First Video of Grandroids

I got permission to post this on my blog:



This is a video of Gollum, the prototype Grandroid, showing off his ability to visually track moving objects. So far that's about all he can do, but eventually every single aspect of his cognition and behavior will be as thoroughly and carefully constructed as his visual tracking is now.

Unlike norns, Grandroids do not see in terms of 'classifiers'. Norns don't see the difference between a block of cheese and a honey pot - both are classified as food, and that's all the norns see them as. From what I understand, their vision is hard-coded, with the game directly stimulating the norns' brain with the right classifier when they look at an object.

Gollum is different. He will actually see 'red round things' or 'square orange things' and have to learn how to recognize them. He sees by projecting rays from his eye (only one of his eyes actually sees so far) and having them detect objects in their path. He also has a reflexive instinct to follow moving objects, which is what this video is showing off.

There's something very special about watching a computerized creature doing the exact same thing that a newborn does, for much the same reason. To know that he can actually see those balls (even if he doesn't yet have the cognitive capacity to recognize what they are or really think anything about them) is simply amazing.

Everyone else agree?

Visible Indication of Executive Dysfunction

Just a random observation.

I have five blogs in my favorite blog list. Of those five, three are maintained by autistic people, one by a psychopath, and one by a 'fat guy in a wheelchair' (as he describes himself). So, four out of the five blogs in my blog list, as well as my own blog, are maintained by people with executive dysfunction. Guess which ones are updated inconsistently?

It seems that executive dysfunction affects blogging as well.

Just Like His Father? Book Review

I just got the book Just Like His Father? by Liane Leedom. (She has a website here.) Unfortunately, it's not as good as I hoped it would be.

Firstly, she doesn't use the terminology right. She recognizes that there's a difference between antisocial personality and psychopath. But she uses psychopathy to refer to only the most severe kinds of psychopaths, and describes all antisocial people as having characteristics (lack of empathy, impaired moral reasoning) that only psychopaths have.

If it was just a matter of using the words differently, that wouldn't be a problem. But then she assumes that researchers are talking about psychopaths when they're really just talking about antisocial personality, and therefore misinterprets the significance of their results. She makes empathy and remorse sound a lot more environmentally determined than they really are, because she combines research on environmental risk factors for non-psychopathic antisocial personality with research on the empathy and remorse problems of psychopaths.

The two most common subtypes of individuals with conduct problems (ODD, CD and/or antisocial personality - often those three form a developmental progression) are psychopathy and behavioral dysregulation. Here's a summary of how the two differ:

• Emotional (affective) empathy for others, and capacity for remorse - A deficit in this area is a defining characteristic of psychopathy while behaviorally dysregulated people range from normal to increased affective empathy. The only people they lack empathy for are their victims, and even then, this isn't always true. When behaviorally dysregulated people lack remorse, it's because they've found an excuse that's sufficient to convince themselves that they did nothing wrong; psychopaths don't need to convince themselves because 'wrong' has no meaning to them.
• Self-esteem - Psychopaths are typically grandiose and narcissistic, meaning their self-esteem is too high. Behaviorally dysregulated people very often have low self-esteem, which can be both a cause and an effect of their conduct problems.
• Depression/anxiety - Fearlessness is one of the classic traits of psychopathy; in addition,  psychopaths rarely get depressed as a side effect of lack of guilt/shame and high self-esteem. In contrast, many behaviorally dysregulated people have mood disorders (depression or bipolar) or anxiety disorders, especially PTSD (the link between trauma and conduct problems is stronger in these individuals than in psychopaths).
• Aggression - Both psychopathy and behavioral dysregulation are associated with reactive aggression, ie aggression that is motivated by strong negative emotions and not intended to achieve a certain goal. However, only psychopaths show an increase in proactive aggression, meaning premeditated, goal-directed aggression. So, while both psychopathic and behaviorally dysregulated kids are likely react to being called names by beating up the name-caller, only psychopathic kids are likely to calmly walk up and pick a fight with a kid so they can steal a toy while the other kid is freaking out.
• Moral reasoning - Psychopaths can sometimes talk a good talk, but they find the concept of right and wrong very difficult to understand. Behaviorally dysregulated people have no such difficulty; they are largely normal in moral reasoning, but their mood and impulse control issues mean that their behavior doesn't match their standards; or else they chose a different moral code to follow (such as loyalty to a gang - psychopaths don't make good gang members).
• Cause - Psychopathy is strongly genetic, with a small environmental component. Behavioral dysregulation is strongly environmental, with a small genetic component (mainly in terms of ADHD and bipolar disorder predisposing to behavioral dysregulation).

