Caution: AI Inside

August 2026

People often ask me if I use AI to put together NGRA’s monthly newsletter. The answer is no. To the degree that it’s possible to produce a digital product manually, that’s how it’s done.

The newsletter is the result of all the grape-research-related things I’ve seen, read or experienced throughout the preceding month. Basically, if I found something interesting (or it seems like others did), I figure you might, too! I write substantially all of the article summaries, highlighting the scientific aspects of each piece. Two wine writers—Kathleen Willcox and Robin Shreeves, whose bylines you may see in the popular wine press and in communications from the New York Wine & Grape Foundation—help proofread and format the newsletter for publication each month.

Have I tried using AI? Of course. I’ve fed some articles into an AI app to see how well it could summarize them. I was deeply disappointed—and disturbed—by how many things it got things wrong…and made up. For example, it assumed an article about swarming drone technology was about bees! I realize that some of AI’s failure here has to do with user-error—e.g., I could probably learn to give it better prompts. But there will always be the need to check its work, which seems just as time-consuming as writing the copy myself. As it stands now, there’s no way I could use AI to produce a publication that wouldn’t embarrass us or betray your trust as a reader. (Never mind the stark warning Bill Gates issued this month regarding the impact AI will have on society and the world if left unchecked.)

But I do know others using AI to great effect. Grape pathologist Katie Gold, ag engineer Yu Jiang and their team at Cornell have created the Aleks agentic AI system to “integrate domain knowledge, data analysis and machine learning…to autonomously conduct data-driven scientific discovery.” They view Aleks as an “autonomous collaborator” helping to enrich and accelerate their research. “Once provided with a research question and dataset,” they explain, “Aleks iteratively formulated problems, explored alternative modeling strategies, and refined solutions across multiple cycles without human intervention.”

And check out this short video, “How I Use AI in Extension Work,” by extension enologist Andreea Botezatu at Texas A&M Agrilife Extension. She explains how she uses AI tools to transform complex technical research into accessible resources like fact sheets, podcasts and visual aids for her wine-industry stakeholders, “reducing friction in communication.” Note that it’s impossible to tell if she’s narrating the video or if it’s AI. (Watch the clip to find out.)

Two grape growers I talked with this month said they love the NGRA newsletter. “But it’s a lot,” they admitted. So, when another colleague told me that, rather than reading long documents she needs to be conversant on for meetings, she uses an AI app to give her an auditory brief while walking her dog, I had a thought. Maybe there’s another way AI could help with the NGRA newsletter—not to make it easier to produce, but to make it more palatable to consume.

So, in the August edition of the NGRA newsletter, navigate to the “In the News” section (or just click here) for a beta test of a podcast-style summary, about 23 minutes long, of the science headlines we cover this month. If you’re a farmer who’d enjoy geeking out to grape research news while you’re driving the tractor, this functionality is for you! Will this feature be available every month? Will we try a different flavor of AI next month—like straight narration vs. this conversational style? Will we use AI to voice more—or all—of future editions? It all depends on you and your feedback! So, take a listen and let me know what you think!

Donnell Brown
President