Weekly Review 30 August 2024:

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

1. There is no way AI is going to replace developers in two years, because the code generated by AI has so many security holes that it will never be accepted in industries like finance, health, infrastructure, defense... https://futurism.com/the-byte/aws-ceo-human-devs-ai
2. If this AI is looking at things like tone of voice, then I'd be in big trouble, the accent where I am from doesn't really inflect at all: https://www.stuff.co.nz/business/350384218/jobseeker-interviewed-100-robots-cant-get-dishwasher-work
3. Lots of organisations are trying out generative AI, but time is running out for them to generate a positive return: https://www.datanami.com/2024/08/22/reality-check-for-genai-deloitte-finds-enthusiasm-tempered-by-adoption-hurdles/
4. Still more legal action over AI companies stealing training data: https://www.informationweek.com/machine-learning-ai/anthropic-hit-with-copyright-lawsuit-over-llm-training
5. AI is a tech field, and tech still has a large gender disparity in it. I've noticed more women studying tech over the years, but change is always slow: https://dataconomy.com/2024/08/21/the-complicated-relationship-between-ai-and-gender-equality/
6. AI is a tool, not a god or a magician. It can make decisions, but so can a coin: https://www.theguardian.com/technology/article/2024/aug/24/yuval-noah-harari-ai-book-extract-nexus
7. The obsession with AI might be distracting organisations from the fundamentals of IT services: https://www.theregister.com/2024/08/22/ai_may_be_distracting_organizations/
8. At least someone's still making money out of AI: https://techcrunch.com/2024/08/24/vcs-are-so-eager-for-ai-startups-theyre-buying-into-each-others-spvs-at-high-prices/
9. I've noticed that while Gen Z is very comfortable with using technology, they lack the knowledge of how it works, which limits their engagement with it. So the results of this survey on their attitudes to AI are not surprising to me: https://techcrunch.com/2024/08/21/this-week-in-ai-gen-z-has-mixed-feelings-on-ai/
10. While I've done research myself into the environmental impact of AI, I don't think that's the main thing faculty should be considering. Rather, the ethical issues with students using generative AI: https://www.insidehighered.com/opinion/views/2024/08/22/gen-ai-discussions-must-prioritize-sustainability-opinion
11. Like every other tool, if you want to use generative AI effectively in your organisation, you have to do the preparatory ground work first: https://www.computerworld.com/article/3488668/want-genai-to-deliver-benefits-you-have-a-lot-of-work-to-do-first.html
12. Is artificial general intelligence really worth it? Will it contribute anything useful? In any case, I don't think generative AI is the pathway to it: https://www.theregister.com/2024/08/22/gartner_agi_hype_cycle/
13. AI being used to screen applications generated by AI, putting everyone back to square one: https://www.informationweek.com/machine-learning-ai/the-search-for-solid-hires-between-ai-screening-and-genai-resumes
14. I expect there will be more and more push-back from AI companies about regulation of AI, a lot of them play so fast and loose with ethics and safety that any regulation is a major threat to them: https://techcrunch.com/2024/08/21/openais-opposition-to-californias-ai-law-makes-no-sense-says-state-senator/
15. More push-back from big tech companies on the regulation of AI: https://techcrunch.com/2024/08/23/meta-and-spotify-ceos-criticize-ai-regulation-in-the-eu/
16. Ways AI is being used in HR. As always, bad training data is causing problems: https://www.datasciencecentral.com/how-will-machine-learning-transform-hr-strategies/
17. Generative AI just isn't that useful: https://www.computerworld.com/article/3488589/ive-got-the-genai-blues.html
18. It is possible to get some value out of generative AI, despite its overall lack of usefulness: https://www.datanami.com/2024/08/21/getting-value-out-of-genai/
19. Slack's AI is vulnerable to a prompt injection attack. How many AI share this security flaw? https://www.theregister.com/2024/08/21/slack_ai_prompt_injection/
20. It seems that Europe is not immune to the funding feeding frenzy around AI: https://techcrunch.com/2024/08/24/the-top-ai-deals-in-europe-this-year/
21. Five free courses on AI: https://www.kdnuggets.com/top-5-free-machine-learning-courses-to-level-up-your-skills
22. Disillusionment is starting to set in around generative AI: https://www.datanami.com/2024/08/21/genai-begins-journey-into-trough-of-disillusionment/
23. When AI hallucination meets Rickrolling: https://futurism.com/the-byte/ai-startup-rickrolling
24. Ways in which AI can contribute to research in physics: https://dataconomy.com/2024/08/23/revolutionizing-physics-how-ai-is-transforming-research-and-discovery/
25. The problem with using AI to fake quotes from real people, is that those people get really annoyed when it happens: https://futurism.com/the-byte/megalopolis-trailer-quotes-chatgpt
26. How AI can assist, rather than replace, software developers: https://dataconomy.com/2024/08/23/ai-assistants-for-software-engineers/
27. Ten free courses on generative AI: https://www.kdnuggets.com/10-free-resources-to-learn-llms
28. An improved data centre architecture for training AI: https://www.nextplatform.com/2024/08/23/this-ai-network-has-no-spine-and-thats-a-good-thing/
29. The stock market is essentially random, and AI are not good at predicting randomness: https://futurism.com/the-byte/ai-picking-stocks-faceplanting