So, clearly, there's a big distinction between the umbrella diagnoses that lump psychopathy and behavioral dysregulation together, and psychopathy itself. Psychopathy isn't just more severe antisocial personality - it's a distinct condition.

So that's my biggest criticism. But not my only one.

Next, regarding impulse control, she talks as if all impulses are bad ones. Impulses can be good, bad or neutral. You can, for example, have the impulse to hug a loved one, or to go play on a swing. Neither of those are bad things to do (as long as the context is appropriate). In addition, it's possible to have too little impulsivity - this is the personality type known as overcontrolled, who are at high risk for anxiety disorders, depression and obsessive compulsive personality disorder. It's important to strike a balance - give into your positive and neutral impulses and resist your negative ones. However, for a child with too little impulse control, a focus on building impulse control is a good thing. Willpower acts like a muscle, and practicing it will make it improve.

I was especially disappointed by her chapter on moral reasoning. She parrots Kohlberg's stages of moral reasoning, which were discovered in the 1950s, and entirely misses all the newer moral reasoning research that is relevant to her topic. In particular, she makes no mention of the moral/conventional distinction, a moral test in which psychopaths show especially notable differences. The vast majority of people draw a distinction between against the rules and morally wrong (causing harm to others) - ranging from 3 to adult (younger kids aren't testable on this test), across many cultures, and including most atypical populations, such as abused children, non-psychopaths with conduct problems, autistic kids, and so forth. Only psychopaths and people with certain kinds of brain injuries fail to draw a distinction between rules and morality.

This research is far more relevant than Kohlberg's stages. While people with conduct problems, and particularly psychopaths, do show a statistical trend towards lower Kohlberg stages, this difference is far less pronounced than the difference on the moral/conventional distinction. (They certainly aren't all fixed at stage 0 - even some psychopaths score at stage 3 or 4, although they're probably espousing beliefs they don't personally share. And a committed gang member would likely score at least at stage 3, because this stage involves ideals such as loyalty to your in-group.)

Another problem, which she only hints at indirectly, is her lack of understanding of the distinction between cognitive and affective empathy. (On page 64 she mentions that blind or deaf kids are often delayed in empathy development, and on page 66 she talks about difficulty reading emotions, with no indication that she recognizes that those two examples are discussing a totally different kind of empathy than the rest of her book is.)

Cognitive empathy is the ability to accurately identify other people's emotions and to imagine another person's perspective; affective empathy is experiencing an emotional response more appropriate to your perception of another person's situation than to your own situation. (I say 'your perception' because inaccurate perception of others doesn't mean a lack of affective empathy, even though it means their emotional response may not fit the other person's actual experience.)

Autistic people typically struggle with cognitive empathy. In addition, individuals with ADHD can often have issues with cognitive empathy as well. Blind or deaf kids are often mildly delayed in cognitive empathy (with the exception of Deaf of Deaf kids) but usually catch up, except in the case of deaf people exposed to language very late. However, all of those people (unless they happen to be psychopaths as well) have entirely normal affective empathy. When they behave unempathetically, it's because they don't know, not because they don't care.

Psychopaths, in contrast, have severe problems with affective empathy. Their emotions always or almost always arise solely from their own situation, regardless of how others around them are doing. If they see someone else is unhappy, they won't feel unhappy themselves, even though they can tell the other person is in distress.

