What Does Watson 2.0 Tell us About the Philosophy, Theory ......What Does Watson 2.0 Tell us About the Philosophy, Theory, & Future of AI? Selmer Bringsjord(1) • Naveen Sundar G.(2)

What Does Watson 2.0 Tell us About the Philosophy, Theory, & Future of AI?

Selmer Bringsjord(1) • Naveen Sundar G.(2)

Rensselaer AI & Reasoning (RAIR) Lab(1,2)

Department of Cognitive Science(1)

Department of Computer Science(1,2)

Lally School of Management & Technology(1)

Rensselaer Polytechnic Institute (RPI)Troy, New York 12180 USA

Oxford UniversityOxford, UK9/21/2013

Tuesday, October 1, 13




















Made possible by the generosity of:

IBMRPI’s pursuit of ITS/IILE

NSF

JTF

AFOSR

ONR x 2

DARPA











Made possible by the generosity of:

IBMRPI’s pursuit of ITS/IILE

NSF

JTF

AFOSR

ONR x 2

DARPA


Simon: “Chess is a lofty target for AI, yes. Go for it!”



“Chess is Too Easy”


“Chess is Too Easy”

I was right :).


What about this, then?





Experts to IBM: “Can’t be done!”




No one asked me.




No one asked me.






u 2 ⌃⇤


u 2 ⌃⇤


u 2 ⌃⇤U = (S, . . .)


u 2 ⌃⇤

U : u �! �U = (S, . . .)


u 2 ⌃⇤

U : u �! �U = (S, . . .)


u 2 ⌃⇤

U : u �! �U = (S, . . .)

A(v1 @ u,R) Â(v2 @ u,R)


u 2 ⌃⇤

U : u �! �U = (S, . . .)

A(v1 @ u,R) Â(v2 @ u,R)(Ab(u) ^ u 2 MedBase) ! t(u) = ‘skin cancer’


u 2 ⌃⇤

U : u �! �U = (S, . . .)

A(v1 @ u,R) Â(v2 @ u,R)(Ab(u) ^ u 2 MedBase) ! t(u) = ‘skin cancer’

$


Many thanks to A Bringsjord for this pair of slides, and discussion.




Autonomy “Magic”


Autonomy “Magic”Understanding 100% of InformationThe characteristic that makes human information unstructured is its form — it does not fit neatly into the rows and columns of a database, but exists in various formats including books, email messages, surveillance video, chat streams, and phone calls that occur across networks, the web, the cloud, and numerous mobile devices. Growing at a rate three times that of structured data, the increasing deluge of unstructured information makes up approximately 90 percent of all information. The challenge for the modern enterprise is to understand and extract value from this rich sea of human information.

Today, with Autonomy, organizations can now process and understand in real time, the meaning of 100 percent of information.







!?!?




!?!?(They sure as heck never consulted with Floridi.)





The little OUP book that could’ve saved HP $6 billion:





The little OUP book that could’ve saved HP $6 billion:


Let’s in fact talk about 100% of semantic information, using Watson 2.0 as a springboard ...



Logic


Logic


MiniMax

ularit

y

Logic


MiniMax

ularit

y

Logic

propo

stional

logic


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL

UIMA outp

ut


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

UIMA outp

ut


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

UIMA outp

ut


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

Art of Infallibility 1

UIMA outp

ut


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...


UIMA outp

ut

FOL


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...


UIMA outp

ut

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...


UIMA outp

utepistemic

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal


UIMA outp

utepistemic

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


UIMA outp

utepistemic

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic

temporal+epistemic+deontic


UIMA outp

utepistemic

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic



UIMA outp

ut

+planning+arg semantics

epistemic

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic



UIMA outp

ut


epistemic

...

FOLSOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic



UIMA outp

ut


epistemic

...

FOL

heterogeneous/visual

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic



UIMA outp

ut


epistemic

...

FOL


SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic



UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Vivid


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

DCEC ⇤

Deontic Cognitive Event Calculus


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

DCEC ⇤


1. natural language semantics (non-Montagovian)2. higher-cognition tests (for Psychometric AI) (false-belief test, deliberative mind-reading mirror test for self-consciousness ...)

4. biz & econ simulation 3. ethically correct robots


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

DCEC ⇤



MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Gödel’s “God Theorem”


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

ITS (Culture, Language, Math)


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

AI-ified Axiomatic Physics!(Synthese)


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...

temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Goodstein’s Theorem!


Licato, J.; Govindarajulu, N.; Bringsjord, S.; Pomeranz, M.; Gittelson, L. 2013. Analogico-Deductive Generation of Godel’s First Incompleteness Theorem from the Liar Paradox. In Proceedings of IJCAI 2013. Pdf

Govindarajulu, N.; Licato, J.; Bringsjord, S. 2013. Small Steps Toward Hypercomputation via Infinitary Machine Proof Verification and Proof Generation. In Proceedings of UCNC 2013. Pdf


http://kryten.mm.rpi.edu/ADR_2_GTheorem_from_LP.pdf

http://kryten.mm.rpi.edu/ADR_2_GTheorem_from_LP.pdf

http://link.springer.com/chapter/10.1007/978-3-642-39074-6_11

http://link.springer.com/chapter/10.1007/978-3-642-39074-6_11

Liar Paradox (L)• Collection of semiformal statements• Syntax not rigorously defined• Represents intuitive understanding of problem

domain

Ending one of AI’s “bad dreams”:

s = “This statement is a lie”There is a statement that is neither

true nor false....

