Sergio Pissanetzky SergioSciControlscom Emergent Inference Any system VISION ROBOT SOFTWARE your mom grab an object computer program eyes cameras sensors translation 100000000 ID: 477456
Download Presentation The PPT/PDF document "Emergent Inference and the Future of NAS..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Emergent Inference and the Future of NASA
Sergio Pissanetzky
Sergio@SciControls.comSlide2
Emergent Inference
Any system
VISION
ROBOT
SOFTWARE
your mom
grab
anobjectcomputerprogrameyescameras,sensorstranslation100,000,000dots of lightmemorymemoryEI in the brainEI in the computerEI in the computer“hi, mom”next moveclasses and objectsyour momwith a hatnewenvironmentdevelopment
learningSlide3
● NASA needs AIT’s for manned and unmanned
space exploration.
● Autonomous robots can fly on their own or help the
astronauts. ● All our current technolgies are analyst-dependent.
● Extricating the analyst from software requires understanding the analyst.● Understanding the analist means understanding the
brain.
Analyst-independent technologies (AITs)Slide4
The brain
The brain is a computer
● Cognition
● In host-guest systems: guest
● We learn● We remember
● We know● We decide● Do we think?
● Do we reason?The brain is an organ
● Unconscious● In host-guest systems: host● Where is the unconscious?● What does it do?● Hermann von Helmholtz● Sigmund Freud● James Clerk Maxwell● Carl Jung● Douglas Hofstadter● Pink Floyd● David Eagleman● Juan Roederer● Jean Piaget● SergioANDI believe that cognition is a giant memory of behaviors, stored as a hierarchy of structures. Then, an unconscious mechanism that von Helmholtz called UI and I call EI generates new structures, and again stores them as behaviors in our cognition for us to use. Slide5
The Great Unsolved Automation Problems
of Software Engineering (GUAPs)
The automation of:
● Object generation (80’s)
● Refactoring (90’s) ● Integration (2000’s)
● Self-programming (2010’s) ● Parallel Programming ● Image Recognition
● Semantic Web ● Ontologies ● Robots
● ... and counting.All these problems have exact mathematical solutions that can be found by emergent inference. Slide6
chess
brain
drive
Jeopardy
play
chess
play Jeopardy
drive a car BRAINSlide7
SOFTWARE
chess program
quiz program
car driving program
chess
sensors
Jeopardy
sensorscarsensorschesscontrolsJeopardycontrolscarcontrolsEMERGENT INFERENCEproblem of Physicsdisorganizedsoftware
set of
programs
sequential
program
pixels
emergent inference
solution
integrated
program
recognized
image
refactored
+ objects
asignementSlide8
Traditional software development cycle
structure
structure
structure
PROGRAM
BRAIN (human analyst)
emergent
inferenceemergentinferenceemergentinference
• • •
• • •
knowledge
data
subject-matter
engineer
fileSlide9
History of mankind
HUMAN LABOR
TASKS
Source of
mechanical power
heavy lifting
dig trenches
propel a boatSource of Emergent Inference program computers control machines creative tasksDecision making- automating systems SINCE THEADVENT OF HUMANS INDUSTRIAL REVOLUTIONINTELLIGENCEREVOLUTIONThe world cost for non-automation is $60 trillion per year.You are hereSlide10
Plan and Conclusions
● EI should be developed in USA.
● NASA may develop EI.
● Step 1: independent verification.
● Step 2: development.● Step 3: application in robotics.
● All current technolgies are analyst-dependent.● Anayst-independent thecnologies are needed for
space exploration, both manned or unmanned.● EI provides the only way to extricate the analyst.Slide11
●
Structural Emergence in Partially Ordered Sets is the Key to Intelligence.
S. Pissanetzky. Lecture Notes in Artificial Intelligence, a subseries of Lecture Notes in Computer Science, LNAI 6830, pp. 92-101. Springer-Verlag.
●
Emergence and self-organization in partially ordered sets. S. Pissanetzky. To appear in Complexity.
● Emergent inference, or how can a program become a self-programming AGI system?
Sergio Pissanetzky. Workshop on Self-programming in AGI systems. AGI-11 conference, Google campus, Mountain View, CA. August 3-6, 2011.
Sergio@SciControls.comwww.SciControls.com2011 publications