Psychopaths aren't entirely normal in cognitive empathy. They have specific difficulty recognizing certain emotions (most often fear). In addition, pre-pubertal psychopaths have a mild delay in development of cognitive empathy, which resolves around their preteens or early teens (my guess is that's when they realize they could benefit from learning to manipulate people). But impairment in cognitive empathy is not one of primary characteristics of psychopathy.

Another issue is that she buys into the widespread myth that criminality is on the rise. It used to be on the rise, back in the 1960s and 70s. The reason was because the Baby Bommers were aged around 15-25 back then, and that's the age group that has the highest rate of criminal behavior. Since the Baby Boomers have outgrown the high-crime age, the rate of crime has gone down dramatically as a result. Furthermore, our preventative efforts have worked - the rate of crime among 15-25 year olds has gone down as well. However, crime reporting has gone up, and along with it, public fear of crime, which she only serves to encourage.

My last criticism is regarding references. She is actually unusually good at providing references for what she's saying - I've seen many books where there are no references given whatsoever. In contrast, she provides a total of 129 references, which I have every intention of tracking down and reading as many as possible. But she has a few claims provided without any references to support them, and they are given in the same tone of certainty as the referenced claims. For example, on page 115, she claims that giving a child with high sensation seeking the intense stimulation that they crave is a bad idea - they'll adapt to it and become even more sensation-seeking. She provides no references for this.

As far as I know, this question has not actually been studied. While it may seem common-sense to her, psychology research has many examples of common-sense ideas that turned out to be wrong (such as the idea that infants will cry more if you promptly pick them up whenever they cry). And you could easily make a counter-argument that if you don't give the kid intense but appropriate stimulation, they'll seek out their own intense stimulation through inappropriate means. I don't know which model is right, and she doesn't either. She should've made this clear in her book, instead of providing her unresearched claim in bold text like it's especially significant. (Another example is her claim of an inverse relationship between aggression and affection, which she presents complete with a illustration to cement the concept. This, too, is unreferenced.)

Last, I personally disagree with her take on social dominance and respecting authority. She argues that to handle a child's strong desire for social dominance, you should socialize them into respecting adult authority figures. However, while this may make the child more obediant while they're a child, it seems to me that this will only strengthen their social dominance orientation in the long term. After all, social dominance orientation is not simply a drive - it's also a worldview. Social dominance orientation has two components; a worldview in which someone always has to be on top, and a strong desire that that person be them. Her advice seems like it's designed to reduce the desire to be on top while strengthening the worldview that someone has to be on top. So the kid learns the lesson that, realistically, a child can't be the boss. But once he grows up, he'll soon realize that he can be the boss if he does the right thing.

Sadly, individuals high on social dominance orientation, though they dearly want to be leaders, are usually terrible leaders because they will toss out morals in favour of more power. And people high in authoritarianism, which also involves the worldview that 'someone must be on top' (they believe good comes from submission to authority), will typically blindly follow their leader even if he does trade morals for power. So even if this strategy works and the kid sticks to following authority, he'll be an easy tool for an unjust ruler. It's important not to overemphasize submission to authority - that's what led ordinary, decent Germans to support Hitler's persecution of the Jews and the others he labeled 'undesirable'.

Despite all these criticisms, it is a good book. She has many bits of excellent advice. Her emphasis on a strong parent-child bond is probably the most valuable message in her book - she's quite right when she says that for fearless children the parent-child bond is extremely important to their development of conscience. And her advice on how to build a parent-child bond (keep your child with you, don't ever tease them, be responsive to their needs, show empathy for their emotions even as you set limits on their behavior) is excellent advice.

Other parts I like are: the recognition that single parenthood, while suboptimal, is often not a choice; the advice to explain the reasons behind rules and tell the kid what he/she should do instead of the misbehavior; recognizing that tantrums can result from a legitimate gripe; and the recognition that impulsive kids who want to please should be handled gently.

