|
Gmail Peter Burgess
(no subject)
1 message
alan longley
Tue, Oct 8, 2013 at 12:49 PM
To: Peter Burgess
Peter,
Hyperbolic trees could be used to give mobile apps the fluidity needed to incorporate your metrics.
The hyperbolic tree patent is owned by SAP. If you recall, I have a meeting with the CEO of Barti-Airtel, India's largest
private sector cell phone company, on the horizon (they have 23,000 engineers). I would like to suggest that they license the use of SAP's
patent on the hyperbolic tree, which is bundled with SAP's Big Data platform.
Licensing the hyperbolic tree patent to Bharti-Airtel could be good for SAP....
Just mashing together old messages here...will work on something more thoughtful....
Hyperbolic trees are useful for massive data sets, but also have ramifications for small scale projects, including design.
Dear Peter,
Thanks for your thoughts, which put me in a good mood and emboldened me to encapsulate some thoughts in the message I sent to Duncan; I sent a copy to you. The hyperbolic tree patent originated at Xerox, as did the mouse and windowing systems, which Steve Jobs and Bill Gates, respectively - did so well with. In the coming weeks I will clarify some of this and write to Veer about it.
You say that you are dangerously over-worked. I am willing to listen to suggestions as to how I might help you and might actually complete some tasks in this regard, but I am quite busy myself. I am elated about some math discoveries in nueral engineering. Economics MUST model brain activity. Why not get it right from the start?
Eye motion is particularly well understood. As you read this, about 1000 neurons in your brain stem are registering information from fifty degrees to you your left to fifty degrees to your right. Each neuron encodes information from a different segment covering a variable number of degrees - though overlapping - a nonlinear encoding system, over the 100 degree range. The most overlap occurs straight ahead or thereabouts. And it is this overlap that enables the decoding, which, surprisingly, is well modelled by linear methods. The heavy overlap in the center is why you are most comfortable staring there, while you must strain to view the sparse overlap at the periphery.
This redundant system of nonlinear encoding and linear decoding is useful for studying economic time series as well as building models of corporate and social systems, though it is largely unexplored and nowhere rigorously exposited. This is too bad for the poor, microfinance must be modeled with such systems. What else do we really have other than the brain to work with, especialy as the future emerges. Would the future exist if there were no brains? That is the future we bring on without brain modeling as a substrate for economics ( and truevalue metrics).
|
|
