Perception means more than taxonomies?
Following is an sanitized version of an exchange I had with a friend of mine about taxonomies in relation to Semantic representations.
First, a Google Tech Talk on ‘Visual Perception with Deep Learning’ — 10 April 2008 :
————————————————————————————
I do appreciate the value of taxonomies. It’s the way most scientific classifications are arrived upon. For example, we have a periodic table structure which provides:
(i.) high-level overview of each of the seven groups of chemicals;
(ii) information about how many outer electrons are in each shell according to their membership of a particular group (this proxies parent-child identities); and
(iii.) how they could possibly connect and interact with another chemical from another group.
I also came across this interesting article on the ’10 Myths of Taxonomies’ which essentially highlights the risks involved in having to adopt someone else’s sets of taxonomies — particularly one in which we have no democratic voice or implementation:
* 10 Myths of Taxonomies — the Montague Review
IS THERE HOPE FOR HUMANS?
There seems to be this postulation amongst the Singularity brigade that
increased AI smartness ===> increased collective intelligence
The counter-argument would be…..
increased AI smartness ===> decreased collective human intelligence
(because they’ll have to think less)
AAAAARGH! QUICK! STOP THE GLOBAL BRAIN EXPRESS!
Personally, I don’t believe the proliferation of information pre-processed by AI necessarily makes each one of us SMARTER. It actually means the knowledge of how that AI was structured to pre-process the information is in the brains of a few select bleeding edge techies and everyone else is simply sheep-like or “dogs eat the dog food” like following whatever results are generated! (Incidentally, this is an anachronistic phrase from Microsoft circa 1988 still apparently being bandied around Silicon Valley — supposedly the cradle of avant garde progress, and let’s bear in mind we’re in 2008 and are supposed to be advancing humans not denigrating their intelligence or identities by ill-thought out associations).
It is possible we humans will erode our own capacity to perceive, to reason, to innovate and to communicate because some of us are too readily abdicating our responsibilities to do so to the machines!
However, we don’t have anything to really worry about because I’m going to share a simple truth…….
THE SINGULARITY IS A PIPE DREAM
The Singularity is unlikely to happen in the way the futurologists insist or according to their timescale of 2040. They all read far too much science fiction and need to more closely proxy terra firma.
Moreover, as evinced by my exclusion from a certain SemWeb some form of ideological eugenics and police state is going to apply to the Global Brain. Not everyone will be invited or have access to plug into it. As much as technologists bleat and spin on like autobots about open source and open house, Terms of Service and contracts (as well as undemocratic feudal warlord oppression of the serfs) will still apply as it has since land laws of C12th England.
TechCo SAYS it’s open this and open that, Google and wannabe minnows alike. However, notice how each tech platform imposes its (sometimes Draconian and contradictory) rules and evicts those whose opinions are different (or threatens to as per the unimaginative and inflexible mindset of the Administrator — please also see Kafka’s writings or Orwell’s ‘Animal Farm’).
This is why the Global Brain is never going to be truly global, open or democratic. The controllers/owners can be dictators and the constituents are the conformists, not necessarily by choice or personal conviction but by consideration towards others which is not reflected in a similar equivalence of respect towards them.
Interestingly, it’s well-documented that it could be the unconventionals who “think out of the box” and may be touched by genius or madness, and create the quantum-level breakthroughs in thinking — please see Einstein, Dali, Dosteovsky, other artistic intellectuals and bipolarly brilliant but insane people (more often than not the world’s greatest comedians).
INTELLIGENCE: THE KIDS IN THE CLASSROOM + ONLINE
Let me make this observation. In a class of 30 kids who are 5 years-old they all have exactly the same access to the same stimulus, reading materials, toys, etc. Yet there will always be 1 kid who is knock out quicker and brighter than their peers.
In the same way, every teenager has the same access to the index of information on the Internet. Yet only 1% of those teenagers will actively go in search of the wealth of knowledge with the specific intent of increasing their own know how.
Technologists and educators imagine that if kids get more tools this will make them smarter.
Not necessarily. It depends on their pre-existing mental orientation, character traits and emotional sense of reward about knowledge. Some of them assign 0 value to knowledge and +1 to Grand Theft Auto rankings.
FUN+GAMES
Certain proponents of the Global Brain theories and the Singularity are probably breathing huge sighs of relief I’m nowhere near them to present obvious and prescient counter perspectives. They can postulate we’ll all have boosted collective IQs (by the osmosis of proximity to smarter AI) without my notes of pragmatic perspective. Oh and we haven’t factored in the ego element of consciousness and self-awareness yet either!
Ack! Taxonomies is one thing. If the perception of the person doing the classification is not spot on or nuanced in the first place the mandolin ends up being related to a string instrument because it looks like violin (plus dyslexia or plain ignorance could be involved).
It’s not the taxonomic tags that show Paris is either a city, a personality or a romantic aspiration which ultimately matters, imo. It’s actually what emotions it evokes and elicits in a person: nostalgia, disregard or longing which help each of us discern meaning from the world and intakes around us.
PERCEPTIONS MORE THAN TAXONOMIES
I’m leaving others to think out the taxonomies issue. The perception dimension has interested me more since childhood. In terms of semantics, graphing the connections that a grape is a fruit that we eat and make wine out of and grows on a vine is less interesting and challenging to me than trying to figure out WHY people like to eat grapes, to drink wine and designate all kinds of adjectives to it — sweet, succulent, tannin aftertaste, globulous, shriveled, meaty, aromatic, green, over-ripe etc. — which reveal their personalities, life aspirations and purchase decisions.
It’s the perception dimension I’m coding a tool to capture…………………….