Social Skills in Video Games

I recently came across this video:

I'm not sure how I feel about the overall point of the video. My impression is that he's probably describing a real phenomenon but overstating it's importance. But at 6:07 he starts trashing video games.
I take exception to his comment that video games are all about the present (what about grinding?) but that's not the worst bit. He says video games don't develop social skills. Since he tossed out Warcraft, I'll discuss the mopst popular game in that series - World of Warcraft - to challenge the idea that you can't build social skills through a video game.
World of Warcraft is an MMORPG - Massively Multiplayer Online Role-Playing Game. This means you play on an online servor along with thousands of other players. You'll see other players wandering around as you walk through the world. Each one of those is another real person, who could chose to interact with. I'd say these incidental interactions are about as common as incidental interactions between people walking down the street. So not common, but it happens.
But World of Warcraft also has more structured multiplayer content, and to succeed at this, you have to interact. I'll talk about all the structured multiplayer content that I have personal experience with, one after another.
Trade
Sometimes players have items that other players want. You can make a lot of gold by trading items to other players. There are two ways to do so - trade chat or auction house (though some people combine the two by advertizing auction house products on trade chat).
Trade chat is a special chat room you enter whenever you're in a capital city. If you want to buy or sell something you can just ask repeatedly on trade chat until you get a taker. People also use trade chat to ask for advice, and to vent when they're frustrated, and occasionally to engage in some of the most bizarre and entertaining conversations in the game. A lot of these conversations take the same skills as the verbal part of real-life conversations.
The auction house is sort of like an ingame version of EBay. You put an item up for auction for a set period of time (usually 24 hours). You post a minimum price for bid, and you can also chose to set a buyout price for people who want the item immediately.
Whether you're trading through trade chat or the auction house, you need to use perspective-taking to make money at it. You have to think about what they will or won't want, and how much they'll be willing to pay for it. The easiest way to answer the second question is to see what others are selling that item for, and price yours near the cheaper end of that range. (I usually set mine as slightly cheaper than the cheapest.)
The way to decide what will sell is to think of what you'd want to buy. Stuff that can only be used by the same people that make it (like arclight spanners) will not sell. Stuff that can be used by a different profession than the one that makes it (eg enchanting rods) will sell quite well. So will raw supplies for professions, because although most people can get the raw supplies they need themselves, it's a lot more convenient to buy it if you have the gold. There are some characters referred to as 'gold farmers' because they'll spend ages gathering supplies that they can sell to others.
Dungeons and Raids
Some zones are designed to be impossible to solo. These fall into two categories - dungeons involve five players, raids take larger numbers. Both dungeons and raids have a team of players fighting very difficult non-player character opponents. And both have designated roles for specific characters, consisting of tanks, healers and damages (a dungeon team is one tank, one healer and three damages, most raids have a similar ratio).
You need to work cooperatively to do dungeons or raids. Firstly, many people may be at different levels of experience with the game, and some bosses have special tricks to them. Very often people will stop before the start of a tricky fight and ask if everyone's played that fight before. If you haven't, they'll explain that you need to get into the shiny triangle when the boss calls for Al-Akir's aid, or two players need to get a stacking debuff on themselves but not let it get to 100, or some trick like that.
You also need to manage conflict. Many people, admittedly, are not very good at this, and many dungeon teams fall apart as a result. But when teams work well together, you can have a very enjoyable experience, and people will often do several dungeons in a row if they have a good team.
Besides the size of teams, dungeons and raids have another difference. Dungeons have a tool for letting the game find you teams, either queueing for a specific dungeon or for a random queue. Raids don't have this system, so you have to find your own team. (You can find your own team for a dungeon too, if you want, or find part of a team and then queue to fill in the missing roles.) There's another city chat room besides Trade chat - it's called Looking for Group. I'm not sure exactly what skills are involved in finding a raid group, but there certainly are skills involved, because I find it really difficult to attract anyone to join my raids. One thing it does take is patience (who said video games are instant gratification?). You're going to have to stand around a long time, repeating your request for a raid group over and over, until you finally gather enough people. And if you decide to try it with the minimum number, you may end up with a frustrating series of whole-team wipes (deaths) because you misjudged your teammates' power.
Battlegrounds
