AI beyond chatbots in 2026: Stanford Computer Science’s Jure Leskovec - part 2/2
2025-12-29_12-14 • 5m 39s
Jure Leskovec (Computer Science Professor)
00:00.480
We
are
extremely
excited
about
uh
the
future.
Because
if
you
think
about
it,
these
AI
models
have
all
been
trained
on
generic
data.
But
what
is
going
to
make
AI
really
powerful
and
have
the
impact
is
to
be
able
to
harness
the
data
that
is
unique
to
every
individual
company.
And
Jure Leskovec (Computer Science Professor)
00:19.240
only
this
way
individual
businesses
are
going
to
be
able
to
to
to
find
kind
of
their
own
unique
strength
and
their
own
unique
way
of
using
AI
that
that
differentiates
them
from
everyone
else
in
the
market.
And
what
my
company
is
focusing
on
is
that
a
lot
of
this
data
is
now
Jure Leskovec (Computer Science Professor)
00:35.920
today
locked
in
the
silos.
A
lot
of
this
data
actually
does
not
come
in
the
form
of
documents
and
text,
but
it
comes
in
a
form
of
transaction
records,
tabular
data,
data
where
house
data.
So,
what
my
company
focuses
on
is
building
specialized
AI
that
can
reason
over
structured
Jure Leskovec (Computer Science Professor)
00:52.700
business
data,
business
records,
supply
chain
records,
transaction
records,
things
like
that.
Autoscroll