/2025/11/07/researchers_detail_legible_software_model/
17. The AI bubble is real and a collapse is coming: https://www.rnz.co.nz/news/thedetail/578574/artificial-intelligence-is-here-to-stay-but-what-about-the-companies-booming-from-it
18. Dealing with the hidden costs of AI: https://www.informationweek.com/it-leadership/how-a-cio-can-detect-and-address-ai-s-hidden-costs
19. Five AI industry terms you should know: https://www.theregister.com/2025/11/10/ai_credibility_buzzwords/
20. AI won't reach superintelligence anytime soon, but it is still useful: https://www.theregister.com/2025/11/11/ai_experts_forecast/
21. Most tertiary students know how and when to use AI, because their teachers told them: https://www.insidehighered.com/news/student-success/academic-life/2025/11/11/faculty-lead-ai-usage-conversations-college-campuses
22. Global divisions in the backend and application layers of AI: https://techcrunch.com/2025/11/11/the-global-race-for-the-ai-app-layer-is-still-on/
23. Excessive use of AI shows such intellectual laziness it's unattractive: https://www.theguardian.com/lifeandstyle/2025/nov/10/chatgpt-dating-ick
24. A smart bandage that uses AI to accelerate wound healing: https://spectrum.ieee.org/smart-bandage-ai-electrical-stimulation