SlideShare a Scribd company logo
1 of 20
Download to read offline
Machine Learning, Artificial General
Intelligence, and Robots with Human Minds
David Peebles
Reader in Cognitive Science
Department of Psychology
April 25, 2017
Cognitive science
Talk outline
Structure
Discuss different types of artificial intelligence (AI)
Human-level artificial general intelligence (AGI)
Human cognitive architecture as a basis for AGI
Introduce a new cognitive robotics project
Key aims
Allay your fears of a possible AI invasion and/or enslavement
by super-intelligent AI systems in the foreseeable future
Enthuse you about the prospect of human-level AI
To entertain you with amusing and informative videos
Recent advances in artificial intelligence
Some recent headlines. . .
Artificial intelligence: We’re like children playing with a
bomb. Sentient machines are a greater threat to humanity
than climate change... (Observer. June, 2016)
Will democracy survive big data and artificial
intelligence? (Scientific American. Feb, 2017)
The AI threat isn’t SkyNet, it’s the end of the middle class
(Wired Business. Feb, 2017)
Stephen Hawking, Elon Musk, and Bill Gates warn about
artificial intelligence (Observer. Aug, 2015)
DESTROYING OURSELVES: Artificial intelligence poses
THREAT to humanity – shock report (Express. Jan, 2017)
Are intelligent machines really upon us?
It depends what you mean by “machine intelligence”
Three types of AI:
Artificial Narrow Intelligence (ANI). Machine intelligence
that equals or exceeds human intelligence or efficiency, but
only in one specific area.
Artificial General Intelligence (AGI). Machines with general
human intelligence, capable of sustaining long-term goals and
intent, which can successfully perform any task that a human
might do.
Artificial Super-intelligence (ASI). Machine intelligence that
is much smarter than the best human brains in practically
every field, including scientific creativity, general wisdom and
social skills (Bostrom, 2014).
The importance of distinguishing AI
The three types of AI are often conflated or confused in the
press and public mind.
All of the recent developments are in Narrow AI due to:
Faster hardware (GPUs)
Huge amounts of data
New statistical machine learning methods (e.g., deep learning
in neural networks)
Main machine learning methods
Supervised learning (classification and regression)
Unsupervised learning (clustering)
Reinforcement learning (optimisation via reward signal)
Narrow AI
Benefits
Rapid conversion of unstructured information into useful and
actionable knowledge.
Narrow AI will provide tools to automate and optimise this
process in science, medicine, business etc.
Costs
Elimination of jobs (mainly white-collar knowledge work).
Concentrations of knowledge, power and resources may
exacerbate inequality.
‘Black box’ systems where humans don’t know why AI
systems are making decisions
Limitations
Require extensive training
Each system is specific to the data set it’s trained on. One
slight change requires retraining
Artificial General Intelligence
“Machines with general human intelligence, capable of sustaining
long-term goals and intent, which can successfully perform any
task that a human might do”.
More interesting for cognitive science
Much more complex and harder to achieve than narrow AI
Some recent examples from Hollywood. . .
Ex Machina (2014) Robot and Frank (2012) Prometheus (2012)
Criteria for advanced autonomous agents
US NSF $16.5M Smart and Autonomous Systems (2016)
Specified five desired capabilities:
Cognisant. Exhibit high-level cognition and awareness
beyond primitive actions.
Taskable. Users can specify desired behaviours and
outcomes in a natural and concise (possibly vague) manner
Reflective. Can adapt behaviour and learn new behaviour
from experience, observation, and interaction with others.
Ethical. Adhere to a system of societal and legal rules. Do
not violate accepted ethical norms.
Knowledge-rich. Semantic, probabilistic and commonsense
reasoning and meta-reasoning about uncertain, dynamic
environments. Able to acquire and understand knowledge.
