Intelligence: ability to contextualize and consequentialise
Monday, October 26th, 2009This is a must-read from the Times. I like these points:
* The more intelligent someone is, the more they can see a fact in terms of other things. The greatest form of intelligence is someone who can make big links between different contexts, such as the scientist F. M. Burnet, who applied the principles of evolution to the immune system. — Professor Susan Greenfield, Director of the Royal Institution;
* The marks of intelligence are alertness, perceptiveness, wit, curiosity, creative responses to opportunities and problems, and the ability to learn quickly from errors. Intelligent people tend not to be mentally lazy or pedestrian, because being smart enough to recognise that one is either or both these things makes for dissatisfaction. Intelligent people are more often than not self-motivating and ambitious and derive pleasure from putting their talents to use. The value of what results depends, of course, on whether the intelligence in question is bent to good or bad ends. — A.C. Grayling, Professor of Philosophy, Birkbeck (ex-Oxford); and
* I don’t equate intelligence with cleverness. I think people who are intelligent have a touch of humanity about them. Their ideas, insight and vision set them apart from others, but they also have an understanding of what makes the world tick and how their ideas can impact for the greater good. Interestingly, as the World Wide Web has evolved so has the concept of collective intelligence, which is best encapsulated in the evolution of Wikipedia. This is a new form of intelligence that could lead to new insights into our understanding of the key challenges that face us as an increasingly global society. — Dame Wendy Hall, Professor of Computer Science, University of Southampton.
It’s interesting because 6 weeks ago this is what I wrote on BBCDigRev:
CONCLUSION: we are intelligent in ways not necessarily captured by current definitions of it or tests to infer it.
I have a similar take on consciousness; we are conscious in ways not necessarily captured by current definitions of it or tests to infer it.
This article is a comment on the Channel 4 program, Race and Intelligence: Science’s last Taboo.
MY VIEWS ON IQ-EQ, TURING TESTS ETC.
I’ve commented previously on this blog about my fun+games with Elbot, the Turing test and what is defined as machine intelligence and consciousness.
IT’S CLEAR THAT THE QUESTION OF INTELLIGENCE FOR ME IS ABOUT WHETHER SOMETHING (MAN OR MACHINE) CAN MAKE SENSE. By this I mean contextualize with diverse, multiple and sometimes contradictory sources of inputs, be able to filter in the perspicacious quality and out the nonsense noise, DNA path the consequences of each option, spark and cross-pollinate some synergistic random creativity (from what seems to be nothing, a vacuum, a pool of ignorance, unconnected silo sources) and…….DO SOMETHING WHICH MOVES THEM, THE PEOPLE AROUND THEM AND THE SITUATION THEY’RE IN FORWARD.
Now, superficially, this may seem to mean intellectual application. It doesn’t. For me, a footballer who can make sense of his terrain, filter out the noise of the crowds and filter in his manager shouting instructions from the sidelines yards away, be aware of where the rest of his team are, who he needs to pass to to move their game forward towards a goal, what paths to take to avoid defenders on the other team and then the random magic of foot connecting with ball to curve it into a net……….is intelligent.
Moreover, Ivy League / Oxbridge / Top 10 MBA school educations and qualifications don’t necessarily equate with intelligence. I’ve worked directly with types like that and some of them are simply incapable of making sense. If they were capable, we wouldn’t have had the global financial crisis in that gravity or of that magnitude and ripple effect. Yes, the nature of markets is that they are cyclical and we are bound to have booms+busts. Nonetheless, smarter anticipatory sense-making BEFORE the crisis could notably have decreased the severity of it and the policies available to take us out of it.
What highly qualified people may be good at is using convoluted jargon and “marketing mantras” but they can be poor at decision-making in any contextualized and consequential way. Unfortunately, people like that do rise into positions of sign-off so US$ billions of savers’ capital is put at risk precisely because those people are incapable of sense-making and risk managing.
Some of them are also simply not mathematically or technically competent enough.
Now, I’ve also been interested in the BBC Digital Revolution’s upcoming documentary which charts the last 20 years of the Web, particularly what it means for our collective intelligence and consciousness. This is one of my observations there about Program 4 which is about how the Web affects our intelligence:
* Baroness Susan Greenfield shares her concerns for our brains under the web’s influence; and
* Nicholas Carr offers his thoughts on the loss of the contemplative mind.
