For some time now I've been trying at a plausible sort of "translation (or some other brain-affecting) microbe" concept. The main idea I'm basing this upon is of biofilms that could collectively work together to read electric fields for certain neural patterns, and if a certain pattern signature is recognized, the biofilm then works together to generate a specific pattern of electric pulses themselves and enact their own response effect back. (Think like a Brain-Computer Interface, but instead of an array of electrodes it's an array of similarly electricity sensing/generating microbes).

The actual function for this concept, so far, is this: the microbes all come together in a biofilm near the part of the brain that does whatever they're targeting, and connect to each other with nanowires to send electric signals so that they can share information and work collaboratively, with electric signals much faster than with chemical. Each microbe has its own sensor for the electric field around it, and the individual inputs all add up to create one big sum representing the whole field of electric activity they detect together. (Note that they're not as long-lived as neurons, so on an individual level the biofilm is in flux as single microbes degrade and join at an equal rate)

Here's the main tricky part: then, they have to have some mechanism to identify if that sum has a recognized neural word-pattern in it (e.g. the pattern for the intention to say "book", so they know to then respond with the pattern to translate that word/meaning). Since these microbes are all identical, at most having two or three different varieties that the huge population would be split among, the mechanism for this identification either needs to be something present chemically inside of them (e.g. protein complex for whichever individual microbe receives the end total "sum"?) or something that all of them come together to do and "understand".

So what's the most plausible physical (electric, biochemical, etc) mechanism, given these constraints, for a microbe biofilm to be able to rapidly--more or less fast enough to keep up with neurons--act kind of like a technological Brain-Computer Interface does, and do these functions:

  1. Adding up sensed electric fields across their biofilm
  2. Reading/identifying the total sum for a signature pattern, and starting a new signal to each other depending on what it is recognized as (if it is)
  3. Passing unique instructions to each other individual by individual for that new signal response

2 Answers 2


Do you absolutely insist on using microbes?

A better candidate is mono-cellular slime mold, which is already proven capable of transmitting electrical signals:

The slime mold is capable of sensing tactile, chemical, and optical stimuli and converting them to characteristic patterns of its electrical potential oscillations. The electrical responses to stimuli may propagate along protoplasmic tubes for distances exceeding tens of centimeters, as impulses in neural pathways do.


The benefit of using slime mold instead of a colony of microscopic organisms, aside from the one that slime molds already are proven to be able to "compute" in electrical impulses, is that a slime mold is essentially one continuous organism, and in fact can be even considered a single giant cell: this allows for the most efficient transfer of current (and thus, information) aside from using a cybernetic implant.

You do not need to design an "intelligent" primitive organism that uses electricity to think, slime molds already are that. All you need is to find a plausible way for the slime mold to send and decode signals faster than the brain, or at least be able to signal preemptively. This is not even a biology problem so much as it is chemical engineering problem: you need to have a mold filled with sufficiently conductive substances and increase the density of the structures the mold uses to modulate the signals. Luckily, both problems can be solved by pumping the mold full of various salts; which depending on type, can promote, inhibit, prevent or direct signal pathways forming. Similarly, adding sugar, various acids, and bacterial agents or agar can stimulate the mold to act and grow in specific ways.

TLDR: fill the brain with Physarum slime mold and pickle brine, wait for results.