Advanced autonomous agents
Long-term autonomy (able to operate unaided over time)
Flexible assistants and agents in applications such as
manufacturing, military, agriculture, health, space etc.
Robots that can be trained for new task using natural
language rather than having to be re-programmed.
Robots that can perform complex task planning without
human intervention
NASA Mars Exploration Rover NASA Valkyrie robot
A key human skill – knowledge transfer
Excel spreadsheet
SPSS spreadsheet
Cognitive architectures
Originated in 1950s but active research programme in 1980s.
Cognitive science – differs from mainstream “narrow” AI and
traditional “divide and conquer” approach of experimental
cognitive psychology.
Theories of the core, immutable structures and processes of
the human cognitive system.
Integration of multiple components of cognition.
Computer architecture
Perception Action
Task
Environment
Learning
Procedural
Perceptual
Learning
Learning
Declarative
Selection
Action
Short−term memory
Long−term memory
Procedural
Long−term memory
Declarative
Human cognitive architecture
Example 1 – Interactive task learning
Rosie (Univ. Michigan). Soar agent able to learn new tasks
specified in natural language and interactive demonstration.
Complex skill integrating natural language processing, vision,
logical reasoning, search and motor control.
V1: Learns several different games and puzzles through
natural language descriptions of the legal actions and goals of
the task.
V2. Learns new goals from single examples, either by being
shown a visual example of the goal or from executing a series
of actions to achieve the goal.
Agent’s hypotheses are refined and corrected by language
input from human instructor.
Video: Simple task learning
Video: Tower of Hanoi puzzle
Example 2 – Mind reading and mental simulation
Cognitive architecture installed on a MDS (Mobile / Dexterous
/ Social) robot
Aim: AA that can construct accurate mental representations
of other agents’ mental states and perspectives in order to
reason about their beliefs and intentions. In order to:
Coordinate actions with teammates effectively
Anticipate teammates’ (potentially ambiguous) actions
Understand shared team goals of the team
Communicate more naturally.
Current tasks include gaze following, hide and seek, task
interruption and resumption, theory of mind.
Agent creates mental model of teammates’ beliefs and
(explicitly stated) goals and predicts what they will do next.
Video: Gaze and gesture interpretation
The ‘robot with a human mind’ project
NAO robot
Hadeel Jazzaa
Lee McCluskey
A flexible research robot
Video: NAO robot
Video: Robot football
The ‘robot with a human mind’ project
NAO robot
NAOqi architecture
Visual
Module
ACT−R Buffers
Environment
Pattern
Matching
Execution
Production
Module
Motor
Problem
State
Declarative
Memory
Procedural
Memory
Control
State
ACT-R cognitive architecture
Aims of the project
Embodied cognition constrained by perceptuo-motor
processing.
Goal directed behaviour. Including self generated goals.
Cognitive control and resource management over different
time scales (including long-term).
High-level cognition. Planning, reasoning, problem solving,
decision making, reflection and metacognition, analogy.
Verbal and nonverbal communication. Interaction and
adaptation through natural language understanding (speech,
text, demonstration.
Constant adaptation and learning. Human scale (i.e.,
relatively fast and not requiring thousands of training trials).
Integration of cognitive functions
Take home messages
Types of AI
Narrow AI is going to change the world radically but is limited
AGI is a much more challenging problem and is a long way off
Super-intelligent AI is a very long way off
AGI:
Requires the integration of many aspects of intelligence
(learning, reasoning, planning, self-monitoring etc.)
Will require insights from cognitive science as human
cognition is currently our best example of AGI
Cognitive architectures are a potentially very useful approach
to developing systems with AGI

More Related Content

What's hot

Knowledge based systems -- introduction
Knowledge based systems -- introductionKnowledge based systems -- introduction
Knowledge based systems -- introductionjkmaster
 
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014lebsoftshore
 
Cognitive Computing - A Primer
Cognitive Computing - A PrimerCognitive Computing - A Primer
Cognitive Computing - A PrimerMarlabs
 
Towards which Intelligence? Cognition as Design Key for building Artificial I...
Towards which Intelligence? Cognition as Design Key for building Artificial I...Towards which Intelligence? Cognition as Design Key for building Artificial I...
Towards which Intelligence? Cognition as Design Key for building Artificial I...Antonio Lieto
 
Artificial intelligence original
Artificial intelligence originalArtificial intelligence original
Artificial intelligence originalSaila Sri
 
Computational Intelligence: concepts and applications using Athena
Computational Intelligence: concepts and applications using AthenaComputational Intelligence: concepts and applications using Athena
Computational Intelligence: concepts and applications using AthenaPedro Almir
 
Artificial intellegence
Artificial intellegenceArtificial intellegence
Artificial intellegencegeetinsaa
 
Artificial Intelligence Course- Introduction
Artificial Intelligence Course- IntroductionArtificial Intelligence Course- Introduction
Artificial Intelligence Course- IntroductionMuhammad Sanaullah
 
What is Deep Learning?
What is Deep Learning?What is Deep Learning?
What is Deep Learning?Ahmed Banafa
 
Artificial Intelligence Techniques In Power Systems Paper Presentation
Artificial Intelligence Techniques In Power Systems Paper PresentationArtificial Intelligence Techniques In Power Systems Paper Presentation
Artificial Intelligence Techniques In Power Systems Paper Presentationguestac67362
 

What's hot (20)

Knowledge based systems -- introduction
Knowledge based systems -- introductionKnowledge based systems -- introduction
Knowledge based systems -- introduction
 
Cognitive computing
Cognitive computingCognitive computing
Cognitive computing
 
Unit 1
Unit 1Unit 1
Unit 1
 
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014
Lebanon SoftShore Artificial Intelligence Seminar - March 38, 2014
 
Cognitive Computing - A Primer
Cognitive Computing - A PrimerCognitive Computing - A Primer
Cognitive Computing - A Primer
 
Towards which Intelligence? Cognition as Design Key for building Artificial I...
Towards which Intelligence? Cognition as Design Key for building Artificial I...Towards which Intelligence? Cognition as Design Key for building Artificial I...
Towards which Intelligence? Cognition as Design Key for building Artificial I...
 
Lect # 2
Lect # 2Lect # 2
Lect # 2
 
Artificial intelligence original
Artificial intelligence originalArtificial intelligence original
Artificial intelligence original
 
Introduction to AI
Introduction to AIIntroduction to AI
Introduction to AI
 
Computational Intelligence: concepts and applications using Athena
Computational Intelligence: concepts and applications using AthenaComputational Intelligence: concepts and applications using Athena
Computational Intelligence: concepts and applications using Athena
 
Artificial intellegence
Artificial intellegenceArtificial intellegence
Artificial intellegence
 
Artificial Intelligence Course- Introduction
Artificial Intelligence Course- IntroductionArtificial Intelligence Course- Introduction
Artificial Intelligence Course- Introduction
 
What is Deep Learning?
What is Deep Learning?What is Deep Learning?
What is Deep Learning?
 
Artificial Brain - Overview 2013
Artificial Brain - Overview 2013Artificial Brain - Overview 2013
Artificial Brain - Overview 2013
 
(Ch#1) artificial intelligence
(Ch#1) artificial intelligence(Ch#1) artificial intelligence
(Ch#1) artificial intelligence
 
Gervigreind
GervigreindGervigreind
Gervigreind
 
Artificial Intelligence Techniques In Power Systems Paper Presentation
Artificial Intelligence Techniques In Power Systems Paper PresentationArtificial Intelligence Techniques In Power Systems Paper Presentation
Artificial Intelligence Techniques In Power Systems Paper Presentation
 
An overview on ai
An overview on aiAn overview on ai
An overview on ai
 
Ai introduction
Ai introductionAi introduction
Ai introduction
 
Unit I What is Artificial Intelligence.docx
Unit I What is Artificial Intelligence.docxUnit I What is Artificial Intelligence.docx
Unit I What is Artificial Intelligence.docx
 

Similar to Machine Learning, Artificial General Intelligence, and Robots with Human Minds

Artificial intelligence in software engineering ppt.
Artificial intelligence in software engineering ppt.Artificial intelligence in software engineering ppt.
Artificial intelligence in software engineering ppt.Pradeep Vishwakarma
 
AI_01_introduction.pptx
AI_01_introduction.pptxAI_01_introduction.pptx
AI_01_introduction.pptxYousef Aburawi
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligenceNimesh_parmar
 
Unit 1 ETI notes.pde`e1eeewdddfffffee1q2
Unit 1 ETI  notes.pde`e1eeewdddfffffee1q2Unit 1 ETI  notes.pde`e1eeewdddfffffee1q2
Unit 1 ETI notes.pde`e1eeewdddfffffee1q2paithankarpratik3
 
EELU AI lecture 1- fall 2022-2023 - Chapter 01- Introduction.ppt
EELU AI  lecture 1- fall 2022-2023 - Chapter 01- Introduction.pptEELU AI  lecture 1- fall 2022-2023 - Chapter 01- Introduction.ppt
EELU AI lecture 1- fall 2022-2023 - Chapter 01- Introduction.pptDaliaMagdy12
 
ARTIFICIAL INTELLIGENCETterm Paper
ARTIFICIAL INTELLIGENCETterm PaperARTIFICIAL INTELLIGENCETterm Paper
ARTIFICIAL INTELLIGENCETterm PaperMuhammad Ahmed
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.pptEithuThutun
 
Artificial Intelligence for Business.ppt
Artificial Intelligence for Business.pptArtificial Intelligence for Business.ppt
Artificial Intelligence for Business.pptFarhanaMariyam1
 
901470_Chap1.ppt about to Artificial Intellgence
901470_Chap1.ppt about to Artificial Intellgence901470_Chap1.ppt about to Artificial Intellgence
901470_Chap1.ppt about to Artificial Intellgencechougulesup79
 

Similar to Machine Learning, Artificial General Intelligence, and Robots with Human Minds (20)

ar
arar
ar
 
Artificial intelligence in software engineering ppt.
Artificial intelligence in software engineering ppt.Artificial intelligence in software engineering ppt.
Artificial intelligence in software engineering ppt.
 
AI_01_introduction.pptx
AI_01_introduction.pptxAI_01_introduction.pptx
AI_01_introduction.pptx
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
Ai introduction
Ai  introductionAi  introduction
Ai introduction
 
Unit 1 ETI notes.pde`e1eeewdddfffffee1q2
Unit 1 ETI  notes.pde`e1eeewdddfffffee1q2Unit 1 ETI  notes.pde`e1eeewdddfffffee1q2
Unit 1 ETI notes.pde`e1eeewdddfffffee1q2
 
EELU AI lecture 1- fall 2022-2023 - Chapter 01- Introduction.ppt
EELU AI  lecture 1- fall 2022-2023 - Chapter 01- Introduction.pptEELU AI  lecture 1- fall 2022-2023 - Chapter 01- Introduction.ppt
EELU AI lecture 1- fall 2022-2023 - Chapter 01- Introduction.ppt
 
ARTIFICIAL INTELLIGENCETterm Paper
ARTIFICIAL INTELLIGENCETterm PaperARTIFICIAL INTELLIGENCETterm Paper
ARTIFICIAL INTELLIGENCETterm Paper
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
AI_Intro1.ppt
AI_Intro1.pptAI_Intro1.ppt
AI_Intro1.ppt
 
Chap1.ppt
Chap1.pptChap1.ppt
Chap1.ppt
 
901470_Chap1 (1).ppt
901470_Chap1 (1).ppt901470_Chap1 (1).ppt
901470_Chap1 (1).ppt
 
Artificial Intelligence for Business.ppt
Artificial Intelligence for Business.pptArtificial Intelligence for Business.ppt
Artificial Intelligence for Business.ppt
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
901470_Chap1.ppt
901470_Chap1.ppt901470_Chap1.ppt
901470_Chap1.ppt
 
901470 chap1
901470 chap1901470 chap1
901470 chap1
 
901470_Chap1.ppt about to Artificial Intellgence
901470_Chap1.ppt about to Artificial Intellgence901470_Chap1.ppt about to Artificial Intellgence
901470_Chap1.ppt about to Artificial Intellgence
 

More from University of Huddersfield

Modelling alternative strategies for mental rotation
Modelling alternative strategies for mental rotationModelling alternative strategies for mental rotation
Modelling alternative strategies for mental rotationUniversity of Huddersfield
 
Multiple representations and visual mental imagery in artificial cognitive sy...
Multiple representations and visual mental imagery in artificial cognitive sy...Multiple representations and visual mental imagery in artificial cognitive sy...
Multiple representations and visual mental imagery in artificial cognitive sy...University of Huddersfield
 
Multiple representations talk, Middlesex University. February 23, 2018
Multiple representations talk, Middlesex University. February 23, 2018Multiple representations talk, Middlesex University. February 23, 2018
Multiple representations talk, Middlesex University. February 23, 2018University of Huddersfield
 
Two methods for optimising cognitive model parameters
Two methods for optimising cognitive model parametersTwo methods for optimising cognitive model parameters
Two methods for optimising cognitive model parametersUniversity of Huddersfield
 
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012University of Huddersfield
 
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...University of Huddersfield
 
A cognitive architecture-based modelling approach to understanding biases in ...
A cognitive architecture-based modelling approach to understanding biases in ...A cognitive architecture-based modelling approach to understanding biases in ...
A cognitive architecture-based modelling approach to understanding biases in ...University of Huddersfield
 

More from University of Huddersfield (12)

Multiple processes in graph-based reasoning
Multiple processes in graph-based reasoningMultiple processes in graph-based reasoning
Multiple processes in graph-based reasoning
 
Modelling alternative strategies for mental rotation
Modelling alternative strategies for mental rotationModelling alternative strategies for mental rotation
Modelling alternative strategies for mental rotation
 
Robert Gordon talk, 21 March 2018
Robert Gordon talk, 21 March 2018Robert Gordon talk, 21 March 2018
Robert Gordon talk, 21 March 2018
 
EWIC talk - 07 June, 2018
EWIC talk - 07 June, 2018EWIC talk - 07 June, 2018
EWIC talk - 07 June, 2018
 
Multiple representations and visual mental imagery in artificial cognitive sy...
Multiple representations and visual mental imagery in artificial cognitive sy...Multiple representations and visual mental imagery in artificial cognitive sy...
Multiple representations and visual mental imagery in artificial cognitive sy...
 
Multiple representations talk, Middlesex University. February 23, 2018
Multiple representations talk, Middlesex University. February 23, 2018Multiple representations talk, Middlesex University. February 23, 2018
Multiple representations talk, Middlesex University. February 23, 2018
 
Two methods for optimising cognitive model parameters
Two methods for optimising cognitive model parametersTwo methods for optimising cognitive model parameters
Two methods for optimising cognitive model parameters
 
Cardiff map-based orientation talk
Cardiff map-based orientation talkCardiff map-based orientation talk
Cardiff map-based orientation talk
 
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012
Graph comprehension model talk, Birkbeck and Toulouse Le Mirail, February 2012
 
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...
Diagrammatic Cognition: Discovery and Design workshop, Humboldt University, B...
 
A cognitive architecture-based modelling approach to understanding biases in ...
A cognitive architecture-based modelling approach to understanding biases in ...A cognitive architecture-based modelling approach to understanding biases in ...
A cognitive architecture-based modelling approach to understanding biases in ...
 
DECISIVe workshop introduction
DECISIVe workshop introductionDECISIVe workshop introduction
DECISIVe workshop introduction
 

Recently uploaded

Quarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsQuarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsCharlene Llagas
 
User Guide: Magellan MX™ Weather Station
User Guide: Magellan MX™ Weather StationUser Guide: Magellan MX™ Weather Station
User Guide: Magellan MX™ Weather StationColumbia Weather Systems
 
Pests of Sunflower_Binomics_Identification_Dr.UPR
Pests of Sunflower_Binomics_Identification_Dr.UPRPests of Sunflower_Binomics_Identification_Dr.UPR
Pests of Sunflower_Binomics_Identification_Dr.UPRPirithiRaju
 
Thermodynamics ,types of system,formulae ,gibbs free energy .pptx
Thermodynamics ,types of system,formulae ,gibbs free energy .pptxThermodynamics ,types of system,formulae ,gibbs free energy .pptx
Thermodynamics ,types of system,formulae ,gibbs free energy .pptxuniversity
 
Immunoblott technique for protein detection.ppt
Immunoblott technique for protein detection.pptImmunoblott technique for protein detection.ppt
Immunoblott technique for protein detection.pptAmirRaziq1
 
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPRPirithiRaju
 
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》rnrncn29
 
Oxo-Acids of Halogens and their Salts.pptx
Oxo-Acids of Halogens and their Salts.pptxOxo-Acids of Halogens and their Salts.pptx
Oxo-Acids of Halogens and their Salts.pptxfarhanvvdk
 
Q4-Mod-1c-Quiz-Projectile-333344444.pptx
Q4-Mod-1c-Quiz-Projectile-333344444.pptxQ4-Mod-1c-Quiz-Projectile-333344444.pptx
Q4-Mod-1c-Quiz-Projectile-333344444.pptxtuking87
 
Microteaching on terms used in filtration .Pharmaceutical Engineering
Microteaching on terms used in filtration .Pharmaceutical EngineeringMicroteaching on terms used in filtration .Pharmaceutical Engineering
Microteaching on terms used in filtration .Pharmaceutical EngineeringPrajakta Shinde
 
Introduction of Human Body & Structure of cell.pptx
Introduction of Human Body & Structure of cell.pptxIntroduction of Human Body & Structure of cell.pptx
Introduction of Human Body & Structure of cell.pptxMedical College
 
bonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlsbonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlshansessene
 
6.2 Pests of Sesame_Identification_Binomics_Dr.UPR
6.2 Pests of Sesame_Identification_Binomics_Dr.UPR6.2 Pests of Sesame_Identification_Binomics_Dr.UPR
6.2 Pests of Sesame_Identification_Binomics_Dr.UPRPirithiRaju
 
User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)Columbia Weather Systems
 
Servosystem Theory / Cybernetic Theory by Petrovic
Servosystem Theory / Cybernetic Theory by PetrovicServosystem Theory / Cybernetic Theory by Petrovic
Servosystem Theory / Cybernetic Theory by PetrovicAditi Jain
 
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdf
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdfKDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdf
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdfGABYFIORELAMALPARTID1
 
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPirithiRaju
 
Gas-ExchangeS-in-Plants-and-Animals.pptx
Gas-ExchangeS-in-Plants-and-Animals.pptxGas-ExchangeS-in-Plants-and-Animals.pptx
Gas-ExchangeS-in-Plants-and-Animals.pptxGiovaniTrinidad
 
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubai
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In DubaiDubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubai
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubaikojalkojal131
 

Recently uploaded (20)

Quarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsQuarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and Functions
 
User Guide: Magellan MX™ Weather Station
User Guide: Magellan MX™ Weather StationUser Guide: Magellan MX™ Weather Station
User Guide: Magellan MX™ Weather Station
 
Pests of Sunflower_Binomics_Identification_Dr.UPR
Pests of Sunflower_Binomics_Identification_Dr.UPRPests of Sunflower_Binomics_Identification_Dr.UPR
Pests of Sunflower_Binomics_Identification_Dr.UPR
 
AZOTOBACTER AS BIOFERILIZER.PPTX
AZOTOBACTER AS BIOFERILIZER.PPTXAZOTOBACTER AS BIOFERILIZER.PPTX
AZOTOBACTER AS BIOFERILIZER.PPTX
 
Thermodynamics ,types of system,formulae ,gibbs free energy .pptx
Thermodynamics ,types of system,formulae ,gibbs free energy .pptxThermodynamics ,types of system,formulae ,gibbs free energy .pptx
Thermodynamics ,types of system,formulae ,gibbs free energy .pptx
 
Immunoblott technique for protein detection.ppt
Immunoblott technique for protein detection.pptImmunoblott technique for protein detection.ppt
Immunoblott technique for protein detection.ppt
 
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR
6.1 Pests of Groundnut_Binomics_Identification_Dr.UPR
 
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
 
Oxo-Acids of Halogens and their Salts.pptx
Oxo-Acids of Halogens and their Salts.pptxOxo-Acids of Halogens and their Salts.pptx
Oxo-Acids of Halogens and their Salts.pptx
 
Q4-Mod-1c-Quiz-Projectile-333344444.pptx
Q4-Mod-1c-Quiz-Projectile-333344444.pptxQ4-Mod-1c-Quiz-Projectile-333344444.pptx
Q4-Mod-1c-Quiz-Projectile-333344444.pptx
 
Microteaching on terms used in filtration .Pharmaceutical Engineering
Microteaching on terms used in filtration .Pharmaceutical EngineeringMicroteaching on terms used in filtration .Pharmaceutical Engineering
Microteaching on terms used in filtration .Pharmaceutical Engineering
 
