HackerTubewatch
Surprise me
Frame from Slime mold form a map of the Tokyo-area railway system Plays via YouTube's privacy-enhanced embed

Slime mold form a map of the Tokyo-area railway system

Harvard Magazine
Demos
YouTube
Cited on HN 5 times, 2017–2021, never outright recommended score 0
Years with an HN recommendation, 2007–2026

Referenced as a vivid illustration of explore-then-exploit optimization and decentralized network design.

A one-minute clip of slime mold placed on a map with food at the cities around Tokyo, growing a network that approximates the Tokyo-area railway system.

“This video shows a slime mold doing exploration/discovery and then exploitation” (HN, 2020)

5 citations · Demos

5 Hacker News comments

From the HN archive; up to 20 kept per video, favoring different commenters. Recommendations first.

@radiorental Oct 2017 Cites it
on HN thread Peter Thiel: Silicon Valley’s monopoly on big growth tech companies is over
Your description made think of the slime mold map of the Tokyo transportation system [1] albeit with a different set of inputs. As with any organic, self propelled, system.. it matures and growth turns to maintenance, and eventually decay. Is the general argument here that SV has become a victim of it's massive organic growth, has it eaten up one of it's resources (housing, healthy diversity, etc) to the point where another city/system can now flourish?? [1] [this video]
Slime mold growing a map of Tokyo railways
@djsumdog Oct 2019 Cites it
on HN thread Paris zoo unveils a slime mould
Slime mold is interesting. There's one species that, if you assemble food sources in the population densities found in Tokyo, will grow to look very similar to the Tokyo subway system: [this video]
Slime mold grows to resemble the Tokyo subway
@ta1234567890 Nov 2019 Cites it
on HN thread Relentlessly Simplify
Read more
This is a video of a slime mold finding food: [this video] It seems to be operating in something like 2 main phases: an initial exploratory phase in which it expands almost in every direction and then a second "optimization" (simplification) phase in which it just keeps the links between the food sources. I think we operate in a similar way. We learn by exploring and then we optimize when "exploiting". Everyone is different though, some like exploring more, others like optimizing more, and there's a big range in between. It also depends on the subject or activity. I personally almost exclusively enjoy the exploration part (learning about things), and don't have much interest in the optimization part (executing repetitively the same thing for gain).
Slime mold exploring then optimizing paths to food
@ta1234567890 Aug 2020 Cites it
on HN thread Young children would rather explore than get rewards
This video shows a slime mold doing exploration/discovery and then exploitation: [this video] Relating it to the article, it seems we function in a similar way to the mold, but our balance between exploration and exploitation changes as we grow up. Also reading other comments, it seems like different people have different balances or proportions between the exploration/exploitation modes. I personally am very imbalanced towards exploration; exploitation makes me demotivate and stop doing very quickly.
Slime mold shows exploration then exploitation
@ta1234567890 Apr 2021 Cites it
on HN thread Feynman: I am burned out and I'll never accomplish anything (1985)
Reminds me of this slime mold video: [this video] The mold “operates” in roughly two modes: exploring and exploiting. It explores to find the food and then it exploits the best paths to the food to consume it. I feel our minds work in a similar way, we explore when learning new stuff and then we exploit when using what we learned to accomplish something. Probably most of the work that majority of people do is mostly exploiting previously learned stuff, with some limited exploration sprinkled into it every now and then.
Slime mold explores then exploits best food paths

Video list and Hacker News comment labels from HackerTube (prompted by @karpathy), released under ODC-By 1.0, built on the open-index/hacker-news archive. Hacker News data is provided by Y Combinator; comments remain the work of their authors. Video titles, channel names and thumbnails belong to their creators and are shown via YouTube. HackerTube Watch is an unofficial edition. Data export of 2026-10-31.

Keyboard shortcuts

/
Search
j k
Next / previous video
r
Surprise me
l
Like (video page)
w
Watch later (video page)
?
This help
Esc
Close