It’s been theorized that the Web will turn (or is turning?) us into non-thinking drones and rewiring our brains negatively. There may have been similar theories when the radio and the television happened.
What we’re showing in this thread is that actually the Web can INCREASE contemplation — with each contextualization by diverse contributors which elicits another contributor or reader (or the BBCDigRev team) to reflect and respond with a supporting / alternative / questioning position towards positive ends.
Positive ends being some type of preservation of the Utopian ideals of the Web first imagined by TBL and others’ genius AND also factoring in the corporate inclusion in the economic model — along with tools for corporate altruism, increased and incentivised online collaboration, and how WE should own its value ecosystems rather than any govt, corporation or Skynet wannabe.
So perhaps the basis of Program 4 shouldn’t be whether the Web is affecting our identities but rather how we’re shaping the Web and each other via online interactions.
INTELLIGENCE + IQ
============
Can we define intelligence without ambiguity or capture it precisely in IQ tests? Answer: NO.
The definitions of intelligence vary, culture by culture, individual by individual. In Chinese cultures someone who can cook well is said to be intelligent as much as if they can pass exams. In Western culture, is David Beckham’s ability to somehow instinctively figure out where his foot needs to hit the ball to bend it into the net not as intelligent as a Professor of Mathematics’ calculations of that projectile (force, velocity, rate of acceleration et al)? In Russia, are the oligarchs who engaged in illegal activities and are now in prison intelligent or is the guy who earns a regular salary and is free to take care of his family intelligent?
Relativism applies.
Besides which, IQ tests as a test for intelligence are quite silly. First of all, they only test for our visual-spatial reasoning — of the kind, “Can you figure out which is the odd one out?” That’s easy. If it’s one of those images with dots and crosses or a flag with stripes, we just have to see whether it’s a reflection / translation / rotation / inversion of the dots+crosses. If the number of dots+crosses changes, then it’s about an algebraic pattern.
They don’t test for our aural abilities to distinguish between people’s voices and their intent. They also don’t test for our actual manual dexterity or the aptitude of our olfactory senses. These too are signs of intelligence. The aural aspect of intelligence is particularly important in Oriental languages where the tone and accents of the vowel can determine whether we’re calling our mothers “Mother / a horse / hemp / measles / a nuisance.” Manual dexterity is a sign of intelligence — look at the artistry of pastry makers, sculptors, fashion designers, glass blowers. That doesn’t get captured in these IQ tests which are supposed to be a measure of intelligence. Ditto the intelligence of the “Noses” in the cosmetics and perfumes industry or even the intelligence of someone being able to smell out changes in the weather or where a piece of fish was caught.
CONCLUSION: we are intelligent in ways not necessarily captured by current definitions of it or tests to infer it.
So returning to machine intelligence because it has implications for Web intelligence. If, for example, we wanted to test a computer’s intelligence using standard IQ tests it would probably pass and be able to match the reflections / translations / rotations / inversions of images. Likewise, it would be able to complete all the computational mathematics ones readily — like “What’s next in this sequence? 2, 3, 5, 7?” Easy, they’re all prime numbers so the next one is 11.
Could a machine as readily get the words differentiation in the IQ test — of the kind, “Which is the odd one out? Dog, dolphin, bat and kangaroo.”
At the most, the machine could identify that they’re all mammals. Then it would apply binomial tree logic to distinguishing that the first is the only one with 4 legs, a dog + a dolphin can swim, a bat is the only one that can fly and a kangaroo carries its joey in its giant pouch, and the dolphin + the bat both have the same maximum hearing ranges (approx 150 Hz).
If you ask an Englishman, they may say, “The dog because if we read the word backwards it says “god” and that’s the only one of the options which can be read and spelt both ways. It’s a palindrome.”
If you ask an Australian, they may say, “The dolphin because it’s the only mammal which exhibits an ability for culture (http://www.newscientist.com/article/dn7475).”
If you ask someone Chinese, they may say, “The bat because it’s the only one in this group not known to have previously been eaten by Man.”
If you ask an African, they may say, “The kangaroo because it can only be found in one land — Australia — but the others can be found in many countries.
So, clearly, our inherited tests for intelligence are flawed and if we apply the same types of tests to machines then we will only end up with a definition of machine intelligence which is flawed.
Solutions?
Well, first, develop better Web tools which enable us to understand US and each other with more insight. Some attempt towards this is starting with the Semantic Web and initial forays into structuring data for contextualization. More, though, needs to be done on the perception, culture and values dimensions but at least it’s a positive evolution for the Web.
Second, discern what the value dimensions are within these diverse cultures.
Third, create more sophisticated collaboration tools that can harness those cultural variances for collective hopes and aims.
It’s also important to note that an increase in functional processing power of machines may not be the same as an increase in the intelligence of machines.
Again, it comes down to what is our definition of intelligence and is it culturally cross-applicable.
Anyone interested in how technologists define intelligence can read about it here:
*http://www.aaai.org/AITopics/pmwiki/pmwiki.php/AITopics/NatureOfIntelligence
Anyone interested in how biologists define intelligence can read about it here:
* Handbook of Intelligence, Robert J. Sternberg —
http://www.amazon.co.uk/Handbook-Intelligence-Robert-Sternberg-PhD/dp/0521596483
* http://www.nature.com/nature/journal/v407/n6803/abs/407470a0.html
AND FINALLY……ABOUT THE GLOBAL FINANCIAL CRISIS + TECHNOLOGY
=============================================
Here’s some food for thought. Ever since it struck there’s been reams of academic argument from the Smith/Friedman schools versus the Keynesian schools. We have the likes of Paul Krugman (Noble Prize for Economics), Joseph Stiglitz (also Nobel Economics), Nouriel Roubini and Niall Ferguson getting into intellectual fisticuffs over the economic models we’ve inherited and how to apply them now.
It’s apparent by now that the Presidents and Prime Ministers are taking their leads from these economic giants.
However, what’s interesting is that the Internet, climate change and the interdependence of nations did not exist in the days of Smith or Keynes in the shape they do now. This seems to be escaping the attentions of the heavyweight economists of today. TECHNOLOGY is facilitating trade and driving it. Electronic systems created by the banks and companies to produce, market, distribute and sell products+services to us are also interdependent — not simply government policies.
So perhaps it may be a good idea to evolve the Smith vs. Keynes models into ones which DO INCLUDE technologies (the Internet, mobile, haptics), climate change and systems interdependency.
Otherwise, if we keep using old tools and economic frameworks which haven’t kept apace with technological advances we shouldn’t be surprised if in 10 years time there’s another recession and risk of stockmarket collapse and values eradication.
See? It’s about a lot more than Chris Anderson’s “theory of free” or “freemium” or whatever. It’s about the whole and holistic economic ecosystem we’re creating online that provides us with tools more sophisticated and current than Smith or Keynes alone.
********
Yes, we’re all in it together moving forward :*).
Yes and I don’t believe that the likes of Twitter or Facebook will be regarded as the apex of Web development. As much as they’re useful and have their own validity, there’s still MUCH MUCH MORE AMAZING INNOVATION, COLLABORATION + SENSE-MAKING AHEAD.
At some point, that’s likely to include a re-imagination and re-configuration of code with some form of pictographics like Chinese and more haptics. I see inklings of how pictographs would level the code playing field; there’s something written in Squeak which allows children aged 11-15 to learn about code in a purely visual way rather than as lines of tags and text.
For true democracy, we’re not only talking about as it applies to adults and people from the same cultures or “intelligence” or “edu-economic status” as us. We’re talking about….EVERYONE.
Who knows? Maybe someone will get an Epiphany/enlightenment soon and figure out a way of incorporating Chinese into RDF in a way which is more than a call to an image file of a Chinese character.
For me, right now, I just want to replace 5-star rating systems with something smarter and more meaningful.
360-2020
The reason for my interest in contextualization and consequentiality is that I don’t believe the Semantic Web is sufficient. Within my lifetime, I’d like to experience the realization of an intelligent CONSCIOUS WEB.
If we can structure content not simply to be able to categorize the simple stuff like people, places, companies, location etc. in RDF, but to actually differentiate QUALITY sense-making content from the noise and also discern the ways people are perceiving and valuing that content, then we will become a whole lot more Enlightened and intelligent.
Even more intelligent is when the on+offline tools to contextualize and consequentiality path the wealth of our content and productivity will enable us to derive more evolved economic models and social ecosystems that advance — rather than atrophy — our species and our humanity towards others.