Meeting Report

ScienceWriters2023: Some Highlights for Editors and Others

Download Article

Conferences with online and in-person components have been taking varied forms. ScienceWriters2023, a joint meeting of the National Association of Science Writers and the Council for the Advancement of Science Writing, consisted of an online-only phase (September 26–October 3, 2023) and an in-person phase (October 6–10, Boulder, CO). The former focused almost solely on communicating science, and the latter included sessions on science. Individuals could register for the full conference or only the online portion. The current report presents highlights of some communication-related sessions from both phases.

Artificial Intelligence Tools: Promises, Prompts, and Pitfalls

By Madison Brown

The opening plenary session of ScienceWriters2023 focused on a topic of considerable current interest: the use of artificial intelligence (AI) tools in science writing and related realms. Moderated by freelance writer and editor Ellen Kuwana, the session featured a panel including journalists, an ethicist, and a technology expert. 

Kuwana began by asking how current AI systems differ from previous ones. The panel indicated that current AI systems (such as ChatGPT and DALL-E) are pre-trained for a variety of tasks, do not need supervision from their creators, and can be more creative than earlier systems. It was noted that ChatGPT is programmed to predict the next word in text, based on the dataset on which it was trained and information gleaned from users. 

Debra Mathews (John Hopkins bioethics researcher), Aimee Rinehart (Associated Press), Jeffrey Perkel (Nature), Melissa Heikkilä (MIT Technology Review), and Michael Madaio (research scientist, Google) spoke largely on the ethics surrounding AI, and how to determine whether something is AI-generated. Currently, Madaio said, it is mainly the user’s task to catch “hallucinations” (instances in which AI generates false information and presents it as fact). The panelists urged journalists to diligently fact check any information from an AI source, and to indicate to their audience that AI was used. They also noted that copyright issues are arising because AI tools are trained on datasets containing books and other materials without the copyright holders’ consent. Kuwana said the need to verify AI-generated content provided “job security for editors and fact checkers.” 

Another area the panel identified as being of concern was the reinforcement of racism and sexism, for example, when AI tools generate images based on their training with internet content. Heikkilä, who is part Asian, discovered that when she used an AI image generator to create pictures of herself, a large proportion were sexually explicit. In contrast, her White colleagues tended to be shown in empowering images. The panelists indicated that responsibility for fighting biases in AI lies in the companies creating the AI.

The panelists cautioned writers to avoid the “hype” of AI and be wary of humanizing this technology. Madaio recommended using the tool for what is best suited for: language-based rather than knowledge-based assistance. Rinehart said an example of suitable use was using AI to draft headlines as starting points to consider. Heikkilä said “excited but skeptical” is the best approach to AI. 

How Editors Can Advance Diversity, Inclusion, Equity, Accessibility, and Justice

By Barbara Gastel

Fairness in various regards has become an increasing priority for editors in the sciences. Accordingly, at this session, a panel of editors and writers discussed ways editors can advance diversity, inclusion, equity, accessibility, and justice. Moderated by Shraddha Chakradhar (deputy news editor, diversity, Science magazine), the panel also included Laura Helmuth, editor-in-chief, Scientific American; Ellen Kuwana, freelance writer and editor; Pakinam Amer, journalist based in Cairo; and Siri Carpenter, editor-in-chief of the science journalism resource The Open Notebook.

Chakradhar described the Science news section’s policies on diversity of interviewed sources. At Science, stories of at least 1000 words, which usually have 4 or more sources, must avoid homogeneous sourcing (such as only male or only White sources). Also, any story about a community must include a source from it. Reporter and editor brainstorm to identify suitably varied sources. “It’s really a partnership between the reporter and the editor,” Chakradhar said.

Others also addressed diversity. Amer noted that source choice should be based on merit, not tokenism. She also said that when local journalists collaborate on work by outside journalists, they need to receive credit, such as through shared bylines. Kuwana encouraged considering diversity of location and career stage. Noting that various issues intersect, Carpenter touched on considerations such as using gender-affirming language, avoiding ableist language, choosing and framing topics suitably, and selecting art. She said The Open Notebook website has many resources on diversity and related subjects.