World of Warcraft has two factions, Alliance and Horde. Most of the multiplayer content is for one faction only (there are two trade chats, single-faction dungeon teams, etc), but battlegrounds involve both factions forming raid-group-sized teams to fight each other. There's a capture-the-flag game, a capture-the-bases games, a defend-and-switch game, and a bunch with combined elements. Each require some degree of cooperation to succeed. Like dungeons, you can queue to play with random players, although there are also 'rated battlegrounds' (which I haven't played) which involve guild teams against each other.
I'll focus on the first battleground you can play, a capture-the-flag scenario called Warsong Gulch. In this game, if you aren't strategically inclined, you're likely to end up in the middle of the field, bashing on the other newbies. The more strategically inclined people either hang back to defend their own base, or run in to steal the flag from the enemy's base. When someone has the flag, they essentially have a huge target painted on them, so other people must defend the flag carrier. In addition, you can only capture the flag if your own flag is at your base, so if both teams are running around with the enemy's flag, you need to kill the enemy flag carrier and return the flag (by clicking it) before you can capture their flag.
Strategic behavior also requires knowledge of what other players will do. While several players hanging back to defend the base is good strategy, one or two hanging back are liable to just get killed when a huge horde comes to take the flag. In lower-level Warsong Gulch, virtually no one hangs back at the very start, so you're better off just rushing forward with everyone else. In addition, while a lone character grabbing the flag can make a pretty good distraction, they have very little chance of capturing unless a group comes to defend them.
Another skill in battleground involves the battleground chat, which can be used to communicate to your entire team. If you just use it to scream about how newbie your teammates are, you won't accomplish much. But if you give simple, clear suggestions for how to play more strategically (eg 'kill the enemy flag carrier' or 'everyone rush the green gate') you can often get many of the less skilled players to follow your commands, thereby improving the entire team. (I've been one of those less skilled players myself, so I know what kinds of commands are easier to follow for newbies.)
The battleground chat can also be used to report on the enemy's movements. 'The enemy flag carrier is on their roof' sends everyone who's paying attention rushing to the roof of the enemy's base. In another battleground, 'farm lost' tells everyone that a huge enemy army has taken the farm and we can't get it back, so it's better to go to other bases instead. Whereas 'farm needs help' means we still have a chance to fight them off, but only if more people come to defend it.
And even the simple battles involve social skills. The enemy players are people, and they require a different style of attack than a computer does. For example, if you use a skill that makes you invisible as long as you stand still (night elf racial skill), players know to check for you where they last saw you. They also know to go for the weaker, high-damage characters or the ones who are healing teammates, so the tank role, so effective in dungeons, is obsolete. Stealthy characters can often be found by area of effect attacks, unless the stealthy character successfully predicts where the AoE will be targeted and avoids it. Certain classes are weaker to certain strategies, such as silencing a paladin so he can't heal himself, or ambushing a high-damage ranged character. One of the most key skills to learn is to keep moving - stationary characters are easier to kill, especially stationary ranged characters.
Guilds
Most of the multiplayer content I've described so far is short term, primarily involving interactions between strangers. Not so with guilds. Guilds are long-term cooperative groups of players, who get their own shared bank, their own chatline, notification when a guild member comes online, and a calender to arrange guild events.
Each guild has it's own culture. Some guilds interact very little, others quite a lot. Guild chat can rival trade chat for bizarre playful conversations. You can ask guild members to help you with difficult solo quests, or give you advice or guidance. Guild members may give other guild members useful items, such as profession supplies or larger inventory bags. Guild members of the same level as you can also be recruited for dungeons, raids or battlegrounds, and some guilds have an established raiding team so they don't need to search for people to raid with. Guilds each have their own rules about things like who gets to remove stuff from the guild bank and whether it's OK to ask guild members for gold.
With a guild, you have long-term contact with the same group of players. If you don't like each other, one or both of you may end up leaving the guild, along with possible friends you've made. You may even be forcibly removed by the guild leader. As a result, it pays to be friendly to your guildies. If you get along, you can easily end up becoming friends, or in rare cases even lovers. (There are married couples who met through World of Warcraft.) In some cases, guilds can become extremely important to people. They are a community, just as much as any real-life community such as a PTA or a sewing circle. The one difference is that the demographic is much broader, including peoples of all ages and from different countries.
So anyone who thinks video games don't build social skills has not played World of Warcraft.

