Weekly Review 27 June 2025

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

1. Bots that scrape content for training AI are now bringing down websites with their voracious demands for data: https://dataconomy.com/2025/06/18/ai-is-breaking-the-internets-memory/
2. Generative AI does not contribute to a gain in the bottom line 80% of the time: https://dataconomy.com/2025/06/17/why-80-percent-of-gen-ai-delivers-no-real-business-gains/
3. There is a critical security vulnerability in Copilot. How many other AI tools also have vulnerabilities in them? https://www.bigdatawire.com/2025/06/17/zero-day-microsoft-copilot-vuln-underscores-emerging-ai-security-risks/
4. What to study if you want to understand agentic AI: https://www.kdnuggets.com/agentic-ai-a-self-study-roadmap
5. There is so much AI generated slop on the internet now that scraping it as training data is making other AI dumber: https://futurism.com/chatgpt-polluted-ruined-ai-development
6. Customers are getting sick of companies replacing real people with AI: https://www.nzherald.co.nz/world/popular-services-keep-adding-ai-some-customers-want-them-to-stop/MKAKEXK3VNDLXEYO26LV2IZ4GQ/
7. I really don't want AI looking over my shoulder all the time I'm working: https://www.extremetech.com/computing/copilot-vision-lets-ai-see-your-windows-11-screen-and-guide-you-in-real
8. AI cheating amongst students is not going to go away. We're going to have to set assessments they can't cheat at instead: https://www.theregister.com/2025/06/16/university_ai_cheating/
9. Mastodon has now banned the use of posts on their platform for AI training. But how can they enforce it? https://www.theregister.com/2025/06/18/mastodon_says_no_to_ai/
10. The use of AI to generate code is increasing, especially in the US: https://www.theregister.com/2025/06/19/us_coders_lead_world_in_ai/
11. If you use AI to filter content, you need to back it up with human confirmation: https://techcrunch.com/2025/06/17/tumblrs-content-filtering-systems-have-been-falsely-flagging-posts-as-mature-users-blame-ai/
12. If you are going to ban people from your platform, you need to back up AI decisions with human review: https://techcrunch.com/2025/06/16/instagram-users-complain-of-mass-bans-pointing-finger-at-ai/
13. AI can make people with delusions and other mental illnesses worse: https://techcrunch.com/2025/06/15/spiraling-with-chatgpt/
14. Using AI to do you thinking for you does make you dumber: https://www.theregister.com/2025/06/18/is_ai_changing_our_brains/
15. Now AI are randomly sharing phone numbers of other users: https://www.theguardian.com/technology/2025/jun/18/whatsapp-ai-helper-mistakenly-shares-users-number
16. Amazon is planning more job cuts, replacing them with AI: https://www.theguardian.com/technology/2025/jun/18/amazon-boss-tells-staff-ai-means-their-jobs-are-at-risk-in-coming-years
17. Adoption of AI is being limited by the lack of clarity around ongoing costs: https://www.computerworld.com/article/4007713/companies-limit-their-ai-use-due-to-unclear-costs.html
18. Training an image generator with no outside data set: https://www.theverge.com/ai-artificial-intelligence/688576/feed-ai-nothing
19. An interactive map of the energy used by AI: https://www.theregister.com/2025/06/19/iea_observatory_ai/
20. The different kinds of intelligence used in business-not just AI: https://www.techrepublic.com/article/differences-bi-ai-analytics/
21. Lack of skills in AI threaten job security more than AI itself: https://www.computerworld.com/article/4006931/ai-isnt-taking-your-job-the-big-threat-is-a-growing-skills-gap.html
22. AI in the workplace is still not living up to the hype: https://www.theregister.com/2025/06/17/ai_not_doing_its_job/
23. Are AI starting to develop their own personas? https://techcrunch.com/2025/06/18/openai-found-features-in-ai-models-that-correspond-to-different-personas/
24. The security issues raised by agentic AI: https://devclass.com/2025/06/17/redefining-identity-security-in-the-age-of-agentic-ai/