> Make sure each cell has it's own data processing organelle.
Why de-localize this process?
Robustness, and because it'll have to be done sooner or later and we want the support to have been in from as early as possible.
> Connect the data processing organelles of neighbouring cells with some kind of low-latency but low-bandwidth signalling
> Create an unique identifier for each cell, a protocol for the assignment and handling of these addresses, and a way to mark data and pieces of *NA with them.
> add a system for keeping track of what data in in which cells, sending *NA pieces to a specified address by passing them from cell to cell, and requesting specific pieces of data through the speedy signalling system.
> Create a system for encoding sensory data for storage and distribution. Distribute important seeming sensory data to relevant cells by a subscription like system.
As in nerve cells? And something like a liver to control chemical balance?
> Create a memory system by tagging important sensory data with a timestamp, storing it indefinitely, and making it retrievable by timestamp matching.
> Create a system for solving complex problems faster by splitting it up into several smaller ones and having the data processing organelles of several cells work in parallel.
As in a brain? (Except ours would be slightly different to a normal brain as each of our cells can contain multiple neurons)
No, more like primitive routers/databases.
> Make sure each cell has it's own data processing organelle.
One is more than enough to process the data input, others would be a waste of time. If you want more power, it would be better to upgrade the existing one. Adding an new DPO to each cell would be redundant and use up a heck of a lot more energy. It would be better to link together each cell with extracellular node/connectors that interface in a network with each adjacent cell, ultimately connecting every single one.
They don't all need to be as good as our curent one, not active and consuming energy all of the time. Being redundant is exactly the point, so that everything is not lost if it gets damaged. Also, it might not be immediately useful but having done it this erly one will avoid a much harder conversion once we grow to big to be controlled by a single cell, and other systems that we DO need at the moment will have developed in the presence of this general architecture and thus work with it more smoothly.
> Connect the data processing organelles of neighbouring cells with some kind of low-latency but low-bandwidth signalling
> Create an unique identifier for each cell, a protocol for the assignment and handling of these addresses, and a way to mark data and pieces of *NA with them.
> add a system for keeping track of what data in in which cells, sending *NA pieces to a specified address by passing them from cell to cell, and requesting specific pieces of data through the speedy signalling system.
> Create a system for encoding sensory data for storage and distribution. Distribute important seeming sensory data to relevant cells by a subscription like system.
The cells are not exactly computers. While this is how the cell would accomplish these tasks, you don't need to come up with every bit of how it works. Most of this falls under upgrading the Data Processing Organelle's speed/connections. Also, all cells have all the ooze's genetic material in their nucleus. *NA is the same everywhere; different cells just read different parts.
I prefer to have some control over the exact implementation so that it fits together well with my future plans. And yes, that's how it works RIGHT NOW, because it has to work that way because a system for transporting the information doesn't exist. Once it does it'll be free of the constraint and able to diverge. Also, As we're going to start accumulating memories we don't want to store that large amount of alredy pretty redundant information in each and every cell.
> Create a memory system by tagging important sensory data with a timestamp, storing it indefinitely, and making it retrievable by timestamp matching.
> Create a system for solving complex problems faster by splitting it up into several smaller ones and having the data processing organelles of several cells work in parallel.
First part would be a specific upgrade/sub-organelle to the DPO, second part would be an upgrade of the existing DPO (by adding subunits and such). Adding more DPO's would be redundant.
Again, it being redundant at the moment is intentional.
> make those pseudopods much longer, thinner, stiffer, and add some glue like grasping proteins to the end. And maybe a simple silicon blade for mechanically cutting up membranes we can't cut chemically.
They are actually about as long as the cell's diameter at their current thickness. The thickness/hardness can be altered as stated at the top. The gluelike proteins might need more advanced ribosomes. If you want it to be much better, I would advise adding motor proteins for control/movement/force. That way, you have enough force to use the blade effectively.
Good idea. Yea, add motor proteins and a blade.
Poll
Author
Topic: Evolution: Origins [Sleek Futuristic Jellyfish][Sporadic Updates!] (Read 141437 times)