Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How insects like bumblebees do so much with tiny brains

bbc.com

Post reply on HN
Print view View original

31–40 of 143 posts

Re: How insects like bumblebees do so much with tiny brains

#31

Post by Tenhundfeld »

JohnGoGoGo wrote: post #16

"With just a few hundred or thousand neurons, you can easily recognise perhaps a hundred faces". It makes me think we are missing something when creating arificial neural networks which needs much more neurons to achieve only this specific task. Maybe artificial neurons are too simplified models compared to biological ones, maybe our training process could be much more efficient?

Reminds me of a fascinating talk by Gerald Sussman at Strangeloop.

We Really Don't Know How to Compute! https://www.youtube.com/watch?v=O3tVctB_VSU

Re: How insects like bumblebees do so much with tiny brains

#32

Post by JackFr »

It's astonishing to me that on the one hand the article upends what had been common sense with experimental results, but then blithely goes on and makes assertions about the nature of cognition based on no other justification than it being the au courant model.

Re: How insects like bumblebees do so much with tiny brains

#33

Post by acqq »

pc86 wrote: post #24
acqq wrote: post #20

Earlier quoted context omitted.

> What should I take away from this? Just that you should not omit with "..." the most important part of the statement for you. You haven't studied physics. Then you see the formula and the textual name of it with the most probable entry in Wikipedia and then you throw the most useful bit for you away and ask what the formula without the name means, even though the name would give you an explanation. pc86's quote: >…

I assumed there was some relation between k T log 2 joules and 20 Watts, probably giving some sort of context to how easily or not easily information in our brains could be lost. The theoretical principles behind the amount of energy required to erase one bit, while interesting, doesn't necessary matter for that relationship, does it?

There is a relation but it is not too important in practice, which you'd find if you'd attempt to read what you avoided... twice.

The theoretical information change limits are theoretical, in practice, there's always significantly more energy used, so 'pizza didn't even have to mention that limit in this case (although knowing that the limits exists is also important in considering what is possible and what simply is not).

The wikipedia article you ignored shows that in the formula T is temperature in Kelvins, i.e. 310 K for a body temperature, and k is of order 1-e28. You get some small energy, log 2 is not relevant for the order of magnitude. Then you need to use time to get just the power in Watts, and which time should you use? Even looking at extreme processes: does killing somebody "erase" all the memory at once? Certainly not, as the processes in the brain are much more complex.... but the destruction of just a pieces of a brain makes it not functioning at all, like all complex systems, etc. And it the end it would be still not much relevant to hang to the formula for any known system, either developed by evolution or by humans.

Also once you google the name of the limit you get a lot of articles, not just the one in Wikipedia. Ditto for the brain power usage.

Re: How insects like bumblebees do so much with tiny brains

#34

Post by throwitawayday »

chongli wrote: post #29
JohnGoGoGo wrote: post #16

"With just a few hundred or thousand neurons, you can easily recognise perhaps a hundred faces". It makes me think we are missing something when creating arificial neural networks which needs much more neurons to achieve only this specific task. Maybe artificial neurons are too simplified models compared to biological ones, maybe our training process could be much more efficient?

If I had to guess: neural networks have to operate on pixel data whereas real neurons don't. Brains and eyes have evolved in tandem. Perhaps what makes them so efficient is that the eyes handle some of the processing as a consequence of their physical shape and characteristics. Look at the eyes of bees. Very different from our own (and from the cameras we build) and perhaps very specialized to the limited set of task…

On eyes in general: the retina is not like a camera pixel array, rather, it extracts all kinds of features and detects motion:

"Eye smarter than scientists believed: neural computations in circuits of the retina." Gollisch, Tim, and Markus Meister. Neuron 65.2 (2010): 150-164.

http://www.sciencedirect.com/science/article/pii/S0896627309...

Re: How insects like bumblebees do so much with tiny brains

#35

Post by adrianN »

pizza wrote: post #3
midgetjones wrote: post #2

Maybe the real question is 'why can we do so little with our giant brains?'

Maybe something along the lines of "our brains would overheat" - there's a small temperature window in which proteins won't denature, it takes k T log 2 joules to erase one bit of information (Landauer's principle), and our brain uses around 20 Watts of power. Maybe tin foil hats make good heatsinks.. Interestingly enough there is some evidence that the Gibbs free energy of ketone metabolism is more thermodynamically…

Blood seems to be a pretty decent cooling agent. Our muscles can generate a lot more waste heat and we manage to get rid of it.

