Hacker News new | past | comments | ask | show | jobs | submit login
What Neurons Look Like as Drawn by Students, Grad Students, and Professors (theatlantic.com)
86 points by rosser on June 1, 2013 | hide | past | favorite | 32 comments



I think the creativity argument is bull. I would bet that the reason why there's a difference between the groups of diagrams is the depth of knowledge in neurons. Undergrads, who I believe we can safely assume have less in depth knowledge of neurons, memorize the physical shape of neurons because their cursory knowledge of neurons. On the other hand, professors have a much more depth (and specialized) knowledge of neurons and therefore may not even know what the neuron as a whole looks like (you could probably get a few PhDs just by studying the signalling mechanism across the neuron membrane). I'm guessing this because for example Figure 22 looks suspiciously like a circuit diagram.

For example, one wouldn't expect an EE/Physicist (the professors) who works on designing MOSFETs (neuron transmitters) to be able to draw a wiring diagram (neurons) that the MOSFETs would be used in, even though that EE/Physicist might be called a chip designer.

Of course, a simpler answer is that professors are busier than undergrads so they just put less time and effort into drawing these diagrams. So Occam's Razor?


I didn't see it as creativity either. Professors who have probably had to draw a lot of neurons over the years have naturally developed a symbolic representation of the object, which is common in many fields.


Of course, a simpler answer is that professors are busier than undergrads so they just put less time and effort into drawing these diagrams. So Occam's Razor?

Yeah, an undergrad, even though this was not a formal exam, would probably feel the need to prove themselves by faithfully reproducing their concept of a neuron. Most professors probably just wouldn't give a damn.


I was immediately struck by the similarity to how Picasso drew a cow (see: http://3.bp.blogspot.com/_yAdxTp2HDQM/S_P7lZz46-I/AAAAAAAAB1... ). It seems the professors have come to the point where they see the essence of "neuronness".


Also reminds me of this Pratchett quote (about "a horse" carved in The Chalk):

> Tain't what a horse looks like. It's what a horse be.


That's a bull not a cow.


A professor of mine was fond of pointing out that the textbook images of neurons were all of rat neurons. Many human cortical neurons have a structure called an apical dendrite which isn't depicted in undergraduate neuroscience textbooks. He was really obsessed with those apical dendrites.


It would be interesting to ask people to draw a resistor. Ignorant savages would draw a sketch of the physical shape minus all useful details (maybe with random color bands in no particular pattern except maybe because they look pleasing; maybe a new palette too); artists would draw something fairly realistic, and engineers would draw a schematic symbol.


I would also argue that the top row is drawn from a biology viewpoint, while the bottom row is drawn from a neuroscience viewpoint. Perhaps this could also explain some of the differences? For example, if you were interested only in neuroscience, why would you bother to draw the fatty layer around the neuron? It isn't useful to you.


It also depends what kind of neuron you're talking about. Plenty of neurons have unmyelinated axons. In particular, most of the neurons in the grey matter on the surface of the neocortex don't have myelin sheaths. It's the fat in the myelin that gives white matter its lighter colour.


I begin to understand what a standard physics text must look like to an outsider.

Utter gibberish at first view, then there are some words you can make out if you look closer and after lots of googling you can remotely comprehend the approximate meaning.


http://onlinelibrary.wiley.com/doi/10.1002/sce.21055/full Link to the research article. The link in the article is dead.


It's not dead, it's behind a paywall, just as on the site you linked. Can anyone post the actual pdf?



Interesting. Who runs libgen.org? How are they in compliance with copyright laws?


Author doesn't mention obvious observation, that professional scientists' drawings looked more like circuit diagrams for the most part whereas the other two groups' looked like portraits of biological neurons of varying accuracy.


I noticed that as well. It struck me that the third group may have been focussing on the function more than the physiology.


Or even more simply: busy high-status people draw sketchier sketches (except for a single outlier). (Maybe the full dataset contradicts this.)


As someone with a strong background in Biology who took several AI classes at an Ivy League school, I found many of my CS professors models oversimplified neurons to absurd levels.

Teaching CS students to use perceptron that can't even change its own behavior is like teaching MechEs to built rockets out of Legos. Sure you can do some cool things with them, but the sooner you move on, the sooner you can make some real progress.

After 2 AI winters, I'm happy to see that many schools like Berkeley are finally starting to combine their neuroscience and AI labs.


