Weekly Review 24 July 2026

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

1. Laptops and smart devices are banned by a law school over concerns of students using AI to cheat: https://www.the-independent.com/tech/laptop-ban-university-ai-backlash-b3013911.html
2. The AI arms race in technical interviews: https://spectrum.ieee.org/technical-interview-ai-arms-race
3. Which jobs are most threatened by AI in New Zealand: https://www.stuff.co.nz/nz-news/361007179/kiwi-jobs-highest-risk-ai-automation
4. South Korea launches a project to build an AI chatbot that all its citizens can use to access government services: https://www.theregister.com/public-sector/2026/07/15/south-korea-to-launch-universal-basic-ai-chatbot/5271566
5. AI data centres are not the blessing for New Zealand that some people say they are: https://www.stuff.co.nz/money/361006005/data-centres-beware-greeks-bearing-gifts
6. Clothing to defeat AI facial recognition systems: https://www.theguardian.com/fashion/2026/jul/17/adversarial-clothing-are-garments-designed-to-confuse-facial-recognition-systems-about-to-go-mainstream
7. An animated font that AI cannot read but humans can: https://www.extremetech.com/internet/ai-cant-read-this-trippy-ghost-font
8. Governments are moving to reduce the number of layoffs caused by AI in call centres: https://www.techtarget.com/searchcustomerexperience/news/366645896/World-leaders-confront-AI-layoffs-more-in-store-for-contact-centers
9. AI can design DNA sequences so that they'll fold into any desired shape: https://spectrum.ieee.org/ai-dna-origami
10. Jailbreaking AI isn't that hard. I had a research student who convinced several to produce phishing emails that couldn't be detected by conventional filters: https://spectrum.ieee.org/jailbreaking-llms
11. References hallucinated by AI continue to appear: https://newsroom.co.nz/2026/07/09/pest-control-report-cites-fake-papers-relying-on-chatgpt-hallucinations/
12. AI is making university faculty more likely to want to retire: https://karenkelsky.substack.com/p/is-ai-making-faculty-more-likely
13. More warnings about the economic impact of the AI bubble bursting: https://futurism.com/artificial-intelligence/ai-bubble-fears-tsmc-nvidia-earnings
14. South Korea is developing its own security AI: https://www.theregister.com/ai-and-ml/2026/07/17/south-korea-making-its-own-security-centric-ai-model/5274034
15. The world's largest open source AI model, with 2.8 trillion parameters: https://www.techrepublic.com/article/news-moonshot-ai-kimi-k3-largest-open-source-model-apac/
16. AI is accelerating the development of quantum computers, which threatens security: https://www.rnz.co.nz/news/business/707600/ai-is-accelerating-quantum-computing-could-our-data-be-vulnerable
17. Australia is establishing an office for AI, to set standards and reduce energy and water usage: https://www.rnz.co.nz/news/world/723387/australia-to-establish-government-ai-office-curb-data-centres-water-use
18. AI is not conscious, but it is developing its own way of thinking: https://www.stuff.co.nz/nz-news/361006118/baffling-discovery-has-scientists-asking-if-ai-already-conscious
19. Improved flash memory can be used to help to speed up recall in AI: https://spectrum.ieee.org/high-bandwidth-flash
20. AI making layoff decisions is not a good idea, when the AI doesn't know about anti-discrimination laws: https://arstechnica.com/tech-policy/2026/07/lawsuit-claims-metas-layoff-decisions-were-made-by-ai-not-humans/
21. It is easy to install backdoors in open AI models: https://www.theregister.com/ai-and-ml/2026/07/16/researcher-poisons-open-weight-ai-model-for-under-100/5273880
22. Salting text into spam emails makes it harder for AI to detect them: https://www.theregister.com/security/2026/07/17/ai-spam-filters-are-getting-suckered-by-old-school-text-salting/5274434
23. More warnings about the economic impact of large scale AI driven job losses: https://futurism.com/artificial-intelligence/economists-warning-ai-job-losses
24. Microsoft is laying off more employees and putting the money into AI: https://www.computerworld.com/article/4196309/with-its-latest-layoffs-microsoft-goes-all-in-on-ai.html