Helmuth discussed equity in paying freelancers. She said editors should be proactive in raising pay rates rather than waiting until writers request increases. She and Carpenter emphasized that the pay should be the same regardless of where the writer is located. “Pay as much as you can afford,” Helmuth said.

Discussing podcasts, Amer said to embrace various accents rather than restricting speakers to those with standard American or British English. Ways to help speakers for whom English is difficult, she said, include sending questions in advance and being willing to re-record answers. If an accent is heavy, she noted, part of a recorded answer can be used and the rest can be paraphrased.

Other advice from the session included the following: 

  • Place expectations about matters such as diversity in writing, for example, in writers’ guidelines and assignment letters. 
  • Guide photographers and artists regarding diversity. 
  • Remember that different locations have different time zones, holidays, and days of rest. 
  • Create a culture of inclusivity at your publication.

A Word Problem: The Hows and Whys of Mathematical Communication

By Francesca Landon-Harding

Many people feel alienated from the mathematical world because of its dry numbers, complex equations, and technical language. This session’s panelists addressed how to navigate these issues to foster a sense of mathematical belonging in everyone.

What does it mean to communicate math? Noelle Sawyer, an assistant professor of mathematics at Southwestern University, broke it down. Drawing on audience answers, Sawyer characterized communicating math as explaining and applying mathematics without unnecessary numbers and jargon. Using “kernel” as an example, she noted that words can mean different things in math than in everyday life.

Sawyer also advised writers to use analogies and images. She said it is fine to use “the black box” rather than presenting every detail. For example, she noted, you do not need to know how to assemble an engine to drive a car. If there is no way around jargon, she said, do your best to define the terms simply. 

Examples are also powerful tools for effective mathematical communication. Sam Hansen, mathematics and statistics librarian at the University of Michigan, Ann Arbor, discussed splitting a cake as a running example to present mathematical approaches to dividing a resource fairly. For complex mathematical problems, Hansen suggested using just one example but solving it in different ways. Hansen suggested using multilevel examples to cater to varied levels of understanding, as has been done in Quanta Magazine. Hansen also recommended including images whenever possible, either making the images oneself or asking mathematicians to create simplified versions. Sawyer added that images give people the intuition and agency to engage with the material. 

Relating mathematical concepts to concrete experiences is another effective strategy. Kenna Hughes-Castleberry, public information officer at the physics institution JILA and freelance writer, said she has found herself lost in the math of quantum computing. She explained that seeing how the math applies to the world around her helped her get grounded. Hughes-Castleberry advised looking at the beginnings and ends of mathematical papers to find those real-world applications. When writing, she suggested phrasing things with “inaccurate accuracy”—in other words, in a way that is easier to grasp even if it is not as exact. For example, instead of saying “the population increased by 90%,” she advised saying “the population almost doubled.” Hansen advised starting pieces with the application to capture interest before introducing the math.

In closing, Hansen borrowed an idea from mathematician and mathematical mime Tim Chartier of Davidson College, advising writers to leave the reader with at least one good mathematical memory. Doing so can distract readers from dredging up doubt about what they might not know.

How To Build a Successful Internship Program: Lessons from the Field

By Sanjida Akter

Internships in science communications can benefit both intern and host. However, they require effort to succeed. At this session, 2 internship supervisors and 2 recent interns shared experiences and perspectives in this regard. 

Tom Ulrich, associate director for science communications at the Broad Institute of MIT and Harvard, directs the science communications internship there. He said this internship focuses mainly on mentorship, not productivity. He also said he tries to expose interns to varied activities, including those in science writing, social media, media relations, public engagement, and internal communication. 

Ulrich accompanies interns during their initial interviews with researchers and gives the interns feedback. In addition, he tries to adapt the internship to each intern’s interests. For example, he arranged for an intern interested in public engagement to include outreach work at a local school.

Sarah Lewin Frasier, who is assistant news editor at Scientific American and coordinates its internship program, said Scientific American has news, graphics, multimedia, and opinion interns. The news intern writes for the web and writes short articles for the print version of the magazine. Depending on their interests, news interns also can do other items, such as working on videos, podcasts, Q&As, book reviews, or infographics; fact-checking; and editing analysis and commentary pieces. Some former interns freelance for the magazine.