Re: How insects like bumblebees do so much with tiny brains

#36

Post by k__ »

midgetjones wrote: post #2

Maybe the real question is 'why can we do so little with our giant brains?'

Can we do so little?

I think not.

I see a brain like a FPGA or a programming language, it can become anything, but it has to be led in the right direction.

Like, some people did Facebook with PHP and some people hacked together one in a million CMS.

Some people wrote a OS in C and I wrote a asteroids clone.

Re: How insects like bumblebees do so much with tiny brains

#37

Post by pavel_lishin »

joshmarlow wrote: post #30

Highly relevant is the Portia genus of spiders ([0], [1]) and apparently other related jumping spiders ([2]). One of my favorite excerpts from [1]: > Harland says Portia’s eyesight is the place to start. Jumping spiders already have excellent vision and Portia’s is ten times as good, making it sharper than most mammals. However being so small, there is a trade-off in that Portia can only focus its eyes on a tiny spot…

Portia spiders are also referenced in Peter Watts' Echopraxia. (Some spoilers could be inferred from thinking about this too hard while reading the novel.)

Re: How insects like bumblebees do so much with tiny brains

#38

Post by joshmarlow »

pavel_lishin wrote: post #37
joshmarlow wrote: post #30

Highly relevant is the Portia genus of spiders ([0], [1]) and apparently other related jumping spiders ([2]). One of my favorite excerpts from [1]: > Harland says Portia’s eyesight is the place to start. Jumping spiders already have excellent vision and Portia’s is ten times as good, making it sharper than most mammals. However being so small, there is a trade-off in that Portia can only focus its eyes on a tiny spot…

Portia spiders are also referenced in Peter Watts' Echopraxia. (Some spoilers could be inferred from thinking about this too hard while reading the novel.)

Yep! Watts' novels are how I first heard about Portia. I can't wait for the third in the trilogy.

In general, I love a good novel with a bibliography.

Re: How insects like bumblebees do so much with tiny brains

#39

Post by brutos »

pc86 wrote: post #24
acqq wrote: post #20

Earlier quoted context omitted.

> What should I take away from this? Just that you should not omit with "..." the most important part of the statement for you. You haven't studied physics. Then you see the formula and the textual name of it with the most probable entry in Wikipedia and then you throw the most useful bit for you away and ask what the formula without the name means, even though the name would give you an explanation. pc86's quote: >…

I assumed there was some relation between k T log 2 joules and 20 Watts, probably giving some sort of context to how easily or not easily information in our brains could be lost. The theoretical principles behind the amount of energy required to erase one bit, while interesting, doesn't necessary matter for that relationship, does it?

Landauer's principle would come into effect if you had a theoretical "Computorium" material that could perfectly turn energy into computation. Then the computation you could produce per volume of space would be limited by it.

Our brains are (very) far from being the perfect computing material (so are our chips).

The theoretical minimum is only useful here to show that our brains cannot do free work. (Irreversible) computation always has a cost.

The 20 Watts are the important number. Our brains have a certain energy budget and have to perform within that constrain.

Re: How insects like bumblebees do so much with tiny brains

#40

Post by pavel_lishin »

Xophmeister wrote: post #28
pizza wrote: post #3

Earlier quoted context omitted.

Maybe something along the lines of "our brains would overheat" - there's a small temperature window in which proteins won't denature, it takes k T log 2 joules to erase one bit of information (Landauer's principle), and our brain uses around 20 Watts of power. Maybe tin foil hats make good heatsinks.. Interestingly enough there is some evidence that the Gibbs free energy of ketone metabolism is more thermodynamically…

Doesn't that imply that people who live in hotter climates have...how can I put this politely?...a disadvantage when it comes to the stability of their brains? I would assume that the brain has a certain amount of redundancy, but has this ever been researched: How hot can it get before people start getting noticeably "stupider"?

Our bodies can cool themselves decently well by sweating. I think that if the climate was affecting you to the point where your brain was getting cooked, the symptom would probably be less "slightly dumber" and more "seizures and organ failure".

But I'm not a doctor or a scientist, so I could be completely wrong. Hopefully someone more informed will weigh in.

Post reply on HN