AI-generated podcast summarizes four recent articles
With advances in artificial intelligence (AI) making huge strides, podcasters look to be on the endangered species list. NotebookLM has cranked out a realistic – if not completely accurate – ten-minute discussion of four of my recent wine related articles exploring Champagne production techniques, accountable sustainability, lower-alcohol Cabernet Sauvignon, and novice wine education.
It was fascinating to see how the algorithem selected four topics from my archive, sussed out the key points of each article, and wove them together into a podcast that I found no more or less compelling than many I have listened to.
For the record, I don’t find podcasts a particularly good use of my listening time. While there are exceptions, I prefer reading to listening to other people audbily processing content.
That said, I encourage you read these brief articles –
Rare perpetual reserve a hallmark at Champagne Palmer
Solera wines are the secret sauce in Champagne producer’s blending process. In a tribute to the rare practice of aging reserve wines via the solera system, Champagne Palmer’s nonvintage Brut Reserve is now labeled La Reserve. The wine spends a minimum of four years on the lees, and the current release is defined by a structured backbone composed of 51% Chardonnay, 30% Pinot Noir, 19% Meunier, and up to 35% reserve wine.
Navigating the sustainability landscape
As the labyrinth of wine industry certifications continues to expand, consumers often find themselves without a compass. Nearly half of American adult drinkers of beverage alcohol (48%) say they are “positively influenced” to buy brands that have demonstrable environmental or sustainability credentials, according to IWSR’s 2021 research. As U.S. wineries rush to embrace sustainable business practices in response to the increasing value of sustainability in the marketplace, farming practices are often their lowest priority.
Sweet wines become a bridge to dry
Novice consumers learn to love wine by starting with sweet wine styles. After more than two decades of teaching Wine & Spirit Education Trust certification courses and academic wine classes, a new assignment at San Jose State University has provided the impetus for a sea change in my approach to educating wine-curious consumers.
Lower-alcohol Cabernet Sauvignon is back in fashion
In 2019, Jordan Vineyard & Winery published a tasting map that listed almost 40 Napa and Sonoma wineries known for making elegant, lower-alcohol wines with less than 14% abv. Among them were producers of Cabernet Sauvignon, such as Corison and Nalle as well as Jordan itself, that have remained steadfast in their dedication to a house style more influenced by the wines of Bordeaux than by the über-ripe, higher-alcohol Cabs that emerged from Napa Valley in the 1980s.
And then spend ten minutes listening to what AI thinks is the relevant take away from all four at DPW’s first AI-generated podcast. Other than the lack of introductions by the hosts, if you didn’t know it was AI, do you think you could tell?
Another example of an AI-generated podcast was posted by a colleague Felicity Carter who used NoteBookLM to generate a seven-minute overview of her archive site. Carter describes the segment as “cringy” but validating as it can be challenging for most of us to see the bigger picture impact of our daily work. As her body of work covers a myriad of topics often in great depth, the AI took an overview approach that lauded her accumen and insights. If you didn’t know Carter, the content would be, in effect, a compelling sales pitch.
Comments on her social media post to Facebook were generally positive with a friend who knows her saying, “The weird part is how scarily accurate it is while being full of things you would never say yourself. It’s actually a pretty good assessment. I’m low-key pleased there are a few tells in the podcast that it’s AI, but they’re pretty subtle.”
Not really for an old man who spent the better part of his life “boots on the ground”, and “boots” in the winery. Could give the AI lots of questions maybe “it” can answer them!