Nutrition science has long been built on a short list of familiar characters: calories, macronutrients, vitamins, a handful of minerals. Network scientist Albert-László Barabási thinks that leaves most of the picture dark.

In a new episode of Quanta Magazine's The Joy of Why, published November 12, Barabási explains how the same organizing principles he used to describe the internet and the cell can be applied to food. His "Foodome" project catalogs the molecules found in what we eat and asks what they do once inside the body, treating diet as a vast molecular network rather than a set of isolated nutrients.

The scale of the gap is the point. Barabási and colleagues have described the "dark matter" of nutrition — the many thousands of food molecules, well beyond the macro- and micronutrients captured in national databases, that science barely understands. Many have no known effect on human health; others may help or harm.

The promise, he argues, is that network science paired with artificial intelligence can start to identify which molecules matter — pointing to compounds that negatively affect health as well as those that could alleviate disease. It is an approach that reframes dietary advice as a problem of mapping relationships rather than counting ingredients.

Whether that can translate into actionable guidance for ordinary eaters remains an open question, but the episode makes the case that food, like the internet and the cell, is a network — and that how it is wired may matter as much as what is in it.