ProSearch’s Post

False overarching narratives are much harder to address than individual facts - audiences look for confirmatory evidence, creating demand for misinformation and conspiracy theories. Media Literacy initiatives increase individuals' ability to read critically and recognise obvious triggers, but are not so good at long-term campaigns. The news industry can help by providing context, challenging narratives when they emerge and being careful not to normalise narratives through "bothsideism".

Enterprise L&D: In August 1979 I learned about Rummler & Gilbert’s 1970 views on Guidance – The Short Way Home. Sometimes you can look it up, get it right, and then, forgetaboutit. If used often enough memorization will come - unless that's inappropriate. And sometimes we don't want memorization - if the Information and Instructions are "dynamic and not static." https://lnkd.in/ebtijJze https://lnkd.in/eNfhjpP

My latest blog post...is the word 'Archived' a dangerous word? I explore the misuse of the word 'archiving' in day-to-day apps/software and if this can cause real damage to organisations both now, and in the future. Some lighter reading for over the weekend!

Just wanted to share our latest article, written by Eileen McDermott, with you!

https://lnkd.in/eH-JrRrY Deep Dive: Transforming Text into Knowledge Graphs with LLM MinYang C.

Just finished re-reading the classic - Toffler’s Future Shock after 50 years! It was amazing then and just as amazing now! 1. The understanding and insight it has given me of of the good and not so good things, happening in the present (eg Application of Ross Ashby’s law in the social context in the last chapter). 2. How open minded reasoning and boundary less thinking can provide useful cues on what is to come(eg OLIVER in the penultimate chapter)

Check out this article!

Have to put this here for peer review and as a source of information.

'Pretrieval practice' enhances engagement with new information. This technique not only aids in the retention of specific facts and concepts but also fosters a deeper comprehension of the material 🧠 https://buff.ly/3v0H0dh

Is human evaluation scalable in LLM-powered systems? In our latest episode, Aaro and I explored the common practice of using domain experts to assess the outputs of LLM systems. This method, while reliable, comes with its drawbacks. It's no secret that having experts manually evaluate every output is costly and time-consuming. Every tweak to the system requires a fresh evaluation, driving up expenses and delaying progress. Is this the best we can do? The lack of scalability creates room for alternative solutions like LLM-as-judge, which you’ll hear us discussing soon. Dive deeper into the topic: https://lnkd.in/dWuySZR9