Information Technology Reference
In-Depth Information
This approach assumes that a group of hits come periodically. The weak synapse
S1 is to provide a single pulse for accuracy. Delay is fixed to match the times
between multiple returns. The symbols 1#2 means one return but not two; 2#3
means two returns in a group, but not three; 3#4 means three returns in a group, but
not four. When a group ends, an enable calculations signal is emitted.
For three hits in a sequence, Hit3, Hit 2, and Hit 1 are true; Hit 4 is false; output
of U1, U2, and U3 is true; output of U4 is false. So U6 has a true output that is
transferred through U7 to provide an enable calculations signal.
Conclusions
Over countless millennia, many billions of neurons and many trillions of synapses
have evolved. Electrical engineers and others have learned how to deal with circuit
complexity as expected in a brain. One approach is to model calculations as arrays
of controlled toggles.
Only the most important attributes are encoded into weights to contribute to
priority, greatly simplifying the calculations. The priority of an image is simply the
sum of these weights.
Simulated qubits in the form of controlled toggles are being proposed as a good
way to compute priority, the computation of which is initiated by a Hit signal from
long-term memory. This computation requires code drawn from a special section of
long-term memory. When Enable Calculations goes true, the priorities are
computed in parallel and made available. Priorities are then compared. That
image with the highest priority is subsequently gated into consciousness.
The rate of images going through consciousness is estimated to vary from
roughly 40 Hz without intervention of a cue editor to about 3 Hz for difficult
situations involving ambiguous cues and multiple returns. A quick decision within
one-third of a second is thus possible with parallel processing, and enhances a
person's chances of survival.
References
1. Franklin S (1995) Artificial minds. MIT Press, Cambridge, MA
2. Anderson J (1983) The architecture of cognition. Harvard University Press, Cambridge, MA
3. Dennett DC (1991) Consciousness explained. Back Bay Books, New York
4. Pagiamtzis K, Sheikholeslami A (2006) Content-addressable memory (CAM) circuits and
architectures: a tutorial and survey. IEEE J Solid State Circuits 41(3):712-727
5. Mano MM (1979) Digital logic and computer design. Prentice-Hall, Englewood Cliffs, NJ
Search WWH ::




Custom Search