Rewards as demotivators

What if all that we knew about what makes us do things would be partially wrong? What if giving people rewards for what they are doing would actually make them do that thing less, instead of doing it more?
In psychology this is called the overjustification effect.
Here is how it works: - extrinsic motivation and intrinsic motivation are not the same thing. Doing something to get something is not the same thing as doing something because you like doing it. Therefore being too goal-focused (and having the goal as something disconnected from the reality of the activity) might make us enjoy the reality of the activity less, shift our focus from the enjoyment, or make us feel anxious while doing it because we are trying to perform better in order to obtain the reward.

Yet why is it that it seems so normal to us that rewards should make people perform better?
How does one start to understand the genealogy of a cognitive mistake?
This theory might seem intuitive through an association to other theories that work, and a wrong inference that, by analogy, the shape of this problem of motivation should be similar.
The two theories that it might thus be associated to which spring to mind are:
- the theory of the survival drive to satisfy one's own needs - the wrong inference goes like this - > if one has an inherent drive to satisfy one's needs, then promising to satisfy them as a reward if that agent does a certain activity should project the drive to satisfy those needs on the activity the agent has to do in exchange.
        mistake 1 - who says the drive is really projected on the activity at hand? (see second associated theory)
        mistake 2 - given more rewards that could potentially satisfy more needs doesn't mean that a satisfaction point isn't reached, when the agent loses drive for having those needs satisfied
        mistake 3 - drive for obtaining something is not the same thing as the drive needed to sustain an activity in a performant, productive fashion
- pavlovian conditioning - if an object of desire is associated with another object for long enough, the symptoms of desire could appear in the presence of the other object
        mistake 1 - it might work with objects and contexts up to a point, but who is to say that one can transfer desire for an object to desire for performing an activity (by offering that object as a reward for performing the activity)
        mistake 2 - as human psychology is quite complex, it is very hard to say if in the same circumstance, the human won't simply feel inadequate for being unable to get the object himself, offended by having to trade an activity for it, and angry that the other person possessing the reward is making it so hard for him to reach the object of desire - all this resentment is bound to make the human feeling not quite so positive about the activity to do.


It seems we have cognitive blindspots, which we refuse to analyse. We all have the experience of not feeling motivated to do something despite the reward we have been promised, or starting from something that we normally love doing, then spoiling it to regimenting it too much and focusing on rules of doing it, the time it takes us to do it or how to move achieve it much faster, although, thus affecting negatively instead of focusing on the pleasure we get while doing it.
Yet this is cognitively unintuitive - it is trendy to think that if only we could give ourselves more rewards we would probably be more productive, or if only we would raise the stake more, put ourselves in competition, put a more anxious deadline or deal with even more discipline with our own work.
So despite having the experiences that would permit us to learn what motivates us, we stick to the norm of what should be motivating, and fail to formulate our own theories of what is really motivating based on our own experience.
That is why when people like Daniel Pink write a book like Drive, which discusses experimental evidence on what really motivates people, they reach a chord which resonates within all of us. They overthrow trends of thought, and verbalise the cognitive reality shadowed by our blindspots. They give our brain the ability to cognitively process and understand something which was previously hidden, yet completely belongs to us, to the way we function, to our experience.

the advantages of fuzzy memory

If cognitive enhancements would be out there on the market, if you could go and buy more memory hard drives for your brain, many of us, that appreciate living in our own mind quite a bit, would buy them.

Some people would love more processing power, some people would love to remember things more accurately and simply be able to store more details of events over longer periods of times. Some people would just love better working memory.

I for one always imagined I would be one of those people that could finally get one of those clearly indexed memories that other people seem to possess - that I will finally remember the titles and authors of the books I read, of the movies I see, of the songs I hear, and be able to quote previous research without having to look it up.

Tape-recorder memory, who says its bad? I say bring it on and let me try it!

Except there might be a big price to pay, bigger than having your nose bleed like Johnny Mnemonic after saying Hit me! (gosh I would love to say that just before they load some AI and Cog Sci Encyclopedia in my brain).

This price might be something quite essential to our human version of cognition, version which we still seem to appreciate more than that of computers, at least at the moment. It might be creativity, and out-of-the-box problem-solving.

It might very well be that what makes us tick and go outside our own knowledge is this very fuzzy version of memory that we have (and blame so much most of the time).

Human memory is many times a process of reconstruction. Yet we might proceed to this reconstruction without actually having all the pieces. We might start from some reference points which we have memorized, then apply our commonsense knowledge to fill in the details. In this process, memories become altered (and yup, it has some disadvantages concerning eye-witness testimony).

