What makes us feel good about our work?

Dan Ariely asks what makes us feel good about our work. Here’s a hint as to what it’s not: nasty reviewers (grumble, grumble).

Something light to think about as we head into Thanksgiving!

Learning socially but not socially learning

How do we distinguish learning from our friends from learning because our friends happen to be around? When I was younger, Goldeneye on the Nintendo 64 was the game to play, but I was sadly N64-less. Did I learn how to play Goldeneye because my friends were good at it and showed me, or because whenever I was around them, Goldeneye was available for me to play? But here’s a fact: I suck at Goldeneye. If I learned anything from my friends vis-a-vis Goldeneye, it was how to be humble in the face of continual defeat.

When animals are foraging for food, they face a similar problem. If they forage on their own, they don’t lose any of their reward (like their self-respect) to other animals. But by foraging socially they are able to increase the likelihood that they will find some food.

One of the biggest problems that social foraging can solve is that of risk-aversion: the preference for guaranteed rewards over risky ones, even when risky ones will be more rewarding over the long-run. In many cases, this preference is a simple reflection of the learning that all animals undergo. Risky rewards have both very large and very small rewards. When given the choice between multiple options, a string of bad luck on the risky option will lead an animal that learns to give up that choice and stick with the less risky option.

Yet learning dynamics are slightly different when an animal is surrounded by other animals. Animals are not identical clones of each other (usually) but have a variety of personal preferences for risk and reward. If you plop sparrows in front of both more and less risky options, of course they’ll generally prefer the risky options. But sparrows come in groups! And when in groups, they can be scroungers, hanging back and waiting to see what others are doing. And this lets them take advantage both of the range of group preferences as well as the range of learning in the group.

Interestingly, individuals only learn about the desirability of an option when they were the ones to have chosen the option but not when they watched (and joined) another individual making a choice. I’m not sure if there is a lesson here on group learning? Perhaps it is better for the group to keep their knowledge uncorrelated so that it is combined their knowledge will be as diverse as possible? But either way, they are not socially learning, not learning how to do something by watching another animal. Rather, they are learning by being in a social group that allows them to take advantage of the learning of each individual in the group.

References

Ilan T., Katsnelson E., Motro U., Feldman M.W. & Lotem A. (2013). The role of beginner’s luck in learning to prefer risky patches by socially foraging house sparrows, Behavioral Ecology, 24 (6) 1398-1406. DOI:

Monday open question: What is the most important open question in your field?

Neuroscience is a field both new and broad. It has roots in psychology, cognition, molecular biology, psychophysics, and more. Although there are some (slightly self-serving) attempts at defining what the open questions are, the sheer diversity of the field lends itself to many possible questions. The answer for those interested in psychology, in economic behavior, in ecology, in vision, in molecular biology, will all be different.

I asked this question on twitter (#openquestions) and got a couple of good questions, but I demand more!

Regardless of field: what are the most important open questions in your field?

Feel free to respond in the comments, on twitter, or on your blog.

Unrelated to all that, 11/22 edition

On the blog

I’ve been trying to find links to neuroscience, and economics/biology/ecology, resources for people who want to hear something serious. Neuro.tv and the Stanford NeuroTalk podcast are both good, and I recently found this set of excellent talks from the NIH.

I discussed how ‘wise’ crowds are, and when that wisdom might fail. It is actually not straight-forward when you should listen to what other people have to say, despite what some in the Economics field thinks.

Whether or not neuroscience is ready for open, ultra-collaborative work is a big question. There are a couple of projects that would probably qualify for that title right now – those being Open Source Brain and OpenWorm – but the more I think about it the more I find that there is no good consensus of what are the big ‘open questions’ in neuroscience. I am beginning to wonder if a resource for collaboration over the internet might be a valuable product…

I also found a couple of cool links on how adaptable people are (and how difficult it is  to separate phenotype from genotype!) and how surprisingly fluorescent arthropods are.

Elsewhere

Because it is known. Prophecy Sciences wants to use neuroscience to improve hiring/sports.

Dear traveller: please don’t think ill of us. We are the last generation. And we are immortal.

Probably a bad idea. Let’s not give this kid a billion dollars.

Let’s do all of it. Doing the things that you’re not supposed to do with Google glass. Can I say that I for one don’t understand the glass hate?

If you stay there too long, you won’t be the same. Are the Andes the most rapidly evolving place on the planet?

I can guess which numbers are even with 95% confidence. Why brains are not computers (ed: except maybe they are, they’re just doing different inference).

Because you want people to understand your stuff. On why engineers and scientists should be worried about color. Very worried.

