It sounds a bit like teamwork.

The arms are part of a team. They are given orders by the head, and report to it. But other than that, they know their job and can communicate directly for simple tasks.

I can already imagine.

- Brain: prey ahead, advance!

- Tentacles: yes sir

- Tentacle 1: I'm pulling on a rock, get ready tentacle 2

- Tentacle 2: I grabbed the rock, you can let go tentacle 1, tentacle 3, get ready

- Tentacle 3: that rock feels weird, alerting brain

- Brain: not a rock, danger, I am shooting ink, let everything go and start swimming

Local multiplayer game coming soon on Steam. :D

A maybe less known fact about octopuses is that they have 9 brains, one central and another in each arm.

“One of the big picture questions we have is just how a distributed nervous system would work, especially when it’s trying to do something complicated, like move through fluid and find food on a complex ocean floor. There are a lot of open questions about how these nodes in the nervous system are connected to each other,” said David Gire, a neuroscientist at the University of Washington and Sivitilli’s advisor for the project.

Given the success of modeling artificial neural networks after biological ones, I wonder if the results of this research will inform future development of distributed artificial neural networks.

I think the root of this is a misunderstanding of neurons in general.

Modern science seems to suggest that if you put a bunch of neurons in a room then it will start questioning quantum physics

Is there an understanding at all on how and when something becomes coordinated and how?

That’s not true at all. Deep learning struggles to learn even basic robotics tasks. Convolutional networks work well for image processing specifically because the network is structured in a way that’s effective for image processing. We don’t yet know how to structure and train neural networks for more advanced reasoning, not in a way that’s data efficient anyway.

I think Emergence is one of the interesting frontiers. How does information emerge from noise?

https://www.wnycstudios.org/story/91500-emergence

https://www.goodreads.com/book/show/8701960-the-information

Very little AFAIK. See, openworm for instance.

A fictional take on this that’s moderately entertaining is Adrian Tchaikovsky’s Children of Ruin. It’s a sequel to Children of Time which I liked a lot.

Came here to just say that. I liked Children of Time more [0].

[0] http://plurrrr.com/archive/2019/06/14.html

Since I learned about the Baader–Meinhof effect, I kept incurring into instances of it, like now: I'm just reading this book

I haven't learned about it yet, so this sort of thing never happens to me... =(

I had a similar idea a few years back for robotics that I jotted down in a note, I'll copy it verbatim here. (It's not rigorous in any way, just sort of a possible vision for the future.)

"For effective AI (for robotics specifically), we need to let separate, hardware driven, purpose built OSes run sensing equipment, and then only use a central OS fusing signals as a moderator/clearing house for those signals. Each separate OS should have autonomy over a certain set of functions to varying degrees, and should be overrideable only by strong consensus of the other systems. The important point is that the central OS not act like a traditional scheduler telling the camera and feelers and whatnot what to do all the time. The nature of the machine will instead be defined by complicated subsystems being moderated, but not fully controlled by a central authority, much like the brain/nervous system."

I wonder if there is any scientific evidence of human organ systems that operate without much input from the brain.

I read somewhere that that the human stomach also has a ton of neurons. So perhaps being an octopus is like being a human with eight gut instincts instead of one.

I knew [Cthulhu](https://en.m.wikipedia.org/wiki/Cthulhu) had to have more than one brain.