Neurons in AI are not supposed to exactly mimic the behaviour of 'real' neurons...


jergosh is right too, but I think the real reason is that academic computer scientists want models they can mathematically reason about. Real neurons are fine and dandy, and can obviously do fantastic things like drive muscles in certain ways to post semantically meaningful content to Hacker News, but we still don't have a general model that can explain how they work in concert at scale. Perceptron's popularity arises not from their accuracy... I'm pretty sure everybody knew they were inaccurate models, though the hope was (falsely) that the inaccuracies wouldn't matter. Their popularity derives from the discovery of the backpropagation algorithm. In fact, if one were to take a mathematical view of the situation, arguably backpropagation is what was really discovered, and perceptrons are incidentally the moving parts in the backpropagation algorithm.

It is not generally useful in AI to have some horrifically complicated algorithm that is based on something vaguely realistic, but that nobody know how to make do anything, or update based on new input, or figure out what's wrong when it doesn't work.


I have no issue with using perceptrons as a learning tool. But they have clung to this model for far too long. The main problem I see is the biogical learning mechanism comes from some feedback mechanism that causes the neurons to grown new connections.

Perceptrons have no way to mimic this and instead resort to sending an error from the output units back toward the input units duringing the learning phase. But when we go to use them, the information only flows forward. So you can never produce a dynamic learning system with these Legos. No matter how much super-glue you use, this rockets never going to make it into space. Yet many of these new deep learning neural nets continue to use back propagation for their learning phase.

I never said we needed EXACT models. And, I know perceptrons are simple to model mathematically and have 40 years of research and tools backing them. But after 40 years, isn't it time we try just a little harder to make something that is a little closer to how biological neurons really work?


"But after 40 years, isn't it time we try just a little harder to make something that is a little closer to how biological neurons really work?"

We have been trying. It has proved to be very difficult to get them to do anything useful, though, which was my point. Dedicated AI researchers are actually intensely aware of the differences between real neurons and their neurons, but getting fake neurons to do anything useful has not been as easy as just "simulate what the neurons do".

There was an interesting breakthrough 6 months to a year ago, some new algorithm that is apparently able to update a more interesting neural net, but I don't know much more than that at the moment.


I agree. From what I've read many of the younger AI researchers are interested. As I mentioned earlier, some school have recently created special labs and programs for them to work together.

My point was just 10 years ago, none of my CS professors showed much interest in neurobiology. Fortunately I think this attitude is changing.


Honestly, I'm not surprised at all. Students need to be thorough to show they understand the material, so their drawings are detailed. Professionals just need to put an idea on paper quickly and easily, so they only draw the details that are relevant.


Working out what are the 'details that are relevant' is perhaps a measure of knowledge in the field?


What a worthless study and article.

It's like if you asked a group of people from the 1930s to "draw an alien", and you compared that to modern people asked to draw an alien.

People from the present would draw the "gray" aliens you see on TV. Are they less creative than the people from the 1930s? No - they're just relating what they think is the accepted or most likely "correct" drawing of an alien.

Then theatlantic.com comes along and writes an article asking "do modern people really understand SCIENCE? Why are their alien drawings all the SAME? Are they missing a significant IMAGINATIVE STEP?"

No they are not - you idiots. Next time ask them to draw an "original interpretation" of some concept if you want them to draw something creative... otherwise people will just draw what they think is the commonly accepted answer among their peers.


Yet the undergrads did NOT draw a modern concept of a neuron. They drew the 1890s conception. So some temporal inversion thingy going on there?


if you were good at something and had to be creative around a whiteboard i'm sure you would use similar shorthand representations. just look at the slang coders on this site use when describing programming ie, 'to hook' onto an API connection for example. then think about sketching code while talking. seriously, this is your best comment today?


Isn't this coming to the wrong conclusion? Students probably just didn't know the extent of the dendrites and their structure, the fact that the axon is usually longer and thinner etc so they drew what they see in Books/magazines. What does that have to do with creativity? If anything, the scientists' drawings are the unimaginative ones, because they are realistic


The higher variety of shapes is what showed increased creativity. It's not anything about the shape in particular.


Again, how is this attributed to "creativity" and not to "scientists know that the structures of dendrites and axons neurons vary widely but still they are quite narrow so that's how they draw them"? Because that's how it seems to me.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: