Artificial Intelligence (AI)Jim Spohrer (IBM)
Seoul, South Korea; October 13, 2016
Consulting Conferencehttp://www.slideshare.net/spohrer/korea-day1-keynote-20161013-v6
10/4/2016 Understanding Cognitive Systems 1
10/4/2016 Understanding Cognitive Systems 2
10/4/2016 Understanding Cognitive Systems 3
Today’s Talk
• Abstract: Artificial Intelligence (AI)• IBM is transforming into a cognitive solutions and cloud platform
company. What opportunities and challenges? What was IBM’s response to a recent White House RFI about Preparing for the Future of Artificial Intelligence?
• Bio:• Jim Spohrer, IBM, Director Understanding Cognitive Systems
• Former Director, University; Service Research; CTO IBM VC Group
• Jim is developing a next generation curriculum to help learners build, understand, and work with cognitive systems. Education: Yale Computer Science (AI&CogSci) PhD, MIT Physics BS.
10/4/2016 Understanding Cognitive Systems 4
Jim SpohrerIBM
Today’s Talk
• Four Cool Tech Topics • Today: AI• If time others
• Definitions• Perspectives
• Industry• IBM
• Relationships• Industry 4.0/IoT
• Challenge & Opportunity• Too hard to build, understand,
and work with AI systems
10/4/2016© IBM UPWard 2016
5
Definition: Intelligence
• Intelligence has been defined in many different ways including as one's capacity for logic, understanding, self-awareness, learning, emotional knowledge, planning, creativity and problem solving. It can be more generally described as the ability to perceive information, and retain it as knowledge to be applied towards adaptive behaviors within an environment or context.
10/4/2016© IBM UPWard 2016
6
Definitions: AI vs IA
10/4/2016© IBM UPWard 2016
7
AI is Artificial Intelligence, orintelligence in machines (smart machines)
IA is Intelligence Augmentation, orpeople thinking and working together with smart machines.
IA is what IBM calls “Cognitive Computing” andthe smart machines are called “Watson Solutions” or
more generally “Digital Cognitive Systems (Cogs)”
Definitions: Types of Cognitive System Entities(symbol and pattern processing systems)
• Socio-Technical (Organization-based)• Businesses• Cities• Nations
• Biological (Brain-based)• People• Animals
• Technological (Computation-based)• Embodied (Robot, Car, Device)• Virtual (Local, Cloud)
10/4/2016 Understanding Cognitive Systems 8
Industry Perspective
10/4/2016 Understanding Cognitive Systems 9
10/4/2016 Understanding Cognitive Systems 10
10/4/2016 Understanding Cognitive Systems 11
10/4/2016 Understanding Cognitive Systems 12
10/4/2016 Understanding Cognitive Systems 13
10/4/2016 Understanding Cognitive Systems 14
IBM Perspective
10/4/2016 Future of AI 15
Request:White House OSTP • The White House Office of Science and Technology Policy is particularly
interested in responses related to the following topics:
• (1) the legal and governance implications of AI;
• (2) the use of AI for public good;
• (3) the safety and control issues for AI;
• (4) the social and economic implications of AI;
• (5) the most pressing, fundamental questions in AI research, common to most or all scientific fields;
• (6) the most important research gaps in AI that must be addressed to advance this field and benefit the public;
• (7) the scientific and technical training that will be needed to take advantage of harnessing the potential of AI technology;
• (8) the specific steps that could be taken by the federal government, research institutes, universities, and philanthropies to encourage multi-disciplinary AI research; and
• (9) the use of open data sets to close fundamental research gaps;
• (10) the role of incentives and prizes to accelerate public benefits;
• (11) any additional information related to AI research or policymaking, not requested above, that you believe OSTP should consider.
10/4/2016 Future of AI 16
A. The Use of AI for the Public Good
• Healthcare
• Social Services
• Education
• Financial Services
• Transportation
• Public Safety
• The Environment
• Infrastructure
10/4/2016 Future of AI 17
B. Social and economic implications of AI • History: Technologies with broad
potential across industries like AI increase:• Productivity• Earnings• Job growth
• Adjustment: Social, learning, and decision making capabilities are key to acceptance of AI in society
• Underserved: Potential to help underserved populations and improve their quality of life
10/4/2016 Future of AI 18
C. Education for harnessing AI technologies• Demand: High demand for AI,
machine learning, data science and related courses; many MOOCs
• Industry Platforms: Growing set of companies provide learners access to cognitive service capabilities in their clouds
• Gap: Need curriculum for learners with no programming or advanced mathematics background
10/4/2016 Future of AI 19
D. Fundamental questions in AI research, and the most important research gaps
• Machine learning and reasoning
• Decision techniques
• Domain-specific AI systems
• Data assurance and trust
• Radically efficient computing infrastructure
10/4/2016 Future of AI 20
E. Data sets that can accelerate AI research• Bottleneck: Develop and validate
data sets that are:• Large and unbiased
• Openly curated
• Publically accessible
• Domains: Novice to expert task performance is hard to get for all occupations and industries
• Models: Incentives to share trained models that require lots of data and compute time to create
10/4/2016 Future of AI 21
F. Multi-disciplinaryresearch• Disciplines: Breadth of disciplines
to tackle issues:• Psychology and cognitive science,
philosophy, design and art, public policy and management, law and regulations
• Systems: Professional associations to tackle industry and system issues, including novice to expert progression on tasks
• Socio-technical system design loop and smart service systems
10/4/2016 Future of AI 22
G. Role of incentives and prizes• Example:
• IBM Watson AI XPrize ($5M)
• Best AI system to empower teams of people to tackle the world’s grand challenges
• TED 2020, finalists present
• I-athlon• More objective scoring
• Rational processes
• Multiple dimensions of intelligence
10/4/2016 Future of AI 23
H. Safety and control issues for AI • Trust & Trustworthiness
• Ethical & Social Norms
• Algorithmic Transparency
• Unexpected Interactions
• Safeguards
10/4/2016 Future of AI 24
Partnership for AI formed
10/4/2016 Understanding Cognitive Systems 25
I. Legal and governance implications of AI • Responsible & inclusive dialogue
• Elevate the dialogue• Algorithmic responsibility
• Individual privacy
• Jobs and workforce transformation
• Safety
• Learn beyond the headlines
• Key: Focus on skills
10/4/2016 Future of AI 26
J. Other issues: Business models • Trusted platform
• Training – Unbiased data• Algorithms - Transparency• Services - Open API Economy• Transactions – Blockchain• Applications – Ethically boosting
creativity and productivity• Governance – Laws and regulations
• Trust takes time• Steam engines – boiler explosions• Cognitive engines – headline hype
10/4/2016 Future of AI 27
Some reactions…
• aitrends• Artificial Brilliance• Futurism• InformationWeek
• Calburn: “IBM: AI Should Stand For ‘Augmented Intelligence’
• PYMNTS• TechCrunch
• Coldewey: “The White House requested input on artificial intelligence, and IBM’s response is a great AI 101”
10/4/2016 Future of AI 28
White House OSTPFollow Up• The White House OSTP received
161 responses and created a 349 downloadable PDF with URL links
• The White House Frontiers Conference is being planned as a follow up meeting• Date: October 13, 2016
• Place: Pittsburgh, PA USA
10/4/2016 Future of AI 29
30
What types of digital cognitive systems?
• Cognitive Build: Outthink Challenge (250K people)• Imagine a digital cognitive system to help you do
something important in your personal or professional lives
• Team to design it and advocate for it, and then everyone votes
• Winners: reduce waste and human suffering, screen for health issues and safety threats, learn life skills and make better choices, find what you are looking for, move around more effectively, provide emotional support, provide IT support, learn about important public policy goals and make better choices
• Types: Tool, Assistant, Collaborator, Coach, Mediator
10/4/2016 Understanding Cognitive Systems 31
Types
• Tool
• Assistant
• Collaborator
• Coach
• Mediator
10/4/2016 Understanding Cognitive Systems 32
Types: Progression of models and capabilities
10/4/2016 Understanding Cognitive Systems 33
Task & World Model/Planning & Decisions
Self Model/Capacity & Limits
User Model/Episodic Memory
Institutions Model/Trust & Social Acts
Tool + - - -
Assistant ++ + - -
Collaborator +++ ++ + -
Coach ++++ +++ ++ +
Mediator +++++ ++++ +++ ++
tool assistant collaborator coach mediator
Build: 10 million minutes of experience
10/4/2016 Understanding Cognitive Systems 34
Build: 2 million minutes of experience
10/4/2016 Understanding Cognitive Systems 35
Build: Hardware < Software < Data < Experience < Transformation
10/4/2016 Understanding Cognitive Systems 36
Understand them…
10/4/2016 Understanding Cognitive Systems 37
Work with…
10/4/2016 Understanding Cognitive Systems 38
Next generation cognitive curriculum
10/4/2016 Understanding Cognitive Systems 39
IBM Cloud Bluemix: Watson APIs are growing…
10/4/2016© IBM UPWard 2016
40
So far (June 2016), 100,000 faculty and students globally given access
10/4/2016 Understanding Cognitive Systems 41
10/4/2016 Understanding Cognitive Systems 42
10/4/2016 Understanding Cognitive Systems 43
IBM Cloud Bluemix: Watson APIs are growing…
10/4/2016© IBM UPWard 2016
44
So far (June 2016), 100,000 faculty and students globally given access
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 45
I have…
Have you noticed how the building blocks just keep getting better?
Learning to program:My first program
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 46
Early Computer Science Class:Watson Center at Columbia 1945
Jim Spohrer’sFirst Program 1972
Brief History of AI
• 1956 – Dartmouth Conference
• 1956 – 1981 Micro-Worlds
• 1981 – Japanese 5th Generation
• 1988 – Expert Systems Peak
• 1990 – AI Winter
• 1997 – Deep Blue
• 1997 – 2011 Real-World
• 2011 – Jeopardy! & SIRI
• 2013 – Cognitive Systems Institute
• 2014 – Watson Business Unit &
• True North Brain Chip
• 2015 – “Cognition as a Service”on IBM Bluemix
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development47
10/4/2016 48
1955 1975 1995 2015 2035 2055
Can better service help us be wiser?
Cognitive Mediator (2035): Tool, Assistant, Collaborator, Coach
Computing: Then, Now, Projected
10/4/2016 49
2035
2055
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 50
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 51
10/4/2016© IBM UPWard 2016
52
10/4/2016© IBM UPWard 2016
53
What might Reality 2.0 look like?
What exists in 2016?
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 54
360,000 100,000 120,000 60,000 150,000
How fast is Artificial Intelligence approaching?
10/4/2016 55
What might it look like?
Jim Spohrer (IBM)Seoul, South Korea; October 13, 2016
Consulting Conferencehttp://www.slideshare.net/spohrer/korea-day1-keynote-20161013-v6
10/4/2016 56
Come visit IBM Research – Almaden in San Jose, CA USA – monthly university day!
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 57
Dedication: Douglas C. EngelbartFather of the mouse and augmentation theory
10/4/2016© IBM 2015, IBM Upward University Programs Worldwide
accelerating regional development 58
But this stuff is still really hard…
10/4/2016© IBM UPWard 2016
59