Ulrich and Frasier also offered guidance on creating internships. Ulrich said he looked at other internships for ideas and consulted members of his team about plans. He said he publicizes the internship through various channels, including organizations and social media. Frasier recommended identifying overlap between activities that would interest interns and tasks useful to the institution. She said that in justifying an internship, the case can be made that it is a good way to cultivate future freelancers.

The speakers noted that in virtual internships, time management can be more difficult for both supervisor and intern than if the intern is on site. They recommended having specific times for the intern and supervisor to meet.

Speaker Caleb Hess interned at Texas Sea Grant and is now communication manager for the Plant Resilience Institute at Michigan State University. He said the internship helped to create both strong writing skills and strong organizational skills. 

Jasmine Johnson, an MPH graduate of the health equity program at the University of North Carolina at Chapel Hill, interned at American Scientist magazine. Her internship was part-time. She said flexibility on both her part and her supervisor’s was helpful.

Both Hess and Johnson said impostor syndrome is an obstacle to seeking internships and functioning as an intern. Johnson said her mentor helped make her comfortable. “The relationship doesn’t stop after your internship,” Johnson observed. For example, internship supervisors can be of continuing help in finding opportunities. And the benefits can flow both ways. “It’s definitely worth it,” Ulrich said.

How to Make Science Comics (Even When You Can’t Draw)

By Abdurrahman Radwan

Comics can be a fine way to communicate science, but what if artistic ability is not in a science writer’s arsenal? This session focused on a solution: collaboration. Moderated by Sarah Zielinski, of Science News Explores, the session consisted mainly of a panel discussion featuring Abrian Curington (a visual storyteller), Maki Naro (cartoonist, illustrator, and science communicator), Matthew R Francis (freelance science writer), and JoAnna Wendel (media relations professional at Pacific Northwest National Laboratory and freelance illustrator).

In response to the question of which stories are best to be treated as comics, Curington said any scientific topic can work, but that a visual adaptation is needed, much as when a novel is adapted for film. Francis added that he and Naro look for items that can be good visual hooks, but that finding a hook in some physics-related topics can be tricky. He added that he seeks a social aspect that can be conveyed. 

The writers and artists also discussed their collaboration. “One of the things that we [editors and writers] really have to do is to learn how to distill, to learn what the important words are,” Francis said about choosing what to present in words and what to present in pictures. He and Naro described how they collaborate in this regard. Naro said he relies on Francis for technical knowledge and creates the visual aspect. Francis and Naro described the process as entailing considerable back-and-forth to create an integrated whole. Zielinski and Wendel described a similar partnership, with repeated exchange of ideas. “I love it,” Wendel said.

The speakers and attendees moved into breakout rooms for an exercise in applying concepts from the session. The exercise entailed visualizing how a comic could portray the recent story in which birds were incorporating anti-bird spikes into their nests. 

A question-and-answer period followed. In response to the questions, the panelists offered tips for thinking in images, listed outlets for publishing science comics, and commented on the role of artificial intelligence. Naro said use of AI tools to generate images is “basically plagiarism,” as the content is derived from others’ work.

A handout from the session is posted at

Brain Management for Freelancers* (*Not Just for Freelancers)

By Julianne Hodges

Being your own manager can be hard, especially when your brain does not want to cooperate. Two experienced freelancers shared their best advice for figuring out how their minds function to make their work go as smoothly as possible.

In “Brain Management for Freelancers,” the hosts discussed balancing the demands of freelancing with mental illness and neurodivergent conditions like anxiety and ADHD, although their advice can help anyone. Julian Anna Nowogrodzki freelanced for 6 years and now balances freelancing with being an opinion editor for Nature. Tara Haelle has been a health and science freelancer for over a decade. After the two presented their tips, the audience was invited to write their own suggestions on large pieces of paper with titles such as “Focusing” and “Asking for Help”.

“This is a session that I wish I had listened to when I was first freelancing,” Nowogrodzki said.

