Forgot your password?
typodupeerror
Science

Slime Mold Demonstrates Primitive Intelligence 40

Posted by Hemos
from the wow dept.
A reader writes "According to BBC News, scientists have just published a paper in Nature demonstrating that slime molds can negotiate the shortest route through a maze, thus demonstrating a form of "cellular computation" which implies a primitive intelligence."
This discussion has been archived. No new comments can be posted.

Slime Mold Demonstrates Primitive Intelligence

Comments Filter:
  • Hofstader is quite correct. You are also probably correct that intelligence exists because it improves survival. However, this does not imply that anything which survives and reproduces more is more intelligent; it is merely well-adapted. Consider the amoeba. It constantly reflows to engulf food and reproduce. Will we now say that ameobas also are intelligent, since you could do the exact same experiment with an amoeba (and get the same results, AFAIK) if you could build a small-enough maze?

    Yes, I do tend to hold to the idea that there is some magic thing called "self-awareness" which "true intelligence" requires. Don't ask me what it is --- I don't know yet. Humans have it. Humans can probably create it in computational systems. However (and IMHO), if it can't produce "cogito ergo sum", it ain't intelligence.
  • But that's really all it is --- a few closed-loop systems. It is my personal belief (and a requirement before I grant another organism the status of "fellow intelligent life-form") that you at least need to have some kind of generalized cognitive process going on. (See further downthread for a rough definition of what I'm looking for.)

    Self-reconfiguration as a response to stimuli can't be intelligence, because if it is, every single living thing on the planet displays rudimentary intelligence. (I make the assumption that most things are not intelligent. If you don't like that assumption, you are free not to use it, but I'll think you're being silly.)
  • I'm sure that the slime mold can discriminate between various food sources, selecting the one that is most favorable. Such behavior is present in unicellular organisms such as bacteria. In fact that have a complex regulatory cascade that selectively utilizes the best energy source available. Only when it has been fully consumed will the cell utilize the next best available food source.
  • However the judge threw it out when the sime molds that came out in support of the defendants showed a modicum of intelligence. Quoth the judge "The test of intelligence is enough of a distiguishing charachteristic that noone would mistake the slime mold for a lawyer."
  • Hey don't knock the Pokemon. You've obviously never seen Pixie. She's hotter than Aeon Flux and Lara Croft.

    --
  • I don't know if that's a true sign of intelligence, however. All you need is a mailing address to get an AOL account. However, it would be interesting to see slime mold manipulate a mouse...
  • by alephnull42 (202254) on Saturday September 30, 2000 @12:54AM (#748848) Homepage Journal
    From the article:
    "Toshiyuki Nakagaki of the Bio-Mimetic Control Research Centre, Nagoya, Japan,..."

    Is it just me or does this sound like the first 5 minutes of a Manga movie?
  • I think it's fairly obvious from the article that the author isn't using the same definition as you. I suppose that all he/she was trying to prove was that the slime mold was capable of more complex reactions in the presence of certain stimulii than was previously thought.

    I disagree with your current definition of intelligence... From here it looks like you view intelligence as a kind of a binary "either ya got it or ya don't" function. My personal beliefs are that while no, I wouldn't use the adjective "intelligent" in describing a slime mold, their behaviour shows certain aspects (ie problem solving and an inate desire to sustain itself). Parts of the meta-intelligence, if you will.

    Bear in mind, a TRULY unintelligent creature would simply bumble-fuck around, maybe finding food, maybe eating it, not caring much one way or the other. It's movement would be truly random, and it wouldn't be effected by either positive or negative reinforcement. Such creatures would almost certainly die out before they could reproduce, unless they were unbelievably resilient to starvation, predators, and intelligent monkeys with a tube of gynolotromin.

    If you aren't impressed by what the mighty slime mold can do, at least be impressed with the design. The thing can execute these complex actions as a system, and doesn't even have a central nervous system.

    -Loooeee teaches his slime mold to roll over and rot
  • Hey don't knock the Pokemon. You've obviously never seen Pixie. She's hotter than Aeon Flux and Lara Croft.

    Am I admitting too much when I say that she is on "Monster Rancher," not Pokemon? Or am I just admitting that I don't watch Pokemon?
    ___

  • You do have a point that intelligence has subcomponents, and that the drive to survive is one of those. OTOH, given that it's impossible to for anything to exist without somehow being designed to persist, I'm not sure that it's exactly a scientific discovery. (Lifeforms adapt to their environments! The sky is blue! Slashdot has trolls! Film at 11!)

    I'm impressed by the design of life in general. A cell is a pretty complex little machine. Whether or not this lends the watchmaker argument credence, I'm not sure. It's more that this mold is somehow being touted as special that bugs me. That, and the fact that when I see animal or software intelligence written about in the popular press, there are always implications of some kind of sentience or human-level ability. Basically, I'm reacting to what I perceive as fluffy hype.

  • ....as are ants, amoebae, and lightening.....
  • Before the race, food was everywhere (To grow it to size). They then put food at the two ends of the maze, and all of the dead end pseudopodia withdrew or died? It just sounds like "grow where the food is..." A skill microscopic life has had for quite a while.
    ___
  • Can you really call this intelligence? Soap bubbles, those darlings of the mathematical community, have long been of interest as they will always form the most efficient network between a set of nodes.

    You could call that intelligence, we tend to call it surface tension.

  • These tests having been completed, it was found that these slimes originate from Washington, D.C., and are closely related to those that can still be found in the Capitol Building and the White House.

    Said one scientist, "The incoherent language that these slimes started spouting as soon as the cameras were turned on them should have been an instant tip-off."

  • Yes, it chose the shortest path out of three possibilities. Not that it's any great intelligence, but it might have been expected to 1) choose any path at random, 2) not realize that the two destinations were the same pieces of food and choose all three paths. Somehow, the mold has some internal sense of location, and is self aware enough to realize when it encounters one of its own pseudopods (and not just another slime mold).
    ---
  • I mostly agree - All the thing did was stay in touch with food, and minimize it's surface area. Duh. That's what they do. I didn't read that any of the investigators used the word "intellgence", I think that is baggage of the air-head "science" writers...
  • Um, what's a Manga movie??
  • the question at heart is..."Will various Slime Molds seeking public office need additional regulatory or equal opportunity help to run for public office under current law?" or is their "de facto" qualification demonstrated by the fact that the vast majority of our elected and appointed officials act as though they equal to or lesser than Slime Molds. I see the potential for Great Confusion here, the Supreme Court may have to review and revise the definition of who (or what) qualifies for citizenship. Also, can fungi and algae be far behind this new liberation movement?
  • In answer to your question, it seems that a toxic substance on some maze walls would cause the slime to alter its body shape to avoid the toxin, though I'm sure tests are already being planned.

    Everyone seems to discuss intelligence as though it has some binary component: you do or do not display intelligence. Haven't we seen far more evidence that intelligence is a continuously evolving trait? All life is intelligent to some extent, while greater organization can lead to higher intelligence.

    Some might claim that human beings can be used as the measure of other species intelligence. However, even if that were true, human beings display such a range of intelligence levels themselves, such a measure could never be useful! Besides, if you believe that human beings represent the ultimate accomplishment in biological intelligence, you oviously haven't been to enough sports venues.
  • by CentrX (50629) on Wednesday September 27, 2000 @01:31PM (#748861)
    Well they're hell of a lot smarter than me. I can never figure out those maze thingies.
  • by Anonymous Coward
    From "Gee that was a Yummy SLIME MOLD" To "The SLIME MOLD Hits! You Die... -more-" ...
  • by Tairan (167707) on Wednesday September 27, 2000 @02:02PM (#748863) Homepage
    In other news, it was discovered that the same slime molds could possibly be smarter than the humans studying them. It was discovered that most humans complete mazes by trying every possible path and then backtracking, or using the time honed method of placing one hand on the right wall and continueing until the end is reached.

  • by pb (1020) on Wednesday September 27, 2000 @02:11PM (#748864)
    I didn't want to see any more articles about the RIAA today!
    ---
    pb Reply or e-mail; don't vaguely moderate [ncsu.edu].
  • by gorilla (36491) on Thursday September 28, 2000 @04:13AM (#748865)
    The bar association has complained that the slime molds could be mistaken for lawyers. In a suit filed in federal court the bar claims that slime molds are "slimy, disgusting, and unwanted, all attributes traditionally associated with lawyers."
  • by Alik (81811) on Wednesday September 27, 2000 @04:35PM (#748866)
    So this thing has demonstrated intelligence because it reshaped its body to move most of its bulk as close to the nutrient source as possible? I'm not mad at the BBC or Slashdot, but I can't believe the scientists who call this intelligence. You could do the same thing with a single neuron and the appropriate mix of growth factors. (At least, given that both axon growth cones and slime mold food-seeking work on a sense-molecule-grow-tubules model, it seems pretty likely to me that you could achieve the same with a single cell.) This doesn't mean that those cells are intelligent.

    I'll grant the idea that you could somehow do computation using a mobile mold as your switching unit. I fail to see why anyone would bother (especially given that slime molds are icky :-) but you could do it. That still doesn't mean it's intelligent. (I'm not even sure one could say that the mold itself is really *computing*, but I suppose the network of cells and integrated signal cascasdes does encode some function.)

    Yes, I'm one of those biased morons who thinks that you've got to demonstrate something like a sense of self or use of language to qualify for intelligence. Then again, given the human race, perhaps the bar's been lowered. :-)
  • ... and that might represent a net icrease in intelligence (!)
  • by jellisky (211018) on Wednesday September 27, 2000 @06:12PM (#748868) Journal
    I don't understand this fascination about biologists trying to prove intelligence through things "figuring" out mazes. Any person with a basic programming knowledge can solve a maze using exhaustive methods.
    Also, it sounds like they ran this experiment once. As the great fortune program will tell you: If reproducibility is going to be a problem, conduct the experiment once.
    I'll be convinced of the intelligence of this slime mold if it, and lots of its relatives, can do a maze without resorting to simple or exhaustive methods, or until the researchers can figure out a better test than exhaustive methods to prove intelligence. Until then, slime mold is a mostly unintelligent fungus-like growth that can solve a maze, like everything else that can move can do...
    Consider me a skeptic.
  • A giant slime was found living in Redmond,WA with its spouse and offspring. The specimen voided all scientific proof of primitive intelligence as soon as we pointed our cameras upon the creature, which immediately started conversing in a seemingly incoherent language tentatively named FudSpeak. Further tests will be performed to find out where these slimes come from.
  • by Tommi Morre (235789) on Wednesday September 27, 2000 @06:45PM (#748870)
    The researchers believe the slime is exhibiting some form of primitive intelligence.

    Obviously, it's been posting as Anonymous Coward.

  • Soon it will be like Fred from GPF [gpf-comics.com] - a good read - better than UF lately, that's for sure.
  • At the very least, when you eat a slime mold,
    your Wisdom should go up a point.

    Sardonica the Aspirant
  • If you take the premise of the article and say that survival equals intelligence, then that implies the evolution is intelligent. This idea is touched on in Akira, but I choice not to believe this one. As the universe ages is edges towards complexity, so more complex systems look more and more intelligent. The illusion of intelligence or intelligence itself is just s side benefit of complexity.