Introduction of Human Body & Structure of cell.pptx
Introduction of Human Body & Structure of cell.pptxIntroduction of Human Body & Structure of cell.pptx
Introduction of Human Body & Structure of cell.pptx
 
bonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlsbonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girls
 
6.2 Pests of Sesame_Identification_Binomics_Dr.UPR
6.2 Pests of Sesame_Identification_Binomics_Dr.UPR6.2 Pests of Sesame_Identification_Binomics_Dr.UPR
6.2 Pests of Sesame_Identification_Binomics_Dr.UPR
 
User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)
 
Servosystem Theory / Cybernetic Theory by Petrovic
Servosystem Theory / Cybernetic Theory by PetrovicServosystem Theory / Cybernetic Theory by Petrovic
Servosystem Theory / Cybernetic Theory by Petrovic
 
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdf
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdfKDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdf
KDIGO-2023-CKD-Guideline-Public-Review-Draft_5-July-2023.pdf
 
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
 
Gas-ExchangeS-in-Plants-and-Animals.pptx
Gas-ExchangeS-in-Plants-and-Animals.pptxGas-ExchangeS-in-Plants-and-Animals.pptx
Gas-ExchangeS-in-Plants-and-Animals.pptx
 
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubai
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In DubaiDubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubai
Dubai Calls Girl Lisa O525547819 Lexi Call Girls In Dubai
 