In the journals

The intrinsic dimensionality of plant traits and its relevance to community ecology. DOI: 10.1111/1365-2745.12187

Toward a neural basis for social behavior. DOI: 10.1016/j.neuron.2013.10.038

Symmetry in hot-to-cold and cold-to-hot valuation gaps. DOI: 10.1177/0956797613502362

Spatial memory and animal movement. DOI: 10.1111/ele.12165

Other

Your weekly image (the newly discovered ‘plant hopper’):

Your weekly tweet:

(Ignore this: 3YG65VYNY7MK )

How to see underwater

Okay, this is a new one to me. A group of Austronesian people known as the Moken – ‘sea gypsies’ – are able to control their pupil dilation in order to see clearly underwater. Unlike some human adaptations, the ability to control pupil contraction is something that anybody can learn.

[via kottke]

Is neuroscience mature enough for a #polybrain project to exist?

The field of mathematics has been doing fantastic things by bringing people together online to solve math problems. Like the wiki or open source approach, math has been doing polymath projects for years now. It all kicked off when Gowers posted the question, is massively collaborative mathematics possible? It’s a great post full of great ideas, and it turns out the answer is a resounding yes with a steady string of successes. One of the advantages of this talent sharing is that you collapse the world. Anyone with talent or interest could contribute. When I was an undergraduate living across the seas (math major, here), I would have loved to help with a neuroscience theory project, even a little bit.

The most recent polymath success – improving the bound between gaps of primes from 70,000,000 to 4,680 – has me wondering: is massively collaborative neuroscience possible? I think the answer is yes, if the community wants it to be.

The hard part for starting such a project is identifying a question. Seven years ago, van Hemmen and Sejnowski published a book on 23 Problems in Systems Neuroscience. Could these be a good set of questions to work on? The more theoretical or computational the better, as we should assume that there will be no new data forthcoming for such a project. Luckily, though, there is at least one data repository (I think there are more?) . In Izhikevich’s green book there is a list of ‘open problems’ for graduate students. But really: what are the open problems in neuroscience? What is amenable to group work? What are the open problems in theoretical neuroscience?

And would the neuroscience community support such an endeavor? Can we have a polymath project or a github?

Fluorescent beauty of arthropods

There’s a great post over at Charismatic Megafauna showcasing the beauty of fluorescent arthropods:

Scorpions have “cuticular fluorescence.” Basically, compounds in their exoskeleton absorb and re-emit ultraviolet light as visible light (light humans can see). The exoskeleton of an arthropod is made from composite materials that are both strong and flexible. It’s the outermost layer, epicuticle, that produces the glow, and it seems to be something that changes chemically as the animals grow…

Some interesting ideas have been proposed for why scorpions and other arthropods glow like this. While humans can’t see UV light, most insects can, and much of the world around uslooks quite different in UV light. Some experiments show that scorpions may use presence of UV light as a way to detect shelter. (They determined this by putting tiny little goggles on the scorpions that blocked their vision…)

Go read more!

When crowds aren’t so wise

Alex Tabarrok recently related a familiar story about the ‘wisdom of the crowds’:

I ask the audience to guess my weight. They all wrote their guesses on a piece of paper. All the guesses was collected and an average guess – the “consensus forecast” – was calculated, while I continued my presentation.

I started my presentation and I naturally started telling why all of my forecasts would be useless – or at least that they should not expect that I would be able to beat the market. I of course wanted to demonstrate exactly that with my little stunt. It was a matter of demonstrating the wisdom of the crowds – or a simple party-version of the Efficient Market Hypothesis.

…I usually think of my own weight as being just below 80 kg…As always I was completely confident that the “survey” result would come in close to the “right” number. So I was bit surprised when the  ”consensus forecast” for my weight came in at 84.6 kg…So once I came back home I immediately jumped on the scale – for once I hoped to show that the Efficient Market Hypothesis was wrong. But the verdict was even more cruel. 84 kg!

And so, Tabarrok concludes, the market does not lie! Or at least, does not spread deliberate falsehoods. Except that’s a bit of a non sequitor, because there are two independent issues here: are crowds wise? And do markets reflect this wisdom?

One reason that crowds might be wise is that they are noise canceling. You and I may be able to guess something like someone’s weight from visual information fairly well but we can never be perfect – even a machine-like ‘optimal decoder’ will give a range of possible values due to little visual and personal quirks. But if we are all noisy in the same way, guessing independently from each other then the noise should disappear. Think of it like a drawing from a gaussian probability distribution – the more guess you make, the closer you get to the correct number.

Crowds can also be wise because they can generate more possible ideas. Just like the old saying goes, everyone is a bit different and will offer slightly different perspectives. Though you don’t even need that difference! A group of foraging animals aren’t all going to be checking the same areas for food, so when one finds a great food source others can join in.

There are a few clear problems that arise from trusting a crowd. A vast literature in ecology is dedicated to the question of when you should use social information instead of just your own personal information. Somewhat unintuitively, when the quality of personal information is lower it is less useful to pool with the crowd! And the size of the crowd matters; individuals in large crowds have an easier time of maintaining high group information even when personal information is of low quality.

