over 6 years on 2010-08-08


two different streams of consciousness of recent personal consideration:

the unexamined life is one thing, but the real issue is humans are network io bound:
consider that we are each a bundle of inputs, processing, and outputs with a limited amount of low latency local storage and low-bandwidth access to higher latency forms of massively scaled networked mildly query-able storage (cave paintings --> books --> internet). a network of peer machines working on centralizing a societal s3.

we all float around in a mesh network, and interact through protocols like english. over time, we find informational/input resources and other machines that engage us and through experience/shared experiences we can slightly up our communication rate/increase bandwidth/etc in local loops to create extra efficiency.

i think right now as a species we are far more network io bound than anything else... we end up re-doing the same experiments/processes over and over because we can't communicate what we learn fast enough -- technology is helping on the margin... fast pipes/copper helps a lot, but we have a long way to go, because as with cable the last mile hop is the hardest.

i have been personally feeling this acutely lately because i feel as though the data-set i am exposed to right now is especially interesting, i am committed to processing on it, and yet i am only able to express in re-usable formats a tiny amount of what i am learning to other nodes, let alone write it to our communal 'disk' in a re-usable format.

i feel like i can only remember 24 hours later a subset of what i perceive / am learning, i feel like i can only express to others in face-to-face high bandwidth communications a fraction of what i a can remember, i feel like i can record in a more useable format a fraction of what i can express... partially because i physically can't record it all, partially because social norms/other considerations prevent me from recording what i am learning.

so, the unexamined life... well, here is my point. the examination of life is essentially an exercise in local compression. it is taking surveys of the sensory input data you are exposed to and pulling out patterns. i think i try to do a good job of this, and i think that generally a lot of people reflect deeply. the issue isn't so much the examination, it is that sharing high quality data through our network is actually very hard. language is very slow and very limited, and are society is constructed in ways that only promote certain types of information sharing.

so, i am willing to posit that the genius of aristotle wasn't his thinking. i can't imagine he was the first person to have his 'insights' and locally come to the answers/ideas he came to. it was his ability to pass out that information in a way that could be spread widely. he did a good packing/compression/distribution job.

i wonder if even our perception of consciousness has something to do with the ratio between local processing speed/local memory vs. access to remote resources... and yes, i know i am sounding a lot like kurzweil / the borg.

and now for something completely different -- the end of meaningful monetary policy:
i think that we are rapidly approaching the utter end of effective monetary policy. the announced 'google price index', which i am both sure will have plenty of flaws, and will be a far more accurate view of the world than the government run cpi, was my most recent trigger for feeling this way.

ultimately, the entire value of the whole economy/everything can be thought of in two pieces - (current assets valued by what people are willing to pay for them today) + (the promise of future assets valued by what people expect others to pay in the future * people's trust in the future).

when we feel really good about the future we may be excited and value the future stuff more.... when we get depressed we value it less... and thus the economy swings.

good governments do all sorts of things to help us trust the future more (like the rule of law and property rights) and thereby created a lot of value in our above equation, bad governments erase a lot of value under the same math**. that really is the only role the government should be able to play in the economy.

that said, there is a bit of a trick/hack that governments have historically been able to deploy beyond their 'trust' factor. they can deploy "monetary policy", which is basically to say that when people are really depressed about the future, they could print pieces of paper to cheer everyone up, and when people were overly excited about the future, they could take away pieces of paper to make everyone chill out.

printing more money / monetary policy in general, at a truly theoretical level shouldn't work... our trust in the government (in our collective society) should already be priced into the equation, and you shouldn't be able to change that ratio at will by 'printing more units of trust/pieces of paper' -- it doesn't make rational sense. if there are $10 in the world representing all current value + future value * a discount rate, printing $10 should have no effect except to change the worldwide denominator.... but miraculously, there is evidence that it did work.

the reason is works is intermediate -- in any practical sense prices have historically been sticky. this is because the fact that there is $10 more in circulation isn't instantly known and absorbed by the economic system... people emotionally like to see charts that go up and to the right regardless of underlying reality, so they feel better just because they have more pieces of green paper.

the point..... well, prices, wages, etc. are not going to be sticky at all for very much longer. when all you need to do is flip a bit to update a price, and when you buy all your inputs on spot market pricing and efficiently use markets to hedge your basket, there is no such thing as sticky prices.... and when sticky prices go, so to goes the ability of governments to deploy monetary policy tactics.

so, ultimately, the cheaper communication gets the closer we could get to removing currency all-together and just transacting on real assets, etc. that is my favorite future state. but long long before we get to that, governments are going to loose their ability to 'print' trust / value and change our minds about the future.

that means that 'monetary policy' will be reduced to what it really is at the core -- which is the government's real ability to tweak the relationship between debtors and creditors (the probabilities of which should be priced in and shouldn't be able to effect anything), and voodoo.

-- if you are interested in this sort of stuff i highly recommend econtalk (http://www.econtalk.org/), especially the most recent podcast on the gold standard and

** in some cases historically when we didn't trust governments all that much, so even when we let them administer money, we forced them to tie it to simple rules we thought we understood and could trade off of (like one piece of paper per unit of gold)... but recently we have let the government have a bit more freedom...

original swl blogposts and letters 2007-2010