Machine Learning, Artificial General Intelligence, and Robots with Human Minds

  • 1. Machine Learning, Artificial General Intelligence, and Robots with Human Minds David Peebles Reader in Cognitive Science Department of Psychology April 25, 2017
  • 3. Talk outline Structure Discuss different types of artificial intelligence (AI) Human-level artificial general intelligence (AGI) Human cognitive architecture as a basis for AGI Introduce a new cognitive robotics project Key aims Allay your fears of a possible AI invasion and/or enslavement by super-intelligent AI systems in the foreseeable future Enthuse you about the prospect of human-level AI To entertain you with amusing and informative videos
  • 4. Recent advances in artificial intelligence
  • 5. Some recent headlines. . . Artificial intelligence: We’re like children playing with a bomb. Sentient machines are a greater threat to humanity than climate change... (Observer. June, 2016) Will democracy survive big data and artificial intelligence? (Scientific American. Feb, 2017) The AI threat isn’t SkyNet, it’s the end of the middle class (Wired Business. Feb, 2017) Stephen Hawking, Elon Musk, and Bill Gates warn about artificial intelligence (Observer. Aug, 2015) DESTROYING OURSELVES: Artificial intelligence poses THREAT to humanity – shock report (Express. Jan, 2017)
  • 6. Are intelligent machines really upon us? It depends what you mean by “machine intelligence” Three types of AI: Artificial Narrow Intelligence (ANI). Machine intelligence that equals or exceeds human intelligence or efficiency, but only in one specific area. Artificial General Intelligence (AGI). Machines with general human intelligence, capable of sustaining long-term goals and intent, which can successfully perform any task that a human might do. Artificial Super-intelligence (ASI). Machine intelligence that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills (Bostrom, 2014).
  • 7. The importance of distinguishing AI The three types of AI are often conflated or confused in the press and public mind. All of the recent developments are in Narrow AI due to: Faster hardware (GPUs) Huge amounts of data New statistical machine learning methods (e.g., deep learning in neural networks) Main machine learning methods Supervised learning (classification and regression) Unsupervised learning (clustering) Reinforcement learning (optimisation via reward signal)
  • 8. Narrow AI Benefits Rapid conversion of unstructured information into useful and actionable knowledge. Narrow AI will provide tools to automate and optimise this process in science, medicine, business etc. Costs Elimination of jobs (mainly white-collar knowledge work). Concentrations of knowledge, power and resources may exacerbate inequality. ‘Black box’ systems where humans don’t know why AI systems are making decisions Limitations Require extensive training Each system is specific to the data set it’s trained on. One slight change requires retraining
  • 9. Artificial General Intelligence “Machines with general human intelligence, capable of sustaining long-term goals and intent, which can successfully perform any task that a human might do”. More interesting for cognitive science Much more complex and harder to achieve than narrow AI Some recent examples from Hollywood. . . Ex Machina (2014) Robot and Frank (2012) Prometheus (2012)
  • 10. Criteria for advanced autonomous agents US NSF $16.5M Smart and Autonomous Systems (2016) Specified five desired capabilities: Cognisant. Exhibit high-level cognition and awareness beyond primitive actions. Taskable. Users can specify desired behaviours and outcomes in a natural and concise (possibly vague) manner Reflective. Can adapt behaviour and learn new behaviour from experience, observation, and interaction with others. Ethical. Adhere to a system of societal and legal rules. Do not violate accepted ethical norms. Knowledge-rich. Semantic, probabilistic and commonsense reasoning and meta-reasoning about uncertain, dynamic environments. Able to acquire and understand knowledge.
  • 11. Advanced autonomous agents Long-term autonomy (able to operate unaided over time) Flexible assistants and agents in applications such as manufacturing, military, agriculture, health, space etc. Robots that can be trained for new task using natural language rather than having to be re-programmed. Robots that can perform complex task planning without human intervention NASA Mars Exploration Rover NASA Valkyrie robot
  • 12. A key human skill – knowledge transfer Excel spreadsheet SPSS spreadsheet
  • 13. Cognitive architectures Originated in 1950s but active research programme in 1980s. Cognitive science – differs from mainstream “narrow” AI and traditional “divide and conquer” approach of experimental cognitive psychology. Theories of the core, immutable structures and processes of the human cognitive system. Integration of multiple components of cognition. Computer architecture Perception Action Task Environment Learning Procedural Perceptual Learning Learning Declarative Selection Action Short−term memory Long−term memory Procedural Long−term memory Declarative Human cognitive architecture
  • 14. Example 1 – Interactive task learning Rosie (Univ. Michigan). Soar agent able to learn new tasks specified in natural language and interactive demonstration. Complex skill integrating natural language processing, vision, logical reasoning, search and motor control. V1: Learns several different games and puzzles through natural language descriptions of the legal actions and goals of the task. V2. Learns new goals from single examples, either by being shown a visual example of the goal or from executing a series of actions to achieve the goal. Agent’s hypotheses are refined and corrected by language input from human instructor. Video: Simple task learning Video: Tower of Hanoi puzzle
  • 15. Example 2 – Mind reading and mental simulation Cognitive architecture installed on a MDS (Mobile / Dexterous / Social) robot Aim: AA that can construct accurate mental representations of other agents’ mental states and perspectives in order to reason about their beliefs and intentions. In order to: Coordinate actions with teammates effectively Anticipate teammates’ (potentially ambiguous) actions Understand shared team goals of the team Communicate more naturally. Current tasks include gaze following, hide and seek, task interruption and resumption, theory of mind. Agent creates mental model of teammates’ beliefs and (explicitly stated) goals and predicts what they will do next. Video: Gaze and gesture interpretation
  • 16. The ‘robot with a human mind’ project NAO robot Hadeel Jazzaa Lee McCluskey
  • 17. A flexible research robot Video: NAO robot Video: Robot football
  • 18. The ‘robot with a human mind’ project NAO robot NAOqi architecture Visual Module ACT−R Buffers Environment Pattern Matching Execution Production Module Motor Problem State Declarative Memory Procedural Memory Control State ACT-R cognitive architecture
  • 19. Aims of the project Embodied cognition constrained by perceptuo-motor processing. Goal directed behaviour. Including self generated goals. Cognitive control and resource management over different time scales (including long-term). High-level cognition. Planning, reasoning, problem solving, decision making, reflection and metacognition, analogy. Verbal and nonverbal communication. Interaction and adaptation through natural language understanding (speech, text, demonstration. Constant adaptation and learning. Human scale (i.e., relatively fast and not requiring thousands of training trials). Integration of cognitive functions
  • 20. Take home messages Types of AI Narrow AI is going to change the world radically but is limited AGI is a much more challenging problem and is a long way off Super-intelligent AI is a very long way off AGI: Requires the integration of many aspects of intelligence (learning, reasoning, planning, self-monitoring etc.) Will require insights from cognitive science as human cognition is currently our best example of AGI Cognitive architectures are a potentially very useful approach to developing systems with AGI