Coupling of rational agents to quantum processes

Abstract

The present invention provides devices, methods, and systems for coupling a rational agent to a quantum process. In particular, the present invention provides rational agents configured to influence a quantum process, or to derive information from a quantum process, and methods and uses thereof.

1 - 20 . (canceled) 
     
     
         21 . A computing system for self-optimizing performance of an artificial intelligence agent executed by the computing system, the computing system comprising:
 one or more processors; and   one or more memories storing instructions that, when executed by the computing system, cause the computing system to perform a process comprising:
 receiving output from a quantum process, wherein the quantum process is influenced in accordance with an intent of the artificial intelligence agent; 
 selecting, based on the received output, an action to be performed by the artificial intelligence agent, wherein the influence of the intent on the quantum process increases a likelihood of the artificial intelligence agent selecting a preferred choice; and 
 performing the selected action in accordance with the intent, wherein, in performing the action, the artificial intelligence agent further exerts influence upon the quantum process, optimizing performance of the artificial intelligence agent. 
   
     
     
         22 . The computing system of  claim 21 , wherein the output from the quantum process is a bitstream received from a hardware device including a quantum random number generator. 
     
     
         23 . The computing system of  claim 21 , wherein the quantum process is influenced in an absence of electrical or mechanical feedback, from the artificial intelligence agent to a hardware device that generates the output. 
     
     
         24 . The computing system of  claim 21 ,
 wherein the artificial intelligence agent is configured to rank two or more actions available to the artificial intelligence agent based on an algorithmically-predicted quality of each of the two or more actions; and   wherein the selecting is performed by selecting an action, from the two or more actions, based on the ranking.   
     
     
         25 . The computing system of  claim 21 ,
 wherein the output from the quantum process comprises a bitstream including at least a first type of bit and a second type of bit; and   wherein the selecting comprises selecting an action mapped to a type of a selected bit in the bitstream.   
     
     
         26 . The computing system of  claim 21 , wherein the artificial intelligence agent is a computer program stored in a memory on the computing system. 
     
     
         27 . The computing system of  claim 21 , wherein selecting the action is further based on a predetermined desirability of an outcome of the action to the artificial intelligence agent. 
     
     
         28 . A method for evaluating an influence of an artificial intelligence agent on a quantum process, the method comprising:
 receiving output, of the quantum process, having a first sequence including two or more values;   ranking two or more tasks to determine at least a higher-ranked task that is ranked higher than a lower-ranked task;   associating the higher-ranked task with a first value of the two or more values;   associating the lower-ranked task with a second value the two or more values;   performing a first iteration, using the output of the quantum process as current output, comprising:
 a) performing the higher-ranked task for each occurrence of the first value in the current output; 
 b) performing the lower-ranked task for each occurrence of the second value in the current output; 
   generating a pseudo-random output, having a second sequence of the two or more values;   performing a second iteration of steps a) and b) using the pseudo-random output as the current output; and   determining deviations between the results of the first and second iterations.   
     
     
         29 . The method of  claim 28 , wherein the output of the quantum process is produced as a bitstream. 
     
     
         30 . The method of  claim 28 , wherein the output of the quantum process is produced by a quantum random number generator configured to convert electrical signals indicative of a quantum process to a corresponding bit pattern. 
     
     
         31 . The method of  claim 28 , wherein the ranking of the two or more tasks is based on an algorithmically-predicted quality of each of the two or more tasks. 
     
     
         32 . The method of  claim 28  further comprising providing an evaluation of the influence of the artificial intelligence agent on the quantum process based on the determined deviations. 
     
     
         33 . The method of  claim 28 , wherein the artificial intelligence agent is controlled, at least in part, by instructions executed by a computing system. 
     
     
         34 . The method of  claim 28 , wherein the artificial intelligence agent is controlled, at least in part, by a neural network. 
     
     
         35 . The method of  claim 28 , wherein the two or more values are two values each corresponding to 1 or 0. 
     
     
         36 . A computer-readable storage medium storing instructions that, when executed by a computing system, cause the computing system to perform a process for identifying intent ability in an agent, the process comprising:
 detecting an output, from a quantum process during and/or following the agent performing one or more actions,
 wherein the one or more actions of the agent were based on a previous output from the quantum process, and 
 wherein the agent does not provide electronic feedback to the quantum process; 
   comparing the output from the quantum process with one or more control values and, in response, determining that the output has at least a specified amount of variance from the one or more control values; and   based on the determining, providing an indication that the one or more actions of the agent are based on intention.   
     
     
         37 . The computer-readable storage medium of  claim 36 , wherein determining that the output has at least a specified amount of variance comprises determining a difference in a) an entropy of the output and b) an entropy of one or more control values. 
     
     
         38 . The computer-readable storage medium of  claim 36 , wherein the one or more control values is a pseudorandom sequence of bits. 
     
     
         39 . The computer-readable storage medium of  claim 36 , wherein the one or more actions of the agent are based on the previous output by the previous output being a sequence of two or more types of values and the agent taking each of the one or more actions based on each action corresponding to a) a place in the sequence and b) an action predetermined to map to each of the two or more types of values. 
     
     
         40 . The computer-readable storage medium of  claim 36 , wherein the agent is controlled, at least in part, by a neural network.