Weekly Review 5 April 2024

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

1. Adobe's ethical generative AI system: https://www.technologyreview.com/2024/03/26/1090129/how-adobes-bet-on-non-exploitative-ai-is-paying-off/
2. Higher impact factor journals are less inclusive of non-English speaking authors: https://phys.org/news/2024-03-prestigious-journals-hard-scientists-dont.html
3. Reinforcement learning enables a robot to learn to walk over uneven terrain: https://wccftech.com/chinas-a-i-powered-robot-nearly-fights-back-after-being-hit-pulled-by-trainer/
4. Job applications from AI that are then processed by AI. It's just machines talking to machines, insulating people from talking to one another: https://futurism.com/the-byte/unemployed-man-ai-apply-5000-jobs-gets-20-interviews
5. Laws that prevent the use of AI to mimic performers voices are being passed: https://www.billboard.com/business/legal/tennessee-elvis-act-protecting-artists-voices-ai-impersonation-1235637934/
6. Making beer better with AI: https://www.technologyreview.com/2024/03/26/1090156/ai-could-make-better-beer-heres-how/
7. AI produces better climate models, faster than traditional approaches: https://www.nature.com/articles/d41586-024-00780-8
8. Google's AI can discover new inorganic crystals at a very high rate: https://www.freethink.com/robots-ai/google-ai-discovers-2-2-million-new-materials
9. Will AI really help preserve the Maori language, or will it propagate a hallucinated facsimile of it? https://www.stuff.co.nz/te-ao-maori/350203962/theres-too-much-stake-not-seize-ai-opportunity-te-reo
10. I prefer to follow the Open Source Initiative's definition of what is open source, which makes a lot of the AI out there NOT open source: https://www.technologyreview.com/2024/03/25/1090111/tech-industry-open-source-ai-definition-problem/
11. Wow, people really don't like AI generated art: https://futurism.com/reddit-ai-art
12. An AI tool for inventorying houses for a moving company, and the trevails its founders went through to get the company running: https://spectrum.ieee.org/zach-rattner-yembo-profile
13. There is no causative link between tweeting about your research paper and the number of citations it garners: https://www.nature.com/articles/d41586-024-00922-y
14. Google's AI search apparently lacks the intelligence to not recommend scam sites: https://www.bleepingcomputer.com/news/google/googles-new-ai-search-results-promotes-sites-pushing-malware-scams/
15. A substantial proportion of researchers in AI are Chinese, and more of them are staying in China rather than emigrating: https://www.technologyreview.com/2024/03/27/1090182/ai-talent-global-china-us/
16. How long until the entire game is generated by AI, not just the artwork? https://futurism.com/the-byte/nvidia-ceo-ai-games
17. The risks of using AI in research articles: https://arstechnica.com/science/2024/03/producing-more-but-understanding-less-the-risks-of-ai-for-scientific-research/
18. AI that learn from other AI: https://dataconomy.com/2024/03/27/self-teaching-ai-models-might-have-been-discovered/
19. This article on a huge AI processor reads like it was written using one: https://www.tomshardware.com/tech-industry/artificial-intelligence/broadcom-shows-gargantuan-ai-chip-xpu-could-the-worlds-largest-chip-built-for-a-consumer-ai-company
20. AI is going to be a big player in this year's elections around the world: https://www.theregister.com/2024/03/29/hillary_clinton_election_ai/
21. The hazards of biased AI in healthcare: https://www.informationweek.com/machine-learning-ai/how-ai-bias-is-impacting-healthcare
22. We desperately need tools to detect deepfakes: https://dataconomy.com/2024/03/28/deepfake-detection-tools-and-techniques/
23. AI can now design novel enzymes: https://www.freethink.com/science/ai-designed-enzymes
24. On the importance of labelling data for use in training AI: https://www.datasciencecentral.com/decoding-ai-success-the-complete-data-labeling-guide/