Comment by fwlr | Hacker News Reader

Comment by fwlr | Hacker News Reader

Parent Full thread

Jack Daniels isn’t meth. It is a complex and specific flavor profile that is enormously successful commercially. This profile was not arrived at via chemistry and biology, and it’s currently infeasible to apply chemistry and biology in a lab to replicate this flavor profile. The same is true of the overwhelming majority of other spirits, too. The only method we have to find your way to these extremely precise locations in high-dimensional flavorspace is to religiously follow the ritual that gets us there. On the road to mass production, spirit distillers have run across many pitfalls that look innocuous, minor changes that seem like they couldn’t possibly affect the outcome but nevertheless do. That’s what I mean by trying to crystallize luck. (Yes, chemistry and biology are the fields of science that _would_ describe this process…)

I'm always amazed that these products can be so accurately locked to an exact flavour for years, regardless of the inputs, which, being plant or animal matter, can vary a lot naturally between batches and between years.

Not only whisky, even things like branded fruit juice, McDonald's burger meat and salad, Nutella and branded chocolate, coffee, to name just a few, seem almost preternaturally consistent.

Juice companies extract the volatile components and them send them to perfumers who transform them into a specific flavor profile which is then added back to the juice. The flavor packs are derived from oranges so they don’t need to be labeled as an ingredient but provide a consistent experience.

> On the road to mass production, spirit distillers have run across many pitfalls that look innocuous, minor changes that seem like they couldn’t possibly affect the outcome but nevertheless do.

Sounds like engineering. They ran into a pitfall, which exposed a process that they can now reproduce in mass production. Follow that up with a hypothesis on why and a way to test it and you've got science. So it is a science.

I’m not trying to claim it is immune to science. I argued it is “luck, not science” in service of the point that it isn’t as easy as “just install some filters”, that carries a bigger risk than commenters might think. Jack Daniels should absolutely do some science to see if installing filters does affect their product. Unfortunately that experiment takes about four years to run, which means four more years of whiskey fungus for the residents.

Reply on news.ycombinator.com