GenAI Helps Some But Doesn’t Score an A+ at Work
The Hispanic Public Relations Association Silicon Valley (HPRA SV), the National Association of Hispanic Journalists Bay Area (NAHJ Bay Area), and Northeastern University Silicon Valley hosted a panel discussion in San Jose about the rise of ChatGPT-like programs and the AI technology that’s powering them.
The discussion revolved around how the field of artificial intelligence has changed recently while also exploring the impacts, challenges, opportunities, and best practices in these industries.
I was a panelist along with Harry McCracken, global technology editor, Fast Company; Susan Gonzales, AIandYou Founder; Michelle Iracheta Redwood City Pulse Founder and Rasika Bhalerao, a Northeastern University instructor with a Ph.D. in computer science from NYU.
By the way, the campus ‘headquartered’ at 75 E. Santa Clara Street in San Jose is stunning with gorgeous and modern furniture and what looked like a fresh paint job with smart pastel colors like peach, battleship gray, and inspiring views, especially from the ninth floor.
Natural Language Processing Expert, Northeastern University Professor Kenneth Church, with his enviable three degrees from MIT and experience as an industry scientist and researcher, moderated.
The panelists reviewed the pros and cons of using ChatGPT in the newsroom. The journalists agreed that there is some usefulness to the application, but it won’t replace human research, writing, editing, and fact-checking. GenAI has to be used by responsible people with care, and if you are using it to write material to be published, it must be fact-checked.
McCracken does primarily in-depth research for Fast Company, which publishes six times per year. He added that media outlets that need to scale more stories might find GenAI more useful. I don’t operate like a CNBC, for example, which focuses more on the story that day, he said.
Journalists agree that stories are better reported when a human visits a place to meet people and ask questions, like at a crime scene. How can GenAI do that properly, the speakers asked.
Panelists cautioned the audience, made up of PR professionals, journalists, and Northeastern San Jose campus graduate students, that the massive use of GenAI for producing internet content will lead to harmful misinformation during the upcoming presidential election.
I suggested that everyone study what a fake photograph looks like. It’s hard to tell these days. The fakes have gotten much better, added Susan Gonzales. There are ways to tell like people next to one another look out of proportion.
Panelists urged audience members to spread the word and warn constituents to watch for misinformation.
I emphasized a limitation that hurts PR professionals. ChatGPT 4.0 and a few other chatbots I have been checking out this week won’t give any information past September of 2021. My job as a media relations pro involves almost all up-to-the-minute information. If I look up my co-panelist Michelle Iracheta via Chat GPT 4.0 app (which costs a pricey $19.99/month) it doesn’t recognize her news outlet, Redwood City Pulse,founded in October of 2021.
I mentioned that startups are using GenAI to write press releases, and the ones I’ve seen are brief and to the point, something reporters appreciate. “Some releases written by humans are 3,000 words long, which people like Harry McCracken do not like.” He shot the other panelist group a look of agreement.However, I added that I would rather write a news release than fill in the GenAI form. I’m not a lover of forms.
Panelists agree that despite the problems, GenAI can write okay and it is going to be useful in research, like when it comes to getting closer to disease cures. Ichareta and McCracken agreed that it was helpful for finding SEO keywords. Ichareta urged the group to be comfortable with technology to remain relevant to the workplace.
Professor Church smartly called GenAI more of a thesaurus, and most panelists agreed it was pretty good at writing stories and poems. I added, but its creative output might be plagiarized, so use it with caution.
Harry McCracken offered valuable pitch tips. He said to summarize points because he doesn’t have the time or energy to read a lot of information. He praised GenAI for its ability to write succinctly, adding that it finally learned how to write correctly.
In summary, while Chat GPT and similar GenAI tools can write and summarize topics well, its information is often massively outdated and may produce plagiarized or ‘alternative’ facts. So use it with caution, and don’t expect it to replace journalists or PR people any time soon.
Or, in my opinion, later for that matter.
PRSA SV Board Member Michelle McIntyre runs a 10 year old PR consulting firm and loves to blog. She grades ChatGPT 4.0 a C+.