30 September 2009

Cognitive Democracy

It is common, and tedious, to hear people discussing the poor quality of their political representatives. They often think those that represent them are mentally deficient or plain mendacious. They may be right on both counts but how would we know before it was too late?

What attracts people to careers in politics? Why do they appear to abandon deeply held principles once embedded in the system? Why do some turn rotten and steal from the societies that elected them? The structure-centric responses to these questions have been hotly debated for centuries. But, ultimately, the answers must be to do with brains.

When you vote for a representative, particularly when you choose on the basis of what you perceive to be their personality, you should have no reason to believe that you have made a logical choice. What, after all, really formed the basis of your choice? Usually you will not be personally acquainted with the politician in question. You may have seen them on the television a few times. They may represent a Party that you feel an affinity with. He/she may be 'the best of a bad bunch'. It's not much to go on.

People will choose to go into politics for a variety of reasons. Some will be 'conviction' politicians with a real sense of what they believe to be morally correct; others will see politics as a useful (and sometimes easy) career ladder; others will spot an opportunity for power and influence. Their capabilities will vary enormously: some will be from the intellectual elite (although the dearth of scientists in political life makes this less likely); others will struggle to think rationally and coherently. Either could end up running a nation.

It would be useful to have the tools to measure these intentions and capabilities before we cast our vote.

There is, of course, massive debate about how these factors can be reliably measured. For example, IQ tests are often discredited. Emotional intelligence in now given more credence but the markers are hard to identify. Psychometrics of one kind or another are often used as part of job interviews. Intensive psychological evaluations are undertaken on patients in psychiatric institutions. The tools are by no means perfect but perhaps they should utilised on those seeking to be our political representatives, and the results made freely available to us.

'League tables' feature often in the news at present: schools, hospitals (including individual surgeons), police forces and so on. League tables of political performance and consistency, while useful, are not what I am talking about here. If a person chooses to put him/herself forward to represent us and to have a measure of say in our lives at this most intimate level do we not need to know a great deal more about their intentions and capabilities before we are in a position to make a logical choice?

It could be argued that this kind of testing would be invasive or a breach of human rights. I would disagree. Testing would be a voluntary part of the qualification process to stand for elected office. No coercion would be required or involved.

It should be an ambition of an enlightened society to be represented by the right people. Not necessarily the best and brightest but a healthy combination of the brightest, most stable, least corruptible, most logical, most nurturing, least mendacious etc. Good eggs not rotten apples. Effective, understanding and striving voices. Not brutish, memetically infectious demagogues.
Achieving this may require assessment of candidate suitability via the most rigorous scientific testing measures available.

Can we rely on the enlightened intellectual and emotional altruism of the few 'incidentals' to shape our future societies? Or do we need to find a humane and reasoned method of bringing just those candidates to the fore?

23 September 2009

Computing with the Multiverse

Quantum computers are real. Their capabilities may be limited at present but they exist. This may not seem like an extraordinary statement until you fully consider what a quantum computer does.

It does not matter that the practical applications of even an advanced future quantum computer would likely be limited to factoring vast numbers for secure encryption purposes. In many ways conceiving of this potential 'killer app' is just a way to secure the funding to build them in the first place. No, what really matters is that the calculations involved just aren't possible within the bounds of this Universe alone.

So where are the calculations being done? They are being done in tandem with the requisite number of equivalent quantum computers, staffed by the requisite number of equivalent copies of the person/people running the quantum computers, in the requisite number of equivalent universes required to complete the calculation.

As the wonderful John Gribbin points out in his recent book 'In Search of the Multiverse', this is not equivalent to the 'phase spaces' used by mathematicians to undertake complex calculations requiring theoretical extra dimensions. 'Phase spaces' work but you cannot use them to do the kind of calculations a quantum computer can do - because the (to all intents and purposes infinite) computing power of the Multiverse is simply not available without one.

If this very real example of how the Multiverse can be used for a practical application leaves you reeling then you are not alone. There is no guarantee that even the operators of these devices are absorbing the full reality of the Multiverse when they talk about 'spin direction' and 'superposition'. If superposition works and can be used for calculation purposes then what does it matter?

At the quantum level - the level of the almost impossibly tiny - 'objects' do not behave as they do at our level. Their position is not fixed but consists of 'clouds' of positional probabilities. They can be in multiple places at the same time. This may seem impossible to us but it is a law of the nature of the quantum level. The property of 'fixedness' only arises at larger scales, including our own. This quantum property can be harnessed for computation by using quantum superpositions to 'represent' binary digits - allowing one to undertake inconceivably large numbers of calculations all at the same time.

My reading of Gribbin's explanation of the process is that the quantum computers in the various universes 'call' for the answer all at the same time. The superpositions allow the calculation to be 'split' and undertaken across the requisite part of the Multiverse. The answer is then instantly 'collapsed' back to the observer in each universe.

Why am I trying to explain this when I don't understand it and do not have the mathematical language to describe it? Because I feel a sense of wonder at this and wish to teach myself a language for describing it to myself.

I want to internalise it and I think it might be enjoyable for others to internalise it too.

16 September 2009

How's my RTPJ?

Have you always been aware of the beliefs of others or did you just grow that way?

With an inhibited Right Temporo-Parietal Junction (RTPJ) would you struggle to understand that others can have beliefs different to yours? Would this inhibition cloud your ability to make moral judgements? Current neuroscientific research using fMRI suggests increased activity in this small brain region when volunteers are tasked with thinking about various situations from the point of view of another human being.

Studies by Rebecca Saxe
and others see the RTPJ as being key to the morality aspects central to a cohesive "theory-of-mind". Her studies have found that the abilities of children to reason out and judge scenarios of "people thinking about thinking people" develop markedly and rapidly between the ages of approximately three and seven years old.

An example would be where a child, with the aid of props, was asked to envision a man putting a sandwich down on a box. The man then leaves and the sandwich gets blown off the box by the wind. A second man comes along and puts his sandwich down on the box, not seeing the one on the ground, then leaves. The child, once given the scenario, is asked which sandwich the first man will take when he returns. According to the Saxe studies the children would respond thus:
  • The 3-year-old says that the first man will take the sandwich on the ground although it is dirty because it is "his". When told that the first man actually takes the one on the box the child expresses surprise - presumably meaning that she cannot understand that the first man would not know that his was the one on the ground, so she thinks it unfair that he took the one on the box.
  • The 5-year-old says that the first man will take the sandwich on the box. This appears to show a more developed understanding of the thoughts of others because the child understands that the first man would think (albeit mistakenly) that his sandwich was the one on the box. However, the 5-year-old, still says that it is "bad" of the first man to take the sandwich on the box. Is this evidence that the child has not yet developed the moral capacity to know that the first man cannot be blamed for not knowing he was mistaken?
  • The 7-year-old says that the first man will take the sandwich on the box. Crucially, she also knows that the first man should take no blame for his mistake because it was a simple accident/misunderstanding.
This is a rather wordy explanation and you can see the Saxe talk on ted.com which should make it all clearer. She also brings up the fascinating results of studies, in adult subjects, where the RTPJ is stimulated (through the skull) via electromagnets. This appears to demonstrate a decreased ability in the subjects to make usual moral judgements while the RTPJ's function is being disrupted.

Other work, such as that of JP Mitchell does not appear to directly contradict the Saxe papers but does, again, bring up the issue of "localisation". He appears to be saying there is no current conclusive proof that the RTPJ is solely responsible for this kind of reasoning, despite the fact it lights up under fMRI when these judgement tasks are undertaken.

It must be very tempting for neuroscientists to fit specific cognitive functions to specific brain regions, especially now that fMRI studies seem to corroborate some of these theories. It's also much easier to explain to laypeople than telling them that fMRI studies suggest increased blood flow in areas that might be associated with a particular function when subjects undertake cognitive tasks that might stimulate blood flow to the region in question. While the "localisations" may be broadly correct there seems to be a bit too much "shoehorning" going on in some of these studies.

There are, of course, ethical issues connected to the potential ability to disrupt a person's ability to make reasoned moral judgements; or to actually change their beliefs by electro-mechanical means! This ability doesn't appear to be on the horizon any time soon. I have no fear of and can see a lot of value in this kind of work and, unlike Ms Saxe, I do think that this will help us to understand "the hard problem" of consciousness.

This is me signing off thinking about me thinking about you thinking about your beliefs about how you think about the thoughts of thinking people.