Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Thursday May 21 2020, @09:20AM   Printer-friendly
from the studying-praying-mantis-style-kung-fu dept.

Praying Mantises: More Deadly Than We Knew:

A praying mantis watches intently as a fly bobs by. In less than a blink, she's snatched it up. When the tape is played back in slow motion, we see the mantis pause and calibrate, almost like an experienced baseball catcher who has realized she's dealing with a knuckleball.

It's an impressive highlight reel. As detailed in a paper published this week in Biology Letters, it's also evidence that mantises strike less like automatons and more like active hunters, calibrating their attacks to more efficiently capture their prey as it flies by at different speeds.

[...] The strike of a praying mantis has two phases. In the first, the approach phase, a mantis extends its arms up and outward.In the second, the sweep phase, the mantis scoops the prey out of the air and pulls it in to eat.

Mr. Rossoni and Dr. Niven found that the mantises did indeed adjust their strike speed, according to how quickly the target was moving. Most of that modulation occurred in the approach phase — when presented with a slower target, the mantises would raise their limbs more slowly or pause in the middle, in a zombielike pose.

And if they initially miscalculated the speed of their prey, the mantises would often "correct their own mistakes" with a similar pause, Mr. Rossoni said. "Considering that some of the strikes are less than a tenth of a second, this is quite extraordinary."

It also adds to a growing conversation about what insects — from wasps that can logically infer to ants that can roll down inclines — are capable of.

"Historically, they were viewed more as almost miniature robots that were following very simple sets of rules," Dr. Niven said. "I think that there is new research coming out that suggests that that rule book might be much more complicated."


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 0) by Anonymous Coward on Thursday May 21 2020, @12:52PM (5 children)

    by Anonymous Coward on Thursday May 21 2020, @12:52PM (#997350)

    Even if bugs are just robots operating on instinct, the depth of instinct required to cope with the complexities of any environment is comparable to "higher" lifeforms. Have any insect brains been completely analyzed, like full map of the synapses level at least? I would imagine it would be very interesting to AI researchers.

  • (Score: 0) by Anonymous Coward on Thursday May 21 2020, @03:20PM

    by Anonymous Coward on Thursday May 21 2020, @03:20PM (#997408)

    The body is the vehicle. The consciousness finds itself in the vehicle. The limitations of the vehicle do not reflect the consciousness. The consciousness explores every available space.

  • (Score: 4, Interesting) by Immerman on Thursday May 21 2020, @06:37PM (3 children)

    by Immerman (3985) on Thursday May 21 2020, @06:37PM (#997490)

    I believe a housefly brain has been fully mapped to the synapse level, and I wouldn't be surprised if a few others have been as as well.

    Of course, insect brains are designed very differently than vertebrates, so there's probably far less benefit for human-like AI research than say, the rat hippocampus that's been similarly mapped.

    For example there's many "neuron equivalent" nodes per cell, presumably so they can pack a lot more "pseudo-neurons" into their tiny heads without the huge cell nuclei wasting a lot of space. And I think I've heard their brains actually remap themselves based on the current activity - e.g. where as mammals have dedicated brain regions for tasting, walking, speaking, etc - an insect may reuse the same brain regions for different tasks.

    Basically, pound-for-pound insect brains appear to be far more sophisticated than our own.

    • (Score: 2) by quietus on Friday May 22 2020, @02:55PM (2 children)

      by quietus (6328) on Friday May 22 2020, @02:55PM (#997868) Journal

      What do you mean with: insect brains are designed very differently than vertebrates?

      • (Score: 2) by Immerman on Saturday May 23 2020, @04:38PM (1 child)

        by Immerman (3985) on Saturday May 23 2020, @04:38PM (#998175)

        I gave some examples on the very next line...

        It seems as though vertebrates were able to "brute force" the development of more sophisticated brains, just adding more and more relatively simple neurons. Primates made a further evolutionary innovation by not scaling neuron size with body size, allowing larger bodies to support far more neurons - the neurons in other vertebrates scale with body size, just as all other cells do, so that you can (very) roughly gauge intelligence by the ratio of brain-to-body mass. In primates in contrast it scales roughly with brain mass alone.

        Insects though likely didn't have that option - or at least had to evolve an alternative when they shrank dramatically in response to falling oxygen levels in the atmosphere(I don't know if anyone knows when modern insect neurology actually evolved). So instead their individual neurons became more sophisticated, doing more with less as it were. I'm no expert on the subject, but the two I can recall hearing of are the examples I gave earlier:

        - Letting one neuron do the work of several (mostly?) independent neurons - Instead of the whole thing firing at once, individual ganglia "branches" can react independently, allowing far more "pseudoneurons" to be packed into the same volume since they can share a nucleus and other bulky cellular infrastructure.

        - Letting whole regions of the brain be dramatically repurposed to control very different activities. A vertebrate brain region mostly does one thing - process a particular sensory input, control the movement of a particular body part, etc. There's some overlap as things get interlinked to aid in other functions, but it mostly seems to be conceptually related stuff - for example much of the visual cortex actually seems to be more appropriately the pattern-recognition cortex, and also gets used in speech recognition, and I think symbolic thought such as math as well, among other things. If I understand correctly, in insects those structures can be repurposed to do very different things. As a made-up example, you might have the same brain structure control either flying or walking, depending on which "mode" it was in.

        • (Score: 2) by Immerman on Saturday May 23 2020, @04:42PM

          by Immerman (3985) on Saturday May 23 2020, @04:42PM (#998177)

          Whoops, just realized - where I said ganglia branches, I meant dendrite branches. Ganglia is something completely unrelated.