Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, and Bluesky):
- Using AI to analyse recordings of surgery. Maybe it'd be more welcome in less litigious countries: https://www.technologyreview.com/2024/06/07/1093338/surgical-safety-technologies-ai-black-box-surgery-operating-room/
- A large proportion of AI projects don't yield positive financial results. https://futurism.com/the-byte/companies-ai-projects-financial-results
- Google's AI generated search summaries are not reliable: https://spectrum.ieee.org/google-ai-search
- So OpenAI licensed their content, and now ChatGPT is hallucinating fake articles: https://futurism.com/the-byte/chatgpt-hallucinating-fake-links-business-insider
- The ongoing arms races between AI generated deepfakes and the AI that can detect them: https://spectrum.ieee.org/deepfake-detection
- Nvidia hardware continues to dominate in AI performance: https://spectrum.ieee.org/mlperf-nvidia-conquers
- Do we still need passwords in the age of AI? https://www.theguardian.com/technology/article/2024/jun/11/apple-password-app-tech-age-of-ai
- Do we really need a PC with built in AI hardware? https://dataconomy.com/2024/06/12/do-we-need-an-ai-pc/
- 40% of workers fear losing their jobs to AI: https://www.computerworld.com/article/2142370/afraid-ai-will-steal-your-job-youre-not-alone.html
- If we're relying on the pope to warn us about ethical issues with AI, we're all in a lot of trouble: https://www.theregister.com/2024/06/12/pope_g7_ai_warning/
- A cheat sheet for ChatGPT: https://www.techrepublic.com/article/chatgpt-cheat-sheet/
- The initial mad rush to get into AI seems to be ending: https://www.theregister.com/2024/06/12/survey_ai_projects/
- Brazil plans to use AI to speed up processing of lawsuits: https://www.theregister.com/2024/06/11/brazil_openai_justice/
- LinkedIn job hunting tools now include AI to write an application for you. Which no doubt will be processed by another AI. https://techcrunch.com/2024/06/13/linkedin-leans-on-ai-to-do-the-work-of-job-hunting/
- AI are running out of data to train on, which is stunting their growth: https://futurism.com/the-byte/ai-running-out-data-smarter
- Better to ask if you really need AI, rather than ask how quickly you can get it: https://www.datanami.com/2024/06/13/why-the-current-approach-for-ai-is-excessively-dangerous/
- A brief description of five (basic) AI algorithms: https://www.kdnuggets.com/5-machine-learning-models-explained-in-5-minutes
- No, an AI is not running for mayor of Cheyenne, Wyoming. But a candidate is using AI, probably the only one being open about it: https://www.theregister.com/2024/06/12/ai_bot_wyoming/
- Integrating AI into business IT systems is fraught, opening up security and privacy issues: https://www.computerworld.com/article/2144103/it-pros-find-generative-ai-doesnt-always-play-well-with-others.html
- Support staff are the unsung heroes of research, and deserve more recognition: https://www.nature.com/articles/d41586-024-01749-3
- Online exams are still a tricky thing to do, especially so when technical problems arise: https://www.nzherald.co.nz/nz/massey-university-forced-to-cancel-online-exams-due-to-technical-problems/B2DSD5DUCBD73O7QHRHI2HJ5JA/
- The skills an AI developer should have: https://www.informationweek.com/machine-learning-ai/key-skills-to-look-for-in-an-ai-developer
- Opportunities for New Zealand in AI: https://thespinoff.co.nz/partner/10-06-2024/will-ai-accelerate-or-relegate-new-zealands-tech-sector
- Apple claims their AI architecture protects personal data, and is making some of their code open to verify it: https://arstechnica.com/ai/2024/06/apples-ai-promise-your-data-is-never-stored-or-made-accessible-by-apple/
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.