Friday, September 19, 2008

No Clue

We live in a world that is full of information. That is nothing new. What is new is that there are three factors that have increased so quickly over the last 8 years that they are on the verge of precipitous descent into uncontrolled states. The three factors are pretty obvious:
  • increasingly free access to information
  • increasing complexity of information items
  • increasing complexity of relationships between information items
A note here: the first is often taken as a good thing; this is not true for the simple reason that the free access is not equal, ubiquitous, instantaneous or rational.

We've been fed a mantra that the information age is good for you. This is true only by comparison to what can happen with insufficient information. The problem is an old one and has two components: 1) we are happy when we have enough information to solve existing problems or answer existing questions; 2) we are happier when we can find new problems to solve and new questions to answer. The two form a cycle, although the viciousness or virtuousity of the cycle is debatable, depending on the kind of problems. There is therefore never enough information, we always want more, and hence an age of increasing access to information is a Good Thing.

The sticking point perhaps is that there is a point at which we can no longer be our own data specialists. In the not-too-recent past, it was possible essentially to have restricted information access (or to restrict information access) and still be expert enough to get the job done. But people have incentives to stand out and hence earn greater rewards. So they recast the roles of professional, expert, master of a discipline, until they stand out. It is a question of whether your information brokerage level (whether it's a skill, a talent, a knowledge base, a database, a process, a capacity for synthesis etc) is outstanding enough.

It never will be, and that will eventually be because the human mind, no matter how complex or augmented, will never know enough. To address this based on my first three points, let's begin with access.

Access to information is a good thing when the information requirement is known in advance. For example, I need to know the best place for beer and pork knuckles in town. I ask people, search websites, do other information-gathering activities, until I have an answer I am satisfied with. Then I go and try out my search results physically. At the end of the experience, I will have updated my own database and am now qualified to offer my own opinion.

However, the questions of "What shall I eat?" (which according to Douglas Adams, is a relatively sophisticated question) and "Where shall I eat?" (which according to Adams is the most sophisticated question in this domain) are a lot more complex if I have no idea at all – or if I have too many ideas, brought on by a vastly greater access to information and travel capability than my ancestors had. With this, and the fact of human variability in opinion, I might face the centipede problem. That's often described as asking a centipede, "Which foot do you first move when you start to move?" and watching it attempt an answer, fail, and never walk again.

Access to information is inherently unfair, as it leverages off information sources and technology in depth, scale, range and speed. We keep talking about free markets, free speech and ideas, things like that; we fail to realise that unless information is ubiquitous, instantaneous, immediately processed, appropriately processed, and acted upon at once, all this is a very shady approximation to anything that could possibly be considered realistic. And yet we persist, perhaps beyond the limit of humanity's ability to make sense of it.

Which leads me to the second point. The information packages and 'chunks' we are presented with are sometimes far too complex. They cannot be evaluated fully or sometimes even usefully by us, our expert systems, or any other information processing tool we have. When the education system was a master teaching five apprentices in a workshop, the possible interactions and processes were far more limited that when the education system became a teacher teaching 25 students in a modern chemistry lab with internet access.

Where in the past, a sound-bite (like 'atoms are indivisible particles which our material universe is made of') could essentially be a workable fragment of reality, the present nature of education and information demands much more (like 'atoms are social constructs which can be considered to be statistical models of...') and sometimes too much more.

The modern sound-bite is of course not just a fragment of text encapsulating a key idea. Watching viral videos on YouTube or looking at campaign slogans and advertising literature, you will see that the entire semiotic domain is exploited at deeper and deeper levels. This is intended to affect us in ways that increasingly complicated analysis say might just be useful as the psychological leverage to push us into buying (into) things. We can no longer evaluate any of this easily; what's interesting is that neither can the merchants of materialism, the nabobs of nationalism, or the emirs of entertainmentism.

My third point is that as information chunks become more complex and more accessible, the number, kind and usefulness level of all their possible relationships spirals way out of control. Who is to know what might possibly be tangentially useful or critically useful? Only in very well-defined domains is this possible; domains such as a deliberately-limited math textbook or an intentionally-limited chemistry lab. With people inventing derivatives of data (information), second derivatives of data (information chunks and packages), third derivatives of data (knowledge packages and information systems), and fourth derivatives of data (information networks and dynamic knowledge bases), we are seeing an approach to the (we think) final level of the true artificial intelligence.

The likelihood is that this 'final' level will probably be centipeded after a while. Yes, it will be like us but smarter in some ways (mostly by more coherent and more rapid processing and access); yes, it will be like us in not knowing what questions to profitably ask for the unknown future.

And so, the correct thing to say is that unless one looks for some other standard by which to look at life, the universe and everything, the answer might as well be '42'. Information doesn't provide us, on its own, with answers to some questions. You might say, "I'm sorry, but I haven't a clue," to the question of what life is all about. Then, I would direct you to try some useful games.

Labels: , , ,

0 Comments:

Post a Comment

<< Home