Thank you, Dave! Yes, I saw this. Interestingly, it is being a touted as a way to prevent fraud, which I think is great. But, If you've got a 1985 DRC Romanee Conti magnum and you want to authenticate it using this technology, don't you have to open it? And that defeats the purpose.
Ah, I hadn't even thought of that use (but I suspect I haven't thought of *most* potential uses). I suppose you can open one out of a set? Then assume if one is authentic the rest will be, too? The part that got me excited was how the AI tech was confirming that different estates/terroirs had distinct chemical components that we'll be able to scientifically correlate to their flavor profiles. Essentially giving us an even greater minute understanding of wine aromas, mouthfeels, and tasting profiles, giving us something objective to line up with our subjective takes. We've always had pH and ABV and RSL, etc., but this opens up a whole new can of worms.
On the other hand, there is something here that speaks to the reality of Terroir. But while the AI sifts through the results of its analysis of the molecules, I don't know if the AI knows WHY or how or what caused them to be found in the wine.
Enjoyed your predictions. Genuinely curious (not trying to provoke), what makes you say the WHO data is untrue?
The most extreme possible conclusion is never correct. Moreover, the analysis applied lacks context and and alternative input.
I have been enjoying VinePair online and their podcast this year.
Love the podcast!
They truly speak out language
Good stuff, Tom, and congrats on wrapping up another year as a wine industry writer par excellence. This was one of the most intriguing articles on AI innovation in the wine world that I've read, I doubt you're unaware but thought I'd share anyway! https://www.newscientist.com/article/2406286-ai-can-tell-which-chateau-bordeaux-wines-come-from-with-100-accuracy
Thank you, Dave! Yes, I saw this. Interestingly, it is being a touted as a way to prevent fraud, which I think is great. But, If you've got a 1985 DRC Romanee Conti magnum and you want to authenticate it using this technology, don't you have to open it? And that defeats the purpose.
Ah, I hadn't even thought of that use (but I suspect I haven't thought of *most* potential uses). I suppose you can open one out of a set? Then assume if one is authentic the rest will be, too? The part that got me excited was how the AI tech was confirming that different estates/terroirs had distinct chemical components that we'll be able to scientifically correlate to their flavor profiles. Essentially giving us an even greater minute understanding of wine aromas, mouthfeels, and tasting profiles, giving us something objective to line up with our subjective takes. We've always had pH and ABV and RSL, etc., but this opens up a whole new can of worms.
On the other hand, there is something here that speaks to the reality of Terroir. But while the AI sifts through the results of its analysis of the molecules, I don't know if the AI knows WHY or how or what caused them to be found in the wine.