cd /news/ai-research/eating-out · home topics ai-research article
[ARTICLE · art-23577] src=futilitycloset.com pub= topic=ai-research verified=true sentiment=· neutral

Eating Out

In the late 1970s, physicist Richard Feynman scribbled a mathematical solution to help a friend decide whether to order a familiar dish or try something new at a Thai restaurant. Berkeley computational cognitive scientist Brian Christian and colleagues have now decoded Feynman's 50-year-old notes and published the optimal stopping strategy in the Proceedings of the National Academy of Sciences. An experiment with 2,520 participants confirmed that people naturally adapt similar linear thresholds, a simple approach nearly as effective as Feynman's original solution.

read2 min publishedJun 6, 2026

In the late 1970s, Richard Feynman visited a Thai restaurant in Glendale, California, for lunch with his friend Ralph Leighton. Leighton wondered whether he should order his favorite dish, the ginger chicken, or try something new. Feynman, on the spot, scribbled out a solution: If the ginger chicken didn’t exceed a certain high threshold, Leighton ought to try a new dish. But the threshold descended over time — on Leighton’s final visit to the restaurant, for example, it would make more sense to choose a meal he knew he’d enjoy rather than to gamble on an untested candidate.

Leighton kept Feynman’s notes, but his mathematical reasoning remained undeciphered for 50 years. Now Berkeley computational cognitive scientist Brian Christian and his colleagues have established Feynman’s argument and published it in the Proceedings of the National Academy of Sciences.

They also ran an experiment with 2,520 participants to see whether people actually follow this advice. They found that “people adapt linear thresholds used in optimal stopping tasks in a way that is sensitive to the underlying distribution — a simple strategy that we show is nearly as effective as Feynman’s solution.”

On this culinary theme: The “dining philosophers problem,” a puzzle in computer science, is described memorably in Wikipedia:

Five philosophers dine together at the same table. Each philosopher has their own plate at the table. There is a fork between each pair of adjacent plates. The dish served is a kind of spaghetti which has to be eaten with two forks. Each philosopher can only alternately think and eat. Moreover, a philosopher can only eat their spaghetti when they have both a left and a right fork. Thus, two forks will only be available when their two nearest neighbors are thinking, not eating. After an individual philosopher finishes eating, they will put down both forks. The problem is how to design a regimen (a concurrent algorithm) such that any philosopher will not starve; i.e., each can forever continue to alternate between eating and thinking, assuming that no philosopher can know when others may want to eat or think (an issue of incomplete information).

(Thanks, Sharon.)

── more in #ai-research 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/eating-out] indexed:0 read:2min 2026-06-06 ·