Unfortunately, this assumes that there’s no talking in the group; nothing is being coordinated. Because when you start coordinating, when one person starts having influence, then suddenly opinions are correlated and can get dragged around in unfortunate ways. Now things are less diverse and you’re in a functionally smaller group (funnily enough, it also makes people more confident in the quality of their personal information).

Of course, this assumes you can trust others. Why should they provide you with their information? In the real world, my ability to get a good deal or an animal’s ability to find food may mean that if you find out, then I can’t get that deal! And the animal can’t get that food! Generally, it is better to trust the information from those whose goals don’t totally overlap with your own. In the wild, animals often do better when they don’t trust non-related animals of their own species (conspecifics) and trust other species (heterospecifics). And they prefer information from animals that have small home ranges, leading to information parasitization from animals that have larger home ranges! (I think that’s a pretty cool concept.)

And that brings us back to markets. Are crowds wise? Sometimes; but more importantly, should we trust the wisdom of the crowds? In a properly designed market, sure. I can’t think of why a one-time, everybody wins market wouldn’t be great. But when there are is low personal information (aka noise traders), group information goes down. When markets are small – and they are often quite finite! – group information goes down. When I can temporarily mislead an opponent about a reward, group information goes down. Markets can always lie.

References

Seppänen JT, Forsman JT, Mönkkönen M, & Thomson RL (2007). Social information use is a process across time, space, and ecology, reaching heterospecifics. Ecology, 88 (7), 1622-33 PMID: 17645008

King AJ, & Cowlishaw G (2007). When to use social information: the advantage of large group size in individual decision making. Biology letters, 3 (2), 137-9 PMID: 17284400

Lorenz J, Rauhut H, Schweitzer F, & Helbing D (2011). How social influence can undermine the wisdom of crowd effect. Proceedings of the National Academy of Sciences of the United States of America, 108 (22), 9020-5 PMID: 21576485

This just in: reptiles are animals too

People are finally figuring out that animals are smart, right? That just because they can’t do algebra they can still figure out some other things okay? It turns out that reptiles are animals too:

“Reptiles don’t really have great press,” said Gordon M. Burghardt, a comparative psychologist at the University of Tennessee at Knoxville. “Certainly in the past, people didn’t really think too much of their intelligence. They were thought of as instinct machines.”…

For evidence of reptilian intelligence, one need look no further than the maze, a time-honored laboratory test. Anna Wilkinson, a comparative psychologist at the University of Lincoln in England, tested a female red-footed tortoise named Moses in the radial arm maze, which has eight spokes radiating out from a central platform. Moses’ task was to “solve” the maze as efficiently as possible: to snatch a piece of strawberry from the end of each arm without returning to one she had already visited…

By using experiments originally designed for mammals, researchers may have been setting reptiles up for failure. For instance, scientists commonly use “aversive stimuli,” such as loud sounds and bright lights, to shape rodent behavior. But reptiles respond to many of these stimuli by freezing, thereby not performing…Scientists may also have been asking reptiles to perform impossible tasks. Lizards do not use their legs to manipulate objects, Dr. Leal said, “so you cannot develop an experiment where you’re expecting them to unwrap a box, for example.”

Now that scientists have gotten better at designing experiments for reptiles, they are uncovering all kinds of surprising abilities. Some of the most intriguing work involves social learning. The conventional wisdom is that because reptiles are largely solitary, asocial creatures, they are incapable of learning through observation.

None of this should be particularly surprising. If reptiles weren’t able to spatially navigate through their environment, I’m not quite sure how they would survive. After all, even fruit flies have spatial memory! But I love lizards, and it’s always important to note that what is ‘intelligent’ for an animal to do is determined by its environment and particular characteristics, not by what some other animal is able to do.

Learning neuroscience without being talked down to

Whenever a nonscientist friend sends me something to read about neuroscience, I usually smile weakly and pretend to read it. Sure, these articles are interesting but they are often imprecise and not as packed with information as I would like. Neuroscience is a large and deep field, and while I know some fraction of it quite well, there is much more that I do not – and the best way to learn about it is to hear from the experts who are not trying to talk down to you.

I guess the field is maturing enough that we are now starting to have access to regularly updated shows in this format. One, in video format, is Neuro.tv with the most recent video posted above. It looks pretty great so far and has a kickstarter for funding to create more episodes. Seriously, go donate.

The other is an interview podcast series by the students at Stanford Neuroscience called NeuroTalk. Every week or so, a neuroscience seminar speaker visits Stanford to give a talk on their work; the students have decided to interview the speakers, have them give a concise description of what they do, and get them to chat for a bit. It’s a fantastic idea and something I wish my current university did…  Here is someone who does decision-making, for example.

Beyond this, you can often see recorded lectures just by going to youtube and typing in a scientist’s name… it is often worth it.