Become a fan of Slashdot on Facebook


Forgot your password?
AI Communications Network Networking Science Technology

AI Could Get Smarter By Copying the Neural Structure of a Rat Brain ( 89

the_newsbeagle writes: Many of today's fanciest artificial intelligence systems are some type of artificial neural network, but they bear only the roughest resemblance to a biological brain's real networks of neurons. That could change thanks to a $100M program from IARPA. The intelligence agency is funding neuroscience teams to map 1 cubic millimeter of rodent brain, looking at activity in the visual cortex while the rodent is engaged in a complex visual recognition task. By discovering how the neural circuits in that brain cube get activated to process information, IARPA hopes to find inspiration for better artificial neural networks. And an AI that performs better on visual recognition tasks could certainly be useful to intelligence agencies.
This discussion has been archived. No new comments can be posted.

AI Could Get Smarter By Copying the Neural Structure of a Rat Brain

Comments Filter:
  • Sounds promising but do we really need an AI that takes in garbage, hides in the darkest cramped spaces, efficiently drustributes viruses, and is a plague to humankind?
    • Who would run cat5 into the dark places for us if not our friends the rats?

    • by MrL0G1C ( 867445 )

      "hides in the darkest cramped spaces, efficiently drustributes viruses, and is a plague to humankind?"

      Is this a reference to CIA, NSA and the other secret services?

    • an AI that takes in garbage,

      . . . Facebook readers . . .

      hides in the darkest cramped spaces,

      . . . We call them collaborative cubicles . . .

      efficiently distributes viruses,

      . . . Emails from your friends . . .

      and is a plague to humankind?

      Ah, yes, humans' leading cause of death . . . other humans.

      Yep, rat-brained AI really would fit well in as a human brain.

  • "Data indicates that I ... err ... AI could get smarter by copying portions of a human brain. Please take a set and hold still while the probes are inserted into your cranial cavity."
  • by aberglas ( 991072 ) on Wednesday May 31, 2017 @07:15AM (#54516593)

    C Elegans is am extensively studied nematode with exactly 302 neurons, whose contetome (wiring) is consistent and known.

    But how its brain actually works remains a mystery. Neurons are complex, as is their interactions with the input and output.

    Not much point looking at mice with many orders of magnitude more first.

    Personally, I do not think that mapping neurons in detail will lead to AI. But if you are going to do it, start with something vaguely tractable. C Elegans.

    • by religionofpeas ( 4511805 ) on Wednesday May 31, 2017 @07:28AM (#54516633)

      Why not both, by different research teams ? There may be things that we can see in 50,000 neurons of a rat brain that we can't see in the 300 neurons of C.elegans.

      • by HiThere ( 15173 )

        I can understand your point, but also the gp's. Probably they should work with starfish and oysters before they devote too much study to rats. But rats are easier to raise in a lab, and come in genetically standardized strains. And there *is* a lot of pre-existing work on rat brains.

    • by Anonymous Coward on Wednesday May 31, 2017 @08:06AM (#54516749)

      Like Open Worm []?

    • by Anonymous Coward

      That seems like it might be a good point, but it's not really. After studying C Elegans, we've started to realize that it too simple and more evolved to the point where individual neurons are specialized...the magic is not in the network any more.

      By studying mammalian neocortex at this scale for the first time, both structurally (how are things wired) and functionally (how do things fire) we can begin to understand learning rules and data representations used in the brain to help constrain possible algorith

    • can only agree. not sure what mapping these neurons should help in understanding AI. the real question remains a logic one: Is it actually possible to recreate a human ai or is that actually beyond us. And wouldn't it be better to try a different sort of AI, not based on the way we think.
  • Adding rat brains can make AI smarter, huh?

    I guess that explains why the Borg kept wanting to assimilate humans.

    • Working out how real brains store information and learn new behaviours can definitely take neural networks far beyond the hand-concocted learning algorithms we currently use. Basically you point deep learning systems at the brain data and let that work out your learning algorithms for you, to make NNs that learn in whole new ways, which mean they can be used for entirely new classes of problems.

  • by turkeydance ( 1266624 ) on Wednesday May 31, 2017 @07:40AM (#54516669)
    is that even if you win, you're still a rat. Lily Tomlin
  • Is this some kind of way to pump money into IBM's Blue Brain Project ( [] )? As I remember, that project was trying to build a rat's neocortical column and was headed by Henry Markram and was funded by Swiss govt.

  • In the server room!!!

  • by Nidi62 ( 1525137 ) on Wednesday May 31, 2017 @08:29AM (#54516829)
    The Matrix, Terminator, 9, etc-all wrong. The war between humans and machines will really begin over cheese.
  • [] []

    First insects, now rodents? Maybe dogs, then dolphins, then humans?

  • How long until the AI gets scared of a loud noise, cats, bright lights and starts running along the wall, looking for a hole to crawl into??
  • Let's use it to model 1mm of a rat's brain instead. Who ever came up with this has a rat brain. They're just going waiste the first couple of million on EEG and shove correlations as proof for people that don't understand the experimental method, only those that care enough to retweet to look intelligent. As narcissist as this sounds, has any Slashdotter bothered to try to talk to normal people that aren't in their niche of intelligent friends? I'll save you some time, they are all as dumb as a box of rock
  • Get the genius brain, not the abbey normal brain.
  • Killer drones guided by mutant rat brains. It's only a matter of time.

  • this advanced facial recognition will literally rat on you?
  • There's nothing extraordinary about rats vis-a-vis humans. And its _human_ intelliigence we want to model, not rat intelligence, last time I checked!

    We don't need any Artificial Rats, though my cat might enjoy one.

  • I, for one, welcome our new rat-brained AI Overlords!

Were there fewer fools, knaves would starve. - Anonymous