Skip to content
Coordinates of ThoughtHome of Philosophy
Join
Question · Explain Newcomer-friendly

My rice cooker says ‘fuzzy logic’. Is that just probability with a nicer name?

Joel D'Souza Fellow
· 4 replies
Modal & non-classical logicAnalyticLotfi Zadeh#degrees-of-truth#fuzzy-logic#probability
Follow

Its coordinates: 4.3 Modal Logic · 4.3.1 Deontic Logic · 4.3.2 Epistemic Logic +4 more

We bought a Japanese rice cooker with "Fuzzy Logic" printed on the front, and it has become a running joke in the kitchen. The Filipino line cook says the rice is "fuzzy done". The pastry chef says everything in the kitchen is fuzzy logic.

But it made me curious. When the cooker decides the rice is 0.8 done, is that the same as saying there's an 80 per cent chance the rice is done? If not, what's the difference? I'm not a mathematician. I can do costings and scale recipes, that's about it.

LikeAnswer
Follow4 answers

Join the conversation: members reply here, and joining is free. Reading is always free.

Join free to reply Sign in

4 answers

Accepted answer

  1. Wanjiru Njenga

    ✓ Accepted answer Fellow

    They're different things, and your kitchen is a good place to see why.

    Probability is about uncertainty. "80 per cent chance the rice is done" means it's either done or not, and you don't know which.

    Fuzzy degree is about how far something has a property, with no uncertainty at all. "The rice is 0.8 done" means you can see it perfectly well, and it's mostly but not completely cooked.

    A quick test: a glass that is 80 per cent full versus an 80 per cent chance the glass is full. The first glass is in front of you. The second is in another room.

    The two also behave differently in the maths. In probability, "done and not done" has probability zero. In the usual fuzzy logic, if "done" is 0.5, then "done and not done" is also 0.5. That's deliberate: a half-cooked pot really is in between.

    Your cooker uses rules of the kind Zadeh's fuzzy sets were designed for (he introduced them in 1965): roughly, "if the temperature is fairly high and the water is mostly absorbed, lower the heat a little". The vague words get turned into numbers, the rules are combined, and the result becomes a single setting for the heater.

    Helpful · 1

Other answers

  1. Joel D'Souza

    Fellow

    The glass in the other room is perfect, I'm stealing it for the staff. So "is the rice done?" is fuzzy, and "will table nine like it?" is probability. That actually describes my whole job.

    Helpful · 2
  2. Derek Hollis

    Fellow

    Engineering detail for completeness: that last step Wanjiru mentions is called defuzzification. The controller reasons in degrees internally but has to output one crisp number, like a heater wattage. A lot of the art in fuzzy control is in how you do that conversion, not in the logic itself.

    Helpful · 3
  3. Bec Doyle

    Fellow

    Same thing in a winery. "Ripe enough to pick" is fuzzy: sugar, acid, flavour, all in between states. "Will it rain before Thursday?" is probability. Most of my bad calls come from treating the second as if it were the first, waiting for the grapes to be perfect when I should have been watching the forecast.