͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏

Forwarded this email? Subscribe here for more

You’re reading the My News Biz newsletter, which I will be sending you every other Thursday. My goal is to help you and other digital media entrepreneurs to find a viable business model that works for you. If you were forwarded this email, you can sign up here.

Reasons for optimism #12: Jeremy Caplan and AI

He shows how to innovate and save time with tools such as ChatGPT

James Breiner

Mar 28

READ IN APP

You’re reading the My News Biz newsletter, which I send every other Thursday. My goal is to help digital media entrepreneurs find viable business models.

Why write about “reasons for optimism” when so much is going wrong in our politics, economy, and environment? Optimism gives us confidence that we can make things better. As I like to say, it’s another day of opportunity.

Jeremy Caplan is director of teaching and learning at the Newmark Grad School of Journalism at the City University of New York. He also does training abroad, such as this session in Germany.

I’m one of the 37,000 subscribers to Jeremy Caplan’s Wonder Tools newsletter, and a fan. We got to know each other when I attended some of his training sessions at the Craig Newmark Graduate School of Journalism, where he is a professor and runs the Entrepreneurial Journalism program. He's also a former Time Magazine reporter.

His newsletter delivers on a promise to help “journalists, teachers, creators, leaders and curious explorers” save time, frustration, and money by using the latest internet resources.

One of his most shared posts from 2023 showed subscribers 7 ways to use ChatGPT. While he points out many of the ways it can be misused and abused, Jeremy focuses on how artificial intelligence tools can help you.

Opposing views. One of my favorite use cases is to diminish blind spots. You could ask the bot to offer views that differ from yours about a TV show, a work of art, a particular writer, or a political position.

Writing help. Another is to get help with headlines. I used ChatGPT and two other chatbots to suggest headlines for one of my blog posts and was surprised at how good some of the suggestions were. I could pick and choose. But the bots need an editor. Human intervention is required.

Jeremy also suggests using it to “cut the flab” from a piece of your writing with a prompt like this one: “Suggest three ways to rewrite this paragraph in a more concise, compact way.”

Even more

Jeremy writes that if we use our imagination, there is almost no limit to the ways we can make the tools work for us.

Here are some of his suggestions:

Boost ChatGPT with new plugins. Each plugin is linked to an online service, “allowing you to plan travel, create diagrams, search contacts and more.” Access to the library of hundreds of plugins requires a $20 monthly fee. I’m thinking of paying for it. I’ve seen how some of them can analyze data and create impressive graphics and charts.
How to prompt effectively. Working journalists and academic researchers know that there is an art to asking the right question and getting a useful answer. Jeremy offers the POP method of prompting: “Persona—Give ChatGPT a role to play; Objective — Provide a clear goal; Parameters — Specify constraints.”
Generate ideas. Some might find this suggestion scary because it involves helping us with our own creative processes. (Will this mean we can be replaced?) But Jeremy shows that the tool can be helpful in “stimulating our thinking.” For example, suggesting questions to ask in an interview, topics to discuss in a video, or pitfalls to avoid in any kind of presentation.
Generate visuals. “ChatGPT Plus allows you to create virtually any image you can imagine,” Jeremy says. “You no longer have to use a separate image generation tool, nor do you have to master complex prompting jargon to create great images.”
If you enjoyed these suggestions, you can read more from Jeremy in his newsletter.

My multilingual assistants

One of my favorite uses of AI tools is for translation. I’m fluent in Spanish, but when I want to translate one of my newsletter articles into that language, I can produce a first draft quickly with tools like Deepl, Bard, Claude, Google Translate, or ChatGPT.

All of these tools have improved greatly in the past year, aided by the power of large language models to use enormous databases for their translations.

The AI tools are best at English-Spanish translation since the databases of documents in these languages are the largest. (English is the language most spoken in the world, and Spanish ranks fourth. Chinese and Hindi rank second and third, respectively.)

A test. I decided to test three translation tools on one of my own texts and found the results were similar. Even Google Translate, which I thought was a bit weaker than the other two, provided a translation that I found to be acceptable for most journalism uses.

I did consult two native speakers of Spanish on the fine point of whether to use the simple past tense or the past imperfect in two places. They differed, but one was more adamant that the past imperfect was preferred.

The weight of words. The hardest thing to translate is the emotional, historical, ideological, or political weight of a particular word or phrase in one language or the other. And those things change over time.

A good example: referring to someone as “an enslaved person” rather than “a slave” is a current trend in the US.

I think it’s a good change. It suggests that a person belonged unwillingly to a particular category of people and that there was an unjust system behind it. The question for a translator to Spanish or another language would be how to capture that subtlety in their own words.