G1• Completely formal statements• Syntax very rigorously defined• Purely mathematical objects: numbers, formal

theories, etc.

s = ?

∃φ∈LA ¬(PA |⎯ φ) ∧ ¬(PA |⎯ ¬φ)...



domain



true nor false....


theories, etc.

s = ?

∃φ∈LA ¬(PA |⎯ φ) ∧ ¬(PA |⎯ ¬φ)...



domain



true nor false....


theories, etc.

s = ?

∃φ∈LA ¬(PA |⎯ φ) ∧ ¬(PA |⎯ ¬φ)...

S• Semiformal statements (but more formal than L)• Syntax somewhat rigorously defined• Somewhat intuitive; deals with stories of reasoners

and utterances made by inhabitants of an island

s = ¬Bs

∃p ¬Bp ∧ ¬B¬p...



domain



true nor false....


theories, etc.

s = ?

∃φ∈LA ¬(PA |⎯ φ) ∧ ¬(PA |⎯ ¬φ)...

S• Semiformal statements (but more formal than L)• Syntax somewhat rigorously defined• Somewhat intuitive; deals with stories of reasoners

and utterances made by inhabitants of an island

s = ¬Bs

∃p ¬Bp ∧ ¬B¬p...


The Singularity Approaches ...





(Good 1965; Chalmers 2010)



If AI is only Ray-1 AI ...



False!






False!



False!False!



False!

And of course ...

False!



... if AI is Ray-3 AI ...


... if AI is Ray-3 AI ...

False!False!


“My, that’s rather negative, even violent. Can we talk about positive views, please?”


SupermindsPeople Harness Hypercomputation,

and More(2003)

Turing Limit

Information Processing

Subjective consciousness, qualia, etc. — phenomena in the incorporeal realm that can’t be expressed in any third-person scheme

persons

animals (chess, go, swimming, flying, locomotion)

Hypercomputation

People Harness Hypercomputation, and More

29

by

SUPERMINDSPeople Harness Hypercomputation, and More

bySelmer Bringsjord and Micael Zenzen

This is the first book-length presentation and defense of a new theory of human andmachine cognition, according to which human persons are superminds. Superminds arecapable of processing information not only at and below the level of Turing machines(standard computers), but above that level (the “Turing Limit”), as information processingdevices that have not yet been (and perhaps can never be) built, but have beenmathematically specified; these devices are known as super-Turing machines orhypercomputers. Superminds, as explained herein, also have properties no machine,whether above or below the Turing Limit, can have. The present book is the third andpivotal volume in Bringsjord’s supermind quartet; the first two books were What RobotsCan and Can’t Be (Kluwer) and AI and Literary Creativity (Lawrence Erlbaum). The finalchapter of this book offers eight prescriptions for the concrete practice of AI and cognitivescience in light of the fact that we are superminds.

SELMER

BR

ING

SJOR

D

AN

D M

ICH

AEL ZEN

ZENSU

PERM

IND

SPeople H

arness Hypercom

putation, and More

KLUWER ACADEMIC PUBLISHERS COGS 29

Bringsjord COGS 29 PB(2)xpr 07-02-2003 16:26 Pagina 1



and More(2003)

Turing Limit



persons


Hypercomputation


29

by




SELMER

BR

ING

SJOR

D

AN

D M

ICH

AEL ZEN

ZENSU

PERM

IND

SPeople H

arness Hypercom

putation, and More



!u!v["kH(n, k, u, v) # "k!H(m, k!, u, v)](programming)



and More(2003)

Turing Limit



persons


Hypercomputation


29

by




SELMER

BR

ING

SJOR

D

AN

D M

ICH

AEL ZEN

ZENSU

PERM

IND

SPeople H

arness Hypercom

putation, and More






and More(2003)

Turing Limit



persons


Hypercomputation


29

by




SELMER

BR

ING

SJOR

D

AN

D M

ICH

AEL ZEN

ZENSU

PERM

IND

SPeople H

arness Hypercom

putation, and More





MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

What is AI for you?


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

What is AI for you?


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Elevated AI only!:

“The ultimate goal of AI is to build a person, or more humbly, an animal.” —C&M

What is AI for you?


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Darwininan “Canine” AI


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

“Monkey” AI


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

“Full-Watson” AI


MiniMax

ularit

y

Logic

propo

stional

logic

semant

ic-web

logic

s

descr

iption l

ogics

fragm

ents

of FOL ...temporal

temporal+epistemic


Lw1,w


UIMA outp

ut


epistemic

...

FOL


Infinitary (AoI 2)

SOL

...

Person-Aspiring AI


What Does Watson 2.0 Tell us About the Philosophy, Theory ......What Does Watson 2.0 Tell us About the Philosophy, Theory, & Future of AI? Selmer Bringsjord(1) • Naveen Sundar G.(2)

Documents