Weekly Review 6 October 2023

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

1. Using a convolutional Neural Network for image classification in PyTorch: https://www.kdnuggets.com/building-a-convolutional-neural-network-with-pytorch
2. How the "Face on Mars" illustrates the problem of bias in data: https://www.datanami.com/2023/09/27/the-face-of-data-bias/ This is a particular problem for AI
3. KrispAI uses deep Neural Networks to filter out background noise for online meetings: https://dataconomy.com/2023/09/26/what-is-krisp-ai-and-how-to-use-it/
4. The AI tools that YouTube is planning to roll out for its creators: https://www.extremetech.com/internet/youtube-teases-ai-powered-creator-tools
5. The opportunities AI offer for Māori: https://www.stuff.co.nz/pou-tiaki/133017132/microsofts-dan-walker-excited-to-see-how-ai-can-disrupt-inequity-for-mori
6. Even though US police like to use tools like facial recognition, most aren't trained in how to use it properly: https://www.gao.gov/products/gao-23-105607
7. The real problem here is that robots.txt is unenforceable, a crawler can always just ignore it. I suspect the real solution is going to be legal, not technical: https://www.theregister.com/2023/09/29/medium_ai_crawlers/
8. Cubomedusae are fascinating creatures that exhibit complex, predatory behaviours despite not having brains. It turns out they can learn, too: https://www.extremetech.com/science/scientists-discover-jellyfish-can-learn-without-a-brain 
9. Detecting ancient biological samples using AI: https://phys.org/news/2023-09-life-mars-planets-ai.html
10. How startups can make use of AI and Machine Learning: https://dataconomy.com/2023/09/29/ai-and-machine-learnings-impact-on-startups/
11. Spotify is using AI to clone people's voices in other languages: https://www.theregister.com/2023/09/27/spotify_now_using_ai_to/
12. How tertiary teachers need to adapt in the era of generative AI: https://www.insidehighered.com/opinion/views/2023/09/28/ai-and-convergence-writing-and-coding-opinion
13. Making AI image processing less biased when handling images of people: https://www.technologyreview.com/2023/09/25/1080191/these-new-tools-could-make-ai-vision-systems-less-biased/
14. The risks shadow AI - unauthorised use of AI tools - pose to an organisation: https://www.informationweek.com/it-leadership/is-your-organization-vulnerable-to-shadow-ai-
15. Large Language Model AI are now so large they are putting pressure on the chip manufacturing industry: https://www.computerworld.com/article/3707412/chip-industry-strains-to-meet-ai-fueled-demands-will-smaller-llms-help.html
16. Amazon is putting US$4b into Anthropic to boost their generative AI technology: https://www.nzherald.co.nz/business/amazon-makes-us4-billion-ai-bet-on-anthropic/U6HPXRNRBREA5OAYX7TQNBNIWM/
17. Cloning television presenters using AI: https://www.nextplatform.com/2023/09/28/bringing-ai-to-reality/
18. An overview of the AI functions embedded into the latest update to Windows 11: https://www.extremetech.com/computing/windows-11s-latest-update-focuses-on-ai
19. Identifying fairy circles in aerial photos using a Neural Network: https://edition.cnn.com/2023/09/28/world/fairy-circles-new-sites-scn/index.html
20. Rather stating the obvious, but it does show how generative AI is the new darling of the hype around artificial intelligence: https://devclass.com/2023/09/26/survey-how-organizations-believe-ai-will-change-devops/