Weekly Review 11 October 2024

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

1. I think it is going to become very easy for police to blame their own errors or manipulations on mistakes made by the AI. And when the AI does mess up, where does the responsibility lie? https://www.theguardian.com/technology/2024/oct/03/artificial-intelligence-police-reports
2. Companies are shifting their AI data centres to Texas, but Texas doesn't have the generating capacity to satisfy their energy demands: https://www.theregister.com/2024/10/04/texas_dc_power/
3. A fortnight after California's law banning AI generated fakes during elections was signed into law, a judge has blocked it: https://dataconomy.com/2024/10/03/judge-slams-californias-ai-deepfake-law/
4. The five AI companies being prosecuted for making misleading claims about what their technologies can do: https://dataconomy.com/2024/09/30/ftc-cracks-down-on-five-ai-companies/
5. Meta's AI glasses will use your queries to train its AI, and you can't opt out: https://www.extremetech.com/internet/meta-will-train-its-ai-enabled-eyeglasses-on-stuff-you-ask-it-to-analyze
6. Sometimes simpler models are better than more complex models in AI: https://www.kdnuggets.com/do-we-really-need-more-complex-models
7. Is regulation needed to prevent AI companies just taking the images of deceased family members and using them as the faces of AI? https://dataconomy.com/2024/10/03/dont-allow-ai-to-profit-from-the-pain-and-grief-of-families/
8. We struggle to define intelligence, defining artificial general intelligence is going to be even harder: https://techcrunch.com/2024/10/03/even-the-godmother-of-ai-has-no-idea-what-agi-is/
9. AI assists in making products recyclable: https://www.informationweek.com/sustainability/ai-dials-up-circularity
10. Some jobs are more likely to be replaced by AI in the near-term than others: https://www.datanami.com/2024/09/27/how-genai-will-impact-jobs-in-the-real-world/
11. Responsible AI must include clear rules around the ownership of data: https://www.datasciencecentral.com/a-user-centric-approach-to-responsible-ai/
12. Computers with embedded AI hardware are coming, and Intel has released a new software tool to help familiarise people with the technology: https://www.extremetech.com/computing/intel-updates-its-ai-playground-software-for-intel-powered-ai-pcs
13. You don't need the cutting-edge tech to develop decent AI models, as China has demonstrated: https://www.theregister.com/2024/10/02/china_telecom_model_trained_local_tech/
14. The contribution of AI to eLearning: https://www.datasciencecentral.com/how-ai-is-shaping-elearning/
15. Ways Google's AI-powered notebook can help you: https://dataconomy.com/2024/10/01/creative-ways-to-use-google-notebook-lm/
16. Was the California AI safety bill really that flawed? And is anything going to replace it? https://www.informationweek.com/machine-learning-ai/newsom-veto-kills-california-ai-safety-bill
17. More ways wearable AI can be misused-instantly doxing someone: https://www.theverge.com/2024/10/2/24260262/ray-ban-meta-smart-glasses-doxxing-privacy
18. Tertiary students' attitudes to AI: https://www.insidehighered.com/news/student-success/life-after-college/2024/10/03/are-ai-skills-key-part-career-preparation
19. OpenAI is demanding exclusivity from its investors. So they can invest in one AI company but not any others? https://www.computerworld.com/article/3544921/openai-demands-investors-shun-rivals-such-as-anthropic-elon-musks-xai.html
20. Optimising database performance to support AI operations: https://www.datasciencecentral.com/large-ai-apps-optimizing-the-databases-behind-the-scenes/
21. The environmental impact of AI: https://www.theguardian.com/technology/2024/oct/02/french-ai-summit-environmental-impact-energy-hungry-tech
22. Text mining in the R language: https://www.kdnuggets.com/how-to-use-r-for-text-mining
23. AI is not ready for use as a sysadmin: https://www.theregister.com/2024/10/02/ai_agent_trashes_pc/