Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, and Bluesky):
- The majority of posts on LinkedIn are generated by AI: https://www.theregister.com/2024/11/28/linkedin_ai_posts/
- Optical data transmission is the key to meeting the computing needs of AI: https://www.theregister.com/2024/11/28/ai_copper_cables_limits/
- Possible legal obstacles to Google's AI plans: https://arstechnica.com/tech-policy/2024/11/google-drags-ai-rivals-into-search-trial-as-judge-entertains-ai-remedies/
- Even though energy supplies are uncertain, investors continue to back building data centres to meet the computing needs of AI: https://www.theregister.com/2024/11/27/investor_datacenter_energy_concern/
- An AI music generator that can produce sounds that have never existed before: https://arstechnica.com/ai/2024/11/nvidias-new-ai-audio-model-can-synthesize-sounds-that-have-never-existed/
- Generative AI hallucinates and is unexplainable. How can organisations trust it? https://www.informationweek.com/machine-learning-ai/how-can-decision-makers-trust-hallucinating-ai-
- What those responsible for IT security need to think about regarding AI regulations: https://www.informationweek.com/cyber-resilience/meeting-ai-regulations-a-guide-for-security-leaders
- The academic and teaching benefits of having a conversation with an AI: https://www.insidehighered.com/opinion/career-advice/teaching/2024/11/27/how-ai-can-help-teach-art-conversation-opinion
- Guardrails for health AI aim to reduce biased models: https://arstechnica.com/health/2024/11/biden-proposes-guardrails-on-health-care-ai-upping-weight-loss-drug-access/
- Building your own AI app on a shoestring: https://www.kdnuggets.com/diy-ai-building-ai-apps-shoestring-budget
- The British don't seem to be all that interested in AI: https://www.theregister.com/2024/11/29/ofcom_online_nation/
- Britain is investing in AI to improve its cybersecurity: https://www.theregister.com/2024/11/26/uk_ai_security/
- Some approaches to protect privacy when using AI: https://www.datasciencecentral.com/ethical-ai-in-data-practices-balancing-innovation-and-privacy/
- Generative AI will hallcuinate citations rather than cite actual publications: https://techcrunch.com/2024/11/29/study-of-chatgpt-citations-makes-dismal-reading-for-publishers/
- Yet another lawsuit from publishers over their material being used to train AI: https://techcrunch.com/2024/11/29/canadian-news-companies-sue-openai/
- An overview of the last two years of ChatGPT and generative AI: https://www.computerworld.com/article/3615039/two-years-of-chatgpt-the-conversation-that-never-ends.html
- The roles you need to fill for a successful AI project: https://www.bigdatawire.com/2024/11/26/building-your-genai-dream-team/
- Building a successful AI strategy: https://www.bigdatawire.com/2024/11/19/tips-on-building-a-winning-data-and-ai-strategy-from-jpmc/
- Cloning Panasonic's founder through AI: https://www.theregister.com/2024/11/29/panasonic_ai_founder/
- AI can't make a bad photo good, to take good photos you need to be a good photographer: https://www.informationweek.com/machine-learning-ai/how-ai-is-revolutionizing-photography
- How to assess when you should use generative AI: https://www.bigdatawire.com/2024/11/27/llms-and-genai-when-to-use-them/
- AI that can detect a driver is dozing off: https://spectrum.ieee.org/driver-drowsiness-detection
- Even AI needs a backup: https://arstechnica.com/tech-policy/2024/11/tech-problems-plague-openai-court-battles-judge-rejects-a-key-fair-use-defense/
- Despite registering their use being mandatory, most British government users of AI are not doing so: https://www.theguardian.com/technology/2024/nov/28/uk-government-failing-to-list-use-of-ai-on-mandatory-register