Monday, July 31, 2023

Writing with Heart and Soul

Simplicity. Specificity. Suprise. Stirring language. Seductiveness. Smart thinking. Social Content. Storytelling. These eight S's are at the heart of "The Science of Strong Business Writing" by Bill Birchard in the Harvard Business Review

The article covers familiar territory but reinforces each point with contemporary research. It effectively describes why business writers would reach their readers effectively by starting sentences with subjects and verbs, limiting longwinded mid-sentence clauses, and using active voice, among other rhetorical standards. It is positively worth the 10-minute read. 

Monday, July 24, 2023

The Impact of Proficiency on Efficiency

If you have ever wondered, as I have, about the impact of language proficiency on work efficiency, read Mariana Lenharo's article "The True Cost of Science's Language Barrier for Non-native English Speakers" in Nature

Reading speed in general affects work productivity. The slower we read, the more time we need to do our jobs. Lenharo's reporting underscores how this fact amplifies for students and employees in science. The article quantifies the extra burden, citing many extra workdays needed for developing nonnative English speakers. I admire the additional effort these people put into their research, reading, and writing. This article epitomizes why they deserve our respect.

A special thanks to Dr. Mary Beth Mulcahy for bringing the article to my attention.

Monday, July 17, 2023

A Focused Reminder about Employing a Successful Writing Process

This blog has covered various aspects of the writing process many times over the past 18 years. (Click here to read them.) Many of those posts are covered well in Mark Rennella's "A (Very) Simple Way to Improve Your Writing" in the Harvard Business Review (January 31, 2022). It's the sort of article that will take you only 10 minutes to read but the rest of your life to practice. Happy learning!

Monday, July 10, 2023

More Phil Vassallo Dramatic Works Available

Performer Stuff, a website catering to actors, musicians, and dancers, has licensed six Philip Vassallo monologues:

Below are the links to Philip Vassallo plays with premiere and publication dates in parentheses:

The Author Makes No Difference (unproduced; published 2014)

Every Day's a Holiday (produced 2021; published 2015)

Everything Means Something Else (produced 2003; published 2004)

The Eye Begins to See (produced 2003; published 2014)

Family Secrets (produced 2005; published 2006)

Hurry Hurry: Twelve Dramatic and Comedic Sketches (produced 2016; published 2015)

The Phoenix (produced 2003; published 2007)

Questions (produced 2021; published 2020)

Questions Asked of Dying Dreams: Four Short Plays (produced 1992; published 2021)

So What If Life's a Cliché? (produced 2011; published 2011)

The Spelling Bee (produced 1993; published 1995)

This World and That One (produced 2021; published 2020)

What Are You Running For? (produced 1993; published 2010)

Monday, July 03, 2023

It's That Easy: ChatGPT

Since everyone seems to be talking about ChatGPT, I thought I would do an experiment of my own with OpenAI. If you are not sure about the meaning, value, or risks of ChatGPT, you read AI's Teachable Moment: How ChatGPT Is Transforming the Classroom for an educational perspective and Best Practices for Using AI When Writing Scientific Manuscripts for a science writer's viewpoint.

Here's the experiment. I logged into my OpenAI account and asked for "an article on the advantages and disadvantages of ChatGPT." What follows, in sans serif font, is not my writing. It is the text ChatGPT generated in seconds. I did not change a word. You judge.

Title: Exploring the Strengths and Weaknesses of Chatbot AI: A Look into ChatGPT

Introduction:

Chatbots have become increasingly prevalent in our digital interactions, offering automated conversational experiences across various platforms. Among these, ChatGPT, based on the GPT-3.5 model developed by OpenAI, has garnered significant attention for its ability to engage in meaningful conversations with users. In this article, we will delve into the strengths and weaknesses of ChatGPT, highlighting its capabilities as well as areas where improvement is still needed.

Strengths:

  1. 1. Natural Language Processing: ChatGPT excels at understanding and generating human-like text. Its advanced natural language processing abilities enable it to comprehend and respond to a wide range of queries, making conversations with users more intuitive and conversational.


  2. 2. Contextual Understanding: The model has the capability to retain and refer back to the context of the ongoing conversation. This contextual understanding enhances the quality of responses and allows for more coherent and relevant interactions.


  3. 3. Knowledge Base and Language Proficiency: ChatGPT benefits from being trained on a vast amount of data, allowing it to possess a wealth of information on a wide range of topics. It can provide general knowledge, explain concepts, and engage in discussions on various subjects, making it a valuable resource for users seeking information.


  4. 4. User-Friendly and Accessible: ChatGPT is designed to be user-friendly and accessible to a broad audience. Its simple interface and ability to integrate with various platforms make it readily available for individuals seeking conversational AI experiences.

Weaknesses:

1. Lack of Factual Accuracy: While ChatGPT can provide information on a wide array of topics, it may not always guarantee complete accuracy. The model's responses are generated based on patterns in the training data, and it may occasionally produce incorrect or misleading information. Users should exercise caution when relying on ChatGPT for factual accuracy and cross-reference information from trusted sources.


2. Sensitivity to Input Phrasing: ChatGPT's responses can be influenced by the way questions or statements are phrased. Slight rephrasing of the same query might yield different answers, leading to inconsistencies. This sensitivity to input phrasing can sometimes result in unexpected or confusing responses, which can impact the user experience.


3. Inability to Clarify Ambiguous Queries: ChatGPT may struggle to seek clarifications or request additional information when faced with ambiguous queries. It lacks the ability to ask for context or clarification, potentially leading to incorrect assumptions or misinterpretations of user intent.

4. Ethical and Bias Concerns: As with any AI model, ChatGPT can exhibit biases present in the training data, reflecting societal biases or propagating misinformation. OpenAI has made efforts to mitigate bias, but there is still work to be done to ensure fairness, inclusivity, and ethical considerations in its responses.

Conclusion:

ChatGPT, powered by the GPT-3.5 model, demonstrates remarkable strengths in natural language processing, contextual understanding, and general knowledge. It offers a user-friendly and accessible conversational AI experience. However, it is not without its weaknesses, such as occasional factual inaccuracies, sensitivity to input phrasing, and limitations in seeking clarifications. Additionally, ethical concerns and biases in AI models remain pertinent issues that need continuous attention. As AI technology evolves, addressing these weaknesses will contribute to enhancing the overall performance and user experience of chatbots like ChatGPT.