LSI SciComm Series: Sophie Bushwick

Forum Hall, Palmer Commons

Audience This is a public event.

I'm afraid I can't do that: The promise and pitfalls of AI for science communication

When the AI chatbot ChatGPT went public just over a year ago, interest in generative artificial intelligence exploded. Enthusiasts touted AI's ability to turn anyone into a writer or an artist, to make workers more efficient and companies richer—that is, if it didn't become powerful enough to destroy humanity altogether. Skeptics pointed out that AI language models spout falsehoods and facts with equal confidence, can churn out disinformation too quickly for fact-checkers to keep up, and threaten writers' jobs and livelihoods.

In science communication, where accuracy and understanding are priorities, is there any place for AI tools? Sophie Bushwick has been covering this technology through the generative AI boom, exploring how artificial intelligence models work and how they're impacting a variety of fields, including medicine and education. She will draw on that experience to discuss AI's strengths and flaws, and what scientists and science communicators should consider when deciding how best to use it.

Speaker

Sophie Bushwick

Science and technology journalist

Sophie Bushwick is a science and technology journalist based in New York City and is currently working as senior news editor at New Scientist. She has more than a decade of experience as a freelance writer and staff editor at outlets including Scientific American, where she covered the recent generative AI boom, Popular Science, Discover Magazine and Gizmodo; and she continues to make regular appearances on Science Friday. Her work spans digital and print, podcasts and radio, TV news and TikTok. In her spare time, she likes to read, run, learn about the evolution of fairy tales and take in as much theater as possible.

Hosts

Life Sciences Institute

School of Information

Science, Technology and Public Policy program, Ford School of Public Policy

Michigan Institute for Data Science (MIDAS)