But other than that it is great! It allows us to mix knowledge from different times. Instead of sticking to the fact and reciting them back and forth, it allows us to build new types of knowledge and memories, new relations. Some of our great mental leaps or inferences might come from having to fill in some missing detail in a fuzzy memory, running a search across present knowledge, finding a match and then adding this match to the previous knowledge, thus constructing a new piece of knowledge through this process of memory reconstruction.

Without fuzzy memories, this would not be possible. Previous details would not fade away, with the possibility of being replaced by other more accurate or new knowledge, in a way which yields new relationships, inferences and moments of insight.

So think again before getting that memory chip, your inability to remember might make you the smart unique constantly self-upgrading intelligence that you are!

Behavioral space versus measured space

They are people that can't measure space. That can't look into the distance at the next signpost and say - yeah man, about 300 m until that one; or about 90 feet to the left of that turn you will find the post office.
Conversely, it is hard for these people to interpret instructions given in this language.
Measuring space in various types of units and having the ability to interpret these measurements by translating them back to some form of representation or ability to use them as cues for the environment is an useful skill.
But there are other skills that relate to space which one can be good at, despite doing poorly with measurements. Behavioral space is one of them.
We are talking in behavioral space terms whenever we say or have the feeling of "my phone is right here" (to my right/left; near my hand, within reach, etc) , or "the coffee cup is over there" (subtext - and I only need to make a step and lean forward to take it).
These types of representations don't talk about measurements, or at least not objective ones (e.g. in metric units) - or perhaps I don't have any friends that normally say "my coffee cup is one meter to the left on my desk).  This way of perceiving space relates to our own body size, ability to move and reach for things (and our experience with that), the time we feel we would need to do so, and our feeling of access and affordance of objects around us based on the things above.
Of course, a subconscious measurement might happen - our brain might measure arm span, arm reach, speed of reach or walk, project movement on the space between us and the object we are thinking about.
But whatever the units we are using (e.g. step size/speed and arm reach) they are not meters or feet.
One can be extremely good at predicting movement in one's own behavioral space, and therefore be able to measure it, but very bad when it comes to projecting slightly more artificial, "objective" units (like meters) on space (be it concrete space we are looking at or an imaginary space we have experienced).

You might be able to catch something falling down a shelf near you, to know exactly how much to lean to get a folder from a drawer - it is rare that we lean to get something and then feel we need to lean a bit more (trying to get something from a very high shelf does not count).
Our brain seems to compute all these distances on the fly and gives us the experience of a flow-like motion. If you think that is a little thing, think again!
Arm span has to be projected on distance to be covered, so that we walk the distance before actually reaching if our arm span is not enough, or move our rolling chair, or lean. We have to compensate for that extra distance somehow, otherwise we would reach out to grab things and only then make a few steps towards them. Instead we move towards these objects and then reach.
We also need to know when to stop, when we are close enough to reach, what is the distance of our arm span, and what is the best orientation of our body in order to have the right orientation for our muscles to coordinate the grab.
Most of the time, we step towards the shelf we need to grab an object from, and our arm starts moving upwards during the last phases of getting close to the shelf. This isn't, in any way, a simple feat. And to give that feeling of flow, that we normally associate with the "simplicity" if out movement, our brain has to learn a lot about our own body, movement, the environment, distance/size, etc.
If you want to understand how great of a feature this sensation of simplicity and flow of movement is, try starting to develop a movement skill you are not used to - or remember how it felt when you learned how to ride a bicycle, how to drive, how to swim or play a musical instrument. As a pianist, I know how developing new types of movement in different speeds feels anything like simple (I swear sometimes it feels like you are being asked to move your ears, or some muscles you have very little control over), and when you finally get simplicity and flow in your movements, your brain has managed to integrate all those movements, understand the distances and speeds involved, and is able to now coordinate them to perfection.
If you learn a new skill which involves a different type of movement (playing darts? pool? badminton?), or different distances, you can let yourself observe how your brain is feeling it's way into motion flow. During such an experience (which can involve a certain degree of frustration), one learns to appreciate more the marvels of our behavioral space and our movement repertoire.