Childhood Psychopathy Screening Test Online

I couldn't find any online test for psychopathy/callous unemotional traits in children, so I decided to make one:

http://www.okcupid.com/tests/the-combined-aspdsdq-test

It's taken from this study. It's meant to be answered by parents, about children between the ages of 4 to 9 years old.

Two Types of Cognitive Empathy

"I want to point out that the definition of cognitive empathy being used in Baron-Cohen and Wheelwright’s paper is quite different from the one that I have been using for some time. In my understanding, cognitive empathy has to do with being able to read nonverbal cues (body language, facial expressions, the expressions in the eyes, and so on) in order to intuitively “tune in” to what another person is thinking or feeling. I have not been using it simply to cover being able to see things from another person’s perspective or to understand the other person’s mental state."
Rachel Cohen-Rottenburg - A Critique of the Empathy Quotient (EQ) Test

This quote made me think of something I've sort of been implicitly aware of, but not explicitly verbalized. The ability to read nonverbal cues, and the ability to mentally represent another person's perspective are actually conceptually distinct abilities, even though they are both common difficulties among autistic people.

Rachel goes on to say that she has difficulty reading nonverbal cues, but doesn't see herself as having difficulty imagining another person's perspective. I'm the same way. In fact, as a creative writer, my guess is that I'm, if anything, better than average at imagining other perspectives - much of my fiction requires that ability in spades. The psychology degree I'm working on also requires that skill. At the same time, I have mild, but significant, difficulty reading nonverbal cues.

Which raises the question - how many people have trouble with one skill but not the other? It's a well-known finding that most older higher functioning autistics pass first-order, and usually second-order, theory of mind tests.The usual interpretation is that autistics have a developmental delay in theory of mind, and if this delay is mild, they will eventually get it. But what if a subset of autistics have no delay in theory of mind at all? Maybe in some cases the social difficulties in autism result solely from nonverbal communication problems, with no higher-order cognitive deficit involved.

Another group who often show delayed theory of mind are deaf people. Typically, Deaf of Deaf have no delays in this skill. Deaf people raised by hearing people but exposed to signing at a young age show a delay of a year or two, oral deaf show more serious delays, and deaf who were first exposed to any language at a late age show severe delays or may never learn theory of mind. Some people have criticized this research by pointing out that standard theory of mind tests are linguistically complex and may be failed simply based on language problems, and indeed deaf people tend to do better on less verbal forms of the tests, but they still show this pattern. In contrast, most deaf people, regardless of language exposure, are acutely aware and responsive to nonverbal cues, supporting a dissociation between the two skills.

But yet, they're still correlated. The majority of autistic people show some degree of difficulty in both theory of mind and understanding nonverbal cues. So why are they linked?

The research on deaf people suggests an explanation: communication problems, in the absence of any cognitive impairment, can be enough to produce delays in theory of mind. How best to learn how other minds work than to hear it from the source? Information relevant for theory of mind can be readily accessed by verbal cues (eg Anne saying 'I think my ball is in the basket' when the child saw Sally move it to the box). It can also be accessed by nonverbal cues, such as Anne looking surprised when she checks the basket and finds nothing there.

Maybe in some cases, a severe difficulty reading nonverbal cues can, like a linguistically deprived environment, result in theory of mind issues simply because the child finds it difficult to gather enough information about how other people see the world.

Furthermore, if these are distinct skills, then what would be some higher-level theory of mind tests, more suitable for adults with mild difficulties? Maybe ask the person to try to get into the head of a person with a very different perspective to theirs - for example, can they imagine what it would be like to live in the 1600s? Or to portray the viewpoint of someone who disagrees with them on a strongly-held belief, like asking me to put myself in the head of a pro-choice person, or a homophobic person. Research should be done into how these abilities relate to autism, to verbal and nonverbal communication skills, and to the standard theory of mind tests.