To refute discouraging thoughts, Nowogrodzki dons a metaphorical “boss hat” to be a compassionate boss to herself. She also maintains a “kudos folder” of compliments received. “So much of freelancing was just managing my own brain,” Nowogrodzki said. “How do I do my best work? What is missing, and how do I provide that for myself? If I’m both the employee and the boss, how can I be a good boss to myself?”

Nowogrodzki and Haelle also discussed asking accountability partners to check up on their progress periodically. They suggested posting a to-do list in a public place like Slack or Facebook. “Even if they’re not actually holding me accountable, just knowing that other people can see it is helpful,” Nowogrodzki said. Haelle added that if she cannot focus on a task after repeated attempts, it is more efficient to take a break than wasting time by getting frustrated.

Finally, Haelle discussed asking editors for accommodations. Now that she has an established career and has demonstrated her abilities, she has no problem asking for help; however, she did not know how to approach the topic early on. She shared a story about an editor she worked with while managing postpartum depression; unprompted, the editor noticed Haelle struggling and lightened her workload. “Editors: If you have the ability to pick up on those things, create a relationship with your reporters where they feel safe sharing things with you,” she said.

Haelle said conditions like ADHD can also be a “superpower”. For example, the way her brain works helps her think of questions that others might not consider. “My ADHD is why I’m a good journalist; my ADHD is also why I have a hard time being a journalist,” she said.

You’ve Got Mail: Newsletters in Science Journalism

By Christina B Sumners

Moderated by Christie Wilcox, newsletter editor at Science, this panel explored how email newsletters can help an individual or organization build and retain an audience.

Flora Graham, who writes the daily email newsletter for Nature, explained how the newsletter has changed her relationship with the audience. “They’ve invited me into their inboxes,” she said, and therefore, “people are expecting me to put them first.” She said she takes that responsibility seriously and tries to create the best newsletter she can for them. “In return, people reply, in their hundreds and thousands, in a very respectful way,” she said—something rarely seen in discourse elsewhere online. The newsletter also serves as one of the biggest traffic drivers to 

Freelancer Melinda Wenner Moyer produces a newsletter that she said provides slightly over a third of her income. She cautioned that building a newsletter audience to the point of profitability takes considerable time and a great deal of work. Still, she said, there are advantages over a blog or a website, mainly that newsletters “land in your readers’ inboxes whenever you want” and help build a brand.

Leigh Kamping-Carder, head of newsletters for The Wall Street Journal, agreed it is not easy, saying, “Newsletters always look like so much less work than they actually are.” She said she finds them satisfying, though, as they reach new readers and build the loyalty and engagement of current ones. She said it is also fun that the newsletter can have personality, rather than needing to follow the strict authoritative voice of a publication. “Within the four corners of the newsletter, you can experiment,” she said.

The panelists had some practical advice for new newsletter creators:

  • Pick a frequency (daily, weekly, or monthly) and stick with it. 
  • Do not obsess over the size of your subscriber list. “A number of people who subscribe to your newsletter have never read your newsletter,” Graham said. Meanwhile, wanting many subscribers disincentivizes keeping the list clean because there is a temptation to keep even outdated email addresses. Click-through rate—the number of people who click on a link within the newsletter—is the more important metric. 
  • Do not assume what time, day, or even season it is where readers are; starting with “Good morning and happy fall” would be disorientating for someone in Australia, where it is evening and spring. 
  • Remember that the design of the newsletter is as important as the text. 
  • Ask yourself how your newsletter meets a need for your audience and distinguishes itself from all others on that topic. “We try to really focus on the audience” and listen to their feedback, Kamping-Carder said. An effective newsletter can range from a list of interesting links to a collection of original feature articles, as long as it serves the audience.

Others’ coverage of some ScienceWriters2023 sessions appears on the National Association of Science Writers website ( and the Council for the Advancement of Science Writing website (


Madison Brown, Francesca Landon-Harding, Sanjida Akter, Abdurrahman Radwan, Julianne Hodges, Christina B Sumners, are graduate students, and Barbara Gastel is a professor, at Texas A&M University.

Opinions expressed are those of the authors and do not necessarily reflect the opinions or policies of the Council of Science Editors or the Editorial Board of Science Editor.