This week, Joanna Wright, our London editor, joins Wei-Shen on the podcast to talk about her feature on how transformer models are benefitting the field of natural language processing (NLP). Then, they talk about having fun at the workplace, and how companies can help provide a sense of fulfillment to their staff.
https://www.waterstechnology.com/technology/4870381/transforming-nlp-a-look-at-new-tools-being-used-by-banks
1:00 – Joanna shares why she thinks mobile phones are the worst thing to happen to our society in the past 20 years.
4:30 – Then, they talk about AI conversational agents (or chatbots) like Google’s Meena, and Mitsuku.
7.30 – They discuss why Google hasn’t released Meena to the public for now, and the issue of bias encoded within AI models.
12:00 – What is a transformer model, and how has it helped with the development of the NLP field?
18:00 – How are data providers like Refinitiv and Bloomberg using transformer models like BERT (Bidirectional Encoder Representations from Transformers)?
24:30 – What are some limitations that exist when using transformer models?
28:30 – They talk about an interesting link (https://daringtolivefully.com/have-fun-at-work) that Jo came across, and how it would be fun to beat our editor-in-chief Tony in a foam sword fight.
38:30 – Wei-Shen and Jo wrap up by discussing how employers could provide satisfaction and fulfillment to employees in other ways, instead of following a how-to-have-fun-at-work type of list.
