Weekly Review 16 February 2024

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

• Generative AI is going to make some parts of filmmaking easier, but it's not going to displace humans entirely from the process just yet: https://www.technologyreview.com/2023/06/01/1073858/surreal-ai-generative-video-changing-film/ 
• Using generative AI for robocalling is now illegal in the US: https://www.cbsnews.com/news/fcc-declares-robocalls-illegal/ 
• This is not going to end well-North Island supermarkets using facial recognition software: https://www.nzherald.co.nz/nz/privacy-commissioner-promises-to-closely-monitor-foodstuffs-facial-recognition-trial/3JXFUBVGJVE47AA2GTVC5E6ASM/ 
• Minorities are going to be disproportionately impacted by facial recognition software in supermarkets. They are almost certainly going to be underrepresented in the training data sets, so there is going to be a lot of false positives: https://www.nzherald.co.nz/kahu/fears-maori-women-will-be-targets-in-foodstuffs-facial-recognition-trial-in-north-island/7RWYY4KZS5EUBNZ77YHBD2MIPI/ 
• It doesn't matter that people will supposedly be comparing results of facial recognition in supermarkets-people are lazy, and will just follow what the machine tell them: https://www.newstalkzb.co.nz/on-air/heather-du-plessis-allan-drive/audio/julian-benefield-foodstuffs-north-island-general-counsel-says-human-element-will-play-key-role-in-facial-recognition-technology-trial/ 
• An AI search engine that seems to be like all the other conversational AI - grievously stupid, with no sense of reality: https://futurism.com/the-byte/ai-search-engine-nyt 
• Using an AI to enhance scans of carbonised ancient scrolls from Herculaneum. Important to bear in mind that this is a machine interpreting the scans, not necessarily what the scrolls actually say: https://www.nature.com/articles/d41586-024-00346-8 
• It's hard enough to get peer reviewers for journal submissions, if anything AI is going to make that worse-a never-ending, self-reinforcing morass of inane sameness in reviews: https://spectrum.ieee.org/ai-peer-review 
• Engineering metamaterials using AI: https://phys.org/news/2024-02-ai-tool-realistic-metamaterials-unusual.html 
• The environmental impact of AI: https://e360.yale.edu/features/artificial-intelligence-climate-energy-emissions 
• Data centres to run AI need so much electricity that coal-fired powerplants are having to be kept online, with all the carbon emissions that come with them: https://futurism.com/the-byte/coal-plants-ai 
• Now there's an ISO standard for the responsible use of AI: https://spectrum.ieee.org/tech-standards-responsible-ai This'll certainly be helpful for the consultancy industry
• Improved sentiment analysis in Python: https://www.kdnuggets.com/sentiment-analysis-in-python-going-beyond-bag-of-words 
• As the hype around AI increases, so too does the demand for engineers to develop it: https://www.datanami.com/2024/02/07/hiring-genai-talent-its-a-matter-of-degree/ 
• Approaches to embed reasoning in to #NeuralNetworks: https://www.datasciencecentral.com/a-neurosymbolic-ai-approach-to-learning-reasoning/ 
• Future roles for AI and #MachineLearning in the IT industry: https://dataconomy.com/2024/02/06/navigating-tomorrow-role-of-ai-and-ml-in-information-technology/
• Not surprising that safeguards around AI can be broken, it's going to be years before we get a good grip on this problem: https://www.theguardian.com/technology/2024/feb/09/ai-safeguards-can-easily-be-broken-uk-safety-institute-finds
• So now AI are being used to determine who does or does not get medical coverage in the USA? I thought the US medical system couldn't get any worse: https://www.theregister.com/2024/02/09/ai_medicare_health/ 
• The tool I used in the past to detect AI generated student essays was pretty rubbish, far too high a false positive rate: https://www.insidehighered.com/news/tech-innovation/artificial-intelligence/2024/02/09/professors-proceed-caution-using-ai 
• A US advisory group on AI regulation has been formed: https://www.computerworld.com/article/3712863/us-creates-advisory-group-to-consider-ai-regulation.html 
• Generative AI enhances human creativity, it doesn't replace it: https://www.theguardian.com/fashion/2024/feb/08/ai-london-fashion-week 
• Overview of Large Language Model AI and how they work: https://www.computerworld.com/article/3697649/what-are-large-language-models-and-how-are-they-used-in-generative-ai.html 
• All the ways deepfakes have been used to scam people: https://www.informationweek.com/machine-learning-ai/dealing-with-deepfakes 
• An AI that is so ethical, it won't do anything: https://techcrunch.com/2024/02/09/meet-goody-2-the-ai-too-ethical-to-discuss-literally-anything/