The thought machine

Computers allow the information-processing paradigm, which has profound implications on both cognitive science and artificial intelligence. Information processing seems to have played for cognition the role that the microscope played for biology.
For a long time I thought that the stumbling block which makes people consider intelligent machines as not truly cognitive was not consciousness, but intentionality. Machines don't have their own goals, they can't choose for themselves to take another path than the one the programmer has endowed them with, neither can they come up with problem-solving skills and methods that haven't been put there by a human.
Processing information for a goal seems to be the defining quality that characterizes living systems, which can brings about cognitive behavior and processes which we don't see in machines.
Yet something is missing from this picture. Sometimes we forget that actually organs keep functioning automatically, just by being alive. That them being alive means they keep attempting to fulfill their function. So the function of our brain might be to process information, be it with a goal or not. It might be in its programming to do so, independent of if the owner has that as a goal or not.
This is where curiosity, as a human/animal trait comes into play. Many humans go out of their way to obtain information which couldn't possibly affect their survival. Yet the drive to acquire more information, to compare it to the previous one, to patch the types of information together, can feel as important for these people as eating (or more, at least until they get really hungry).
So it might be that the information-processing machine is simply in overdrive for some people. That it takes over. And of course they are survival advantages in that (if you are in the right job to manifest that talent, and if you can make a difference to the community through what you are processing).
Perhaps we should also ask ourselves why it is that our information-processing machine seems to always be on. Do we really need to do that much information processing? Is that an inherent part of the human condition, that we would have needed to develop this constant stream of thought/awareness?
Of course we feel that our thought machine sleeps at times - we zone out and can't think of the things we would like to think of, can't solve the things that we would like to solve.
Sometimes our minds seem blank and sometimes our ability to focus is slippery. Our machine can't seem to come to terms with what we want to direct it towards, and escapes our goal, while swimming heavily in its own purposes and plays at information-processing. It keeps on asking us questions we don't want or have the time to answer then and there, instead of helping us solve what we put in our schedule as a priority. It keeps focusing on things that seem unrelated to our task, albeit interesting for the context of our lives or the way we understand the world and universe and things.
That is one of the reasons for which some people use diaries. It is not because they need to constantly say what they did that day. But because they need to give their information processing machine a place to express its concerns, and process all those things that are in the back of our head, and that otherwise would stand in the way of our normal productivity. And they are questions in our head that we can't escape, that dominate our mental landscape in such a powerful way, that the old advice applies "the only way to escape a temptation/desire is to fulfill it".
In this context, the only way to get over that preoccupation that our mind has for certain subjects is to give it free rein to think about them at will and in a focused manner, so that we can then direct it at will ourselves after it has settled into having the answers that it needed, or being closer to them.
The mind-body system might not be a dualistic machine, but our own mind can be one at times, and we can remark and reflect on the difference between the things that we want to think about and the things that seem to "think us".
Our information-processing machine is not a completely tamed one, and perhaps it shouldn't be. As it knows better what our deep goals, interests and questions are than our daily task planner ever could.

Back to the intentional/non-intentional debate, I think there that our information-processing device just plays at times. It is a ludic device that takes pleasure (and rewards are significant for motivating behavior) from being active and trying to think and "solve" the world around us.
This can be easily seen in perception illusions, where we can't help ourselves but to solve certain patterns to some known common pattern. Or we can experience it when we look with no purpose at some patterns on the wall while thinking of something else and realize after a few minutes that our mind has kept on trying to arrange those patterns in various configurations.
The major implication of that thought is that it means we are more machine-like and less intentional than we thought. And that many times it is the machine that drives us forward, not the intention.
So if information processing is a function, and our brain can't help but do it, where does that leave the debate on the importance of intentionality as being a main feature that differentiates human cognition (our benchmark) from machine cognition?
First of all having this ludic information processing machine is not an easy process to replicate in itself. I don't know of any artificial cognitive machine that just keeps on adding random observations and drawing random conclusions and inferences out of them. Are they rules or general guiding principles which apply to our mental play? To be found out.
Second, our intentionality - our goals - might predispose us to gather information about how we can solve those even during our ludic time. We might be collecting information on the weird way the patterns are distributed on the wall and associate it through some far analogy to one of the important problems we are struggling with (and which is goal related) - in fact my research focuses on this process.
Second, there is a definite degree of independence, free will and randomness (read uniqueness if you like) that comes with this ludic information-processing machine, and probably defines the human cognitive experience. To be constantly only goal-directed is, after all, what we imagine machines to be. (very funny that we think that, but we also think that us having goals and free will is what makes us human).
Third, our experience in the world and in our thoughts is sometimes what makes new goals emerge. Which is not an experience replicated by a machine yet. And I don't mean refining goals purposefully - like in the case of creating subgoals in ACT-R. I mean simply thinking about some random things then suddenly deciding some possibility is attractive enough to be worth investigating, or experiencing things and then deciding to make something our goal just because we like it (or without particular knowledge of what cognitive goal that investigation might serve).
Therefore I don't see the idea of our brain being an information-processing machine that does just that all day long cause it likes to and that is its function as something that makes intentionality less important. Perhaps it is just bridging a little bit more the gap of our minds being mechanisms that have their own function, and drive us with the expression of this function, rather then other way around (us setting goals, and constantly directing our machine forward).
How would a neural implementation or model of either intentionality or ludic information processing look like? That is what I would really like to see.