2001-Dreams and reality

Hans-Georg.Stork@cec.eu.int
2001 - dreams and reality
Weaving the Web ... of the future
… for a safe voyage in knowledge space

Today’s dreams are not new:
HAL
… recognizes who you are
… understands what you say
… surmises what you feel
HAL knows !
… talks to you

The inventors of HAL
ignored (e.g.):
But 2001 reality is quite
different from 1968
dreams ...
Vannevar Bush’s MEMEX (1945)
Ted Nelson’s XANADU (1960)
Moore’s Law (1965)

Yet, some of the old
dreams are in fact on the
verge of coming true…
(well, to some extent …)
But the
dimensions
are different !

Weaving the Web … of the future
They can be big, but
most are small; they are
everywhere and they are
interconnected.
Computers are not huge
isolated supermachines.
They are in business !

Computers communicate ...
… but do they “understand”
what they tell each other ?
They don’t, unless ...

… they are told
… or told to learn
… the meaning of what
they are talking about.
To unleash the full potential
of networked resources
these resources have to
have clearly defined
SEMANTICS.

Mobility
gateway
XML
RDF
OIL
CC/PP
DAML-O
MPEG
SMIL
Formal languages are the
stepping stones to the
‘Semantic Web’
Tim Berners-Lee, 2000

Conjecture:
In an adaptive and open,
collaborative and automated,
multimedia and pervasive,
trustable Semantic Web …
populated by content and
context “aware” systems and
appliances ...
content semantics must be
grounded in well defined
reference ontologies
Another view of OIL, the Ontology Inference Layer

Semantics:
… the meaning of things
… to be aware of the
meaning of things ...
to know ...
… the ‘Semantic Web’ ...
… a Web of
Knowledge ??

But what kind of knowledge ?
E = mc2 ?
eip
+ 1 = 0 ?
1066 - Battle of Hastings ?
Client X prefers product Y ?
… a rose by any other name
would smell as sweet …

If we wanted to bring a HAL-like
being into existence, what are
the millions of things that
should be used to prime HAL's
knowledge pump?
How should they be represented
inside the machine so that it
can use them efficiently to
deduce further conclusions
when needed, just as we would?
Who will do the actual entering
of all that data?
Douglas B. Lenat

An Allegory with Venus and Cupid
probably 1540-50
BRONZINO ( 1503 - 1572)
C
EN
SO
R
ED
?
The challenge of knowledge
extraction from multimedia
documents:
It should indeed be done
automatically and with great
care.
Machines must learn to make
the difference.
HAL technologies (and more,
of course) may help !
(HAL = Heuristic Algorithmic Learner)

But remember Isaac Asimov’s robot
commandments (and the fate of
USSS Discovery):
• a robot must not injure a human
being;
• a robot must obey orders of a
human being, but not one that
would violate the first principle;
The ultimate challenge (so far !):
to have machines ground themselves
in the real world and have them share
with us, via a Semantic Web, the
knowledge they acquire.
…and the Joy-Kurzweil controversy which may not be so off the wall after all ...

2001-Dreams and reality

More Related Content

Similar to 2001-Dreams and reality (20)

2001-Dreams and reality