    For more on complexity I would read "Complexity: the Emerging Science At the Edge of Order and Chaos" by M. Mitchell Waldrop. We read this after reading "Chaos" by Gleick. These books and the class change my perspective on the universe and changed it from a static thing to a wonderful complexity.
  • I think that we need to take a step back and look at what we (and this article) define intelligent to be. The slime mold has clearly demonstrated a problem solving ability - albeit to a very simple problem. While I wouldn't cheat off of this slime mold on my next calculus test, I would grant that it's behavior shows some signs of a rudimentary intelligence. I think that it would be interesting to see how "intelligently" a slime mold could handle more complex problems... what if portions of the maze were laced with a substance that was toxic to the mold? Would it avoid them successfully, or would it kill itself trying to get to the food? What if by going through a more circuitous route, the mold could reach a different type of food - one with more nutrients that would allow it to grow faster? How quickly can the slime mold to rearrange itself for these configurations? Can it "learn" to prize one food source more than another? Again, this may not be displaying "intelligence" in the cut-and-dried sense that you are mentioning, but it would be a display of how well the various closed loop control systems that guide the slime molds interact with each other (for example, in the poisoned maze example above, how successfully do the "find food" behaviour mechanism and the "avoid danger" mechanism interact?) But then again, maybe you can view all intelligence as a sum of how efficiently one's stimulus-detection-response closed loop feedback systems interact. -Loooeeeee steps around the bungee pit and eats a steak dinner
  • by hucke (55628)
    ...slime molds will be getting AOL accounts.
  • Apparently my editor doesn't like white space...sorry about that! ;D

    Here it is again, easier to read.

    ________________________________________________ _

    I think that we need to take a step back and look at what we (and this article) define intelligent to be. The slime mold has clearly demonstrated
    a problem solving ability - albeit to a very simple problem. While I wouldn't cheat off of this slime mold on my next calculus test, I would grant
    that it's behavior shows some signs of a rudimentary intelligence.

    I think that it would be interesting to see how "intelligently" a slime mold could handle more complex problems... what if portions of the maze were laced with a substance that was toxic to the mold? Would it avoid them successfully, or would it kill itself trying to get to the food? What if by going through a more circuitous route, the mold could reach a different type of food - one with more nutrients that would allow it to grow faster? How quickly can the slime mold to rearrange itself for these configurations? Can it "learn" to prize one food source more than another?

    Again, this may not be displaying "intelligence" in the cut-and-dried sense that you are mentioning, but it would be a display of how well the various closed loop control systems that guide the slime molds interact with each other (for example, in the poisoned maze example above, how successfully do the "find food" behaviour mechanism and the "avoid danger" mechanism interact?)

    But then again, maybe you can view all intelligence as a sum of how efficiently one's stimulus-detection-response closed loop feedback systems interact.

    -Loooeeeee steps around the bungee pit and eats a steak dinner
  • Slime Mold Gets Slashdot Id
    Earns +2 Karma bonus in record time.
  • So this thing has demonstrated intelligence because it reshaped its body to move most of its bulk as close to the nutrient source as possible?

    Umm, when you get down to it, isn't that pretty much all that intelligence does? That is, isn't it the prime goal of intelligence to move closer to food to increase survival rates and be able to reproduce more than the less intelligent competitors?

    I wouldn't be quick to state that the slime mold is intelligent, but I wouldn't be too ready to write off its accomplishments.

    After all, Douglas Hofstadter makes the point in Godel, Escher, Bach: An Eternal Golden Braid [barnesandnoble.com] that when dealing with artificial intelligence, the common definition of AI is whatever we can do that computers haven't accomplished yet. That is, once a computer beat a human at chess, then a lot of people quit using that as an indicator of AI. Let's not make that mistake here; give the little chap his credit!

  • In a similar experiment, scientists put Monica Lewinsky and President Clinton's legacy at seperate exits to a maze. In multiple tries, Clinton chose Lewinsky each time, proving once again that not all slime is intelligent.

  • Sure any programmer could colve this, but the programmer IS intellegent. The computer isn't intellegent, it's the programmer all along. Well the slime CAN do what you can do, so I think it's impressive.
  • Just drop food in random places on a mousepad that has a wireless, waterproof mouse between the slime mold and the food:)

Saliva causes cancer, but only if swallowed in small amounts over a long period of time. -- George Carlin

Working...