  • $\begingroup$ Slime molds are microbes. $\endgroup$
    – Monty Wild
    Apr 12 at 11:54
  • $\begingroup$ @MontyWild I specifically mentioned Physarum slime mold which is monocellular, and definitely not "micro" as it can be a single cell the size of a dinner plate. $\endgroup$ Apr 12 at 13:45
  • $\begingroup$ Apologies for late reply, I was out for a while; I like this idea though! It seems to work with what I'm looking for a lot better but still retains the biological/organic aspect. My main questions before I definitely go with the slime mold idea are 1. Like the biofilm, can the slime mold still form from individual micro-cell organisms? since part of the idea was the "translator 'microbes'" getting to the brain area individually and congregating there... (cont.) $\endgroup$
    – inkwell87
    Apr 22 at 0:27
  • $\begingroup$ And 2. is there a way the slime mold "nervous system" can keep its own electric signals from affecting the actual brain neurons it is near? Like, is there some sort of membrane boundary for the slime mold tubes that could block the electricity from signals in them from "getting out" and affecting the potential and behavior of the nearby brain neurons, but still allow purposeful outward electric pulses to come through and affect the neurons when the slime mold is actually supposed to do so? (Or would the voltage for signaling in the slime mold be low enough that that wouldn't be a problem?) $\endgroup$
    – inkwell87
    Apr 22 at 0:30
  • $\begingroup$ Slime molds are effectively single celled, but if you grind them to bits, they just grow and reconnect if possible, or grow back from the biggest patch. You can start with a dust of "slime microbes" which will later grow connections, and become one giant cell. All the signals in in the slime mold are internal, within the cell, so there is no "leak" of electricity. As to how to get the slime mold to affect neurons: I would fill or coat SOME neurons with nutrients, so that slime mold would grow into them, making the ends of slime tendrils the only neural connections to touch neurons. $\endgroup$ Apr 22 at 6:42

This doesn't sound like a job for microbes, this sounds like a job for nanoengineered machines. You're trying to do things that single-celled organisms can't do, but that a designed machine or collection of machines might be able to do easily. Cells talk to each-other with slow chemicals. Making them talk to each-other quickly is hard. Cells don't do a lot of processing by themselves, they form neural networks for that. Capable cellular neural networks are big... we call them brains.

We're getting to the point where we will soon be able to build atomic-scale processors, processors so powerful that we'd probably be able to build the processing power of a human brain into an object the size of a single cell. We can make glucose-oxygen fuel cells. You can be far more sure that a machine that you design and is built by another machine cannot reproduce than you can be sure that a microbe that you modified won't mutate and infect your patient with some hideous GMO plague... or just make them sick.

So, you inject your patient with these machines. They use the patient's glucose and oxygen to power them. They move to where they're needed, they assemble themselves around the target area, orient themselves by which other machines they're next to, and they can communicate electrically with one-another and share processing power. It's then just a matter of tracking what the neurons are doing and decoding that.

Sure, these tiny machines might generate a bit of heat while they do their thing, but the human body is really good at getting rid of waste heat... an extra hundred watts should be fine, as long as they don't cook the neurons they're trying to analyse.

  • $\begingroup$ Not all cell-to-cell communication is chemical; electrical synapses exist. Granted, microbes don't form synapses, but there's no fundamental reason why they couldn't if you could convince them it's beneficial for them. Nanomachines might be better... but they're not what OP wanted. $\endgroup$
    – Cloudberry
    Apr 11 at 16:13
  • $\begingroup$ @Cloudberry OP may have wanted microbes, but realistically they're not going to be able to do the job. Nanites will suspend my disbelief far, far better than GMO microbes. $\endgroup$
    – Monty Wild
    Apr 12 at 2:56
  • $\begingroup$ where do you draw the line between "nanomachines" and "genetically engineered microbes"? $\endgroup$ Apr 12 at 9:50
  • $\begingroup$ @GoingDurden Quite simply, with GM microbes, you start with a microbe and... genetically modify it. With nanomachines, you assemble them atom by atom from the basic elements. With the first, you start with something and change it a bit until it does what you want. With the second, you start with nothing and build the tool you want. $\endgroup$
    – Monty Wild
    Apr 12 at 9:55
  • 1
    $\begingroup$ @MontyWild I understand that, but at the level of complexity we are talking about, you would either have to engineer nanomachines so complex, efficient, and biocompatible that they are essentially bacteria in all but name, or genetically engineer bacteria to the point that they are essentially purpose-built machines. You would simply converge in the middle between these two options, baceuase entirely "dry" nanomachines of ythis complexity are impossible, and entirely biological bacteria not good enough. $\endgroup$ Apr 12 at 10:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .