Chaining banker bots together

OK, we have seen that banker bots manage queues, and bot define the number 1 has having 1 or 2 bets in their queue.  This is the same as the Free Banking open market for Seigniorage shares in their banker bot because the queueing banker bot will adjust its effective sample rate when the queue gets too long.  This is the basic idea in the theory of everything, match variance to quant by adjusting local queue size. And adjusting quant size is equivalent to adjusting sample rate, so it all comes down to spectral adjustments in finite networks.

Coming up next, how to the bots arrange themselves in networks. They adjust queue size between each other. This will be automatic because they are all no arbitrage, whether betting or taking bets.  I will go into that as I figure it out. But the end result, for the currency network, is a bunch of bot managing one currency each, and they all attach themselves to the savings and loan bot managing the cirtual currency. The savings and loan bot will subdivide itself as needed to match its member bank variance with the number of member banks, making the bid to cover nearly constant throughout. What that means is that collectively, they are making compact polynomials our of compact polynomials, doing the finite lag. Now the reason finite log is sow doable, when before is was difficult, is because the TOE assumes a self adapting process, it is making the unit 'thing' relatively right sized so finite log works.

Making the unit 'thing' properly sized is why the semantic networks need a precision of at least six bits, it lets them adjust its 'meaning' close to the unit 'meaningful'. Once the artificial brain has stableized a unit ;meaningful' then it has a causal ordering of all things and its minimum redundancy groups become the theory of the collective. It has leaned the aggregate theory of operation.

Learning to drive:
For the bot, learning to drive is a matter of adjustable huffman encodeing of all the  sensory inputs, which dividing up the various stages to have the same queueing length.  When that process is done, then the output is one of 2^n-1 possible actions that will be taken in response to the vaious inouts.  These are the minimum redundant action, steeal slightly left or right, brake, accelerate, etc.When the feat is learned, then all the possible actions are matched to sets of inputs, and optimally divergent from each other. Splitting and regrouping banker bot until the maximal causality structure is found is like defining the Shannon encoder, but you are allowed to change SNR along the way.