When
I
was
a
young
lawyer,
I
spent
hours
in
dark
dusty
warehouses
paging
through
documents,
looking
for
something
relevant
and
important.
At
the
time,
I
didn’t
think
it
did
much
for
me
and
I
still
don’t.
It
was
my
belief
then
and
now
that
much
of
the
mundane
work
young
lawyers
were
expected
to
do
really
didn’t
train
them
to
do
much
of
anything
other
than
maximize
billable
hours.
Moreover,
having
to
do
that
kind
of
work
was
depressing
and
lowered
worker
satisfaction.
That’s
why
I’ve
been
a
vocal
advocate
for
the
theory
that
AI’s
takeover
of
mundane
legal
work
won’t
necessarily
harm
junior
lawyer
training.
So,
while
future
on-the-job
training
may
be
different,
it
doesn’t
have
to
mean
younger
lawyers
wouldn’t
ultimately
just
as
good
if
not
better
than
today’s
lawyers
for
a
whole
lot
of
reasons.
Being
passionate
and
defending
a
theory
can
often
be
a
good
thing.
However,
it’s
sometimes
good
to
question
that
theory
every
now
and
then
just
to
make
sure
there
isn’t
some
nuance
that
you
need
to
be
aware
of.
It’s
time
to
at
least
question
my
own
assumptions.
It’s
always
good
to
recognize
that
things
are
never
as
simple
as
your
beliefs
make
it
seem.
Gut
Instinct
and
Science
Recent
research
by
Daniel
Kahneman
and
Gary
Klein
in
a
substantive
work
entitled
The
Foundational
Kahneman-Klein
Study
“Conditions
for
Intuitive
Expertise:
A
Failure
to
Disagree
looks
at
the
unconscious
recognition
of
patterns
a
person
has
previously
seen
and
experienced
when
presented
with
a
problem.
These
patterns
are
then
applied
to
the
situation
presented
to
find
solutions
to
a
problem.
This
theory
may
have
significant
implications
for
how
we
think
about
legal
training,
particularly
the
grunt
work
AI
is
replacing
This
process
leads
you,
for
example,
to
“sense”
something
is
wrong,
although
it’s
really
not
sensing
at
all.
The
theory
is
called
Recognition-Primed
Decision
Making
(RPD).
When
I
was
more
experienced,
for
example,
I
usually
could
predict
when
the
other
side
would
approach
settlement
because
I
had
been
through
enough
cases
to
see
patterns
of
behavior
in
the
other
side.
Experienced
lawyers
though
unknowingly
call
this
their
gut
instinct
when
making
decisions
and
recommendations.
If
RPD
is
valid,
then
it
stands
to
reason
that
the
more
patterns
you
see,
the
more
ability
you
would
have
to
apply
those
patterns
to
future
situations
and
the
better
your
so-called
gut
instinct
will
be.
Pattern
Recognition
in
Legal
Training
Let’s
take
document
review
and
research,
the
bane
of
many
young
lawyers’
existence.
The
argument
can
be
made
that
document
review,
the
sifting
through
contracts
or
discovery
materials,
due
diligence,
simple
legal
research,
or
drafting
routine
motions
exposes
young
lawyers
to
recurring
patterns:
common
clauses,
the
patterns
and
links
between
documents
that
you
see
as
you
review
typical
legal
risks,
the
pattern
of
judicial
reasoning
that
leads
to
conclusions,
the
patterns
of
judicial
reasoning
that
enables
prediction.
This
repetitive
work
exposed
lawyers
to
thousands
of
variations
of
legal
problems.
For
example,
a
junior
lawyer
reviewing
hundreds
of
contracts
might
start
noticing
red
flags
like
ambiguous
indemnity
clauses
or
risky
termination
provisions.
Once
I
was
exposed
to
a
situation
involving
a
potential
ambiguous
noncompete
clause,
I
could
use
what
I
had
seen
in
similar
cases
so
that
I
recognized
the
clause
in
question
could
lead
to
trouble.
Over
time,
this
repetition
could
embed
a
mental
database
of
“what
looks
wrong”
or
“what
feels
right,”
which
manifests
as
gut
instinct.
Of
course,
real
life
is
more
nuanced
than
that.
We
can’t
just
say
that
the
old
ways
that
younger
lawyers
gained
experience
meant
they
were
necessarily
better
lawyers
than
those
who
come
up
in
age
of
AI.
Merely
reviewing
endless
emails
with
little
legal
significance
without
more
might
not
sharpen
instincts
as
much
as
analyzing
a
few
pivotal
documents
that
AI
can
enable.
Simple
Repetition
Is
Not
Enough
Tools
like
legal
research
platforms
or
predictive
analytics
can
expose
lawyers
to
more
patterns,
faster,
by
surfacing
relevant
cases,
trends,
or
outcomes
that
would
take
years
to
encounter
manually.
Moreover,
you
don’t
hone
contract
negotiation
skills
by
reviewing
a
bunch
of
contracts.
You
don’t
learn
how
to
create
a
successful
litigation
strategy
by
doing
routine
motions
over
and
over.
It
takes
more.
Of
course,
there
are
studies,
particularly
Ericsson’s
work
that
suggest
that
repeated
exposure
to
specific
tasks
can
build
some
sort
of
intuitive
judgment.
But
Ericcson’s
studies
on
expertise
show
that
effective
skill-building—real
expertise—
requires
more
than
just
repetition.
It
requires
what
he
calls
“deliberate
practice.”
Deliberate
practice
refers
to
“a
special
type
of
practice
that
is
purposeful
and
systematic.
While
regular
practice
might
include
mindless
repetitions,
deliberate
practice
requires
focused
attention
and
is
conducted
with
the
specific
goal
of
improving
performance.”
A
junior
lawyer
using
AI
might
develop
a
richer
“database”
of
patterns
than
one
slogging
through
manual
document
review.
Indeed,
AI
may
eliminate
or
at
least
reduce
the
need
for
the
gut
instinct.
Data
analysis
may
replace
the
gut
instinct
with
fact-based
data.
Indeed,
those
who
subscribe
to
my
theory
about
the
impact
of
AI
on
younger
lawyers
typically
pooh-pooh
the
whole
notion
of
a
gut
instinct,
viewing
it
as
little
more
than
wild
ass
guesses
versus
reliance
on
data
and
facts.
(Interestingly,
even
if
you
subscribe
to
RPD,
humans
are
basically
doing
just
what
AI
and
data
analytics
do
inside
their
brains:
they
are
looking
at
past
experaince
(data)
and
looking
for
patterns.
AI
can
just
do
it
thousands
of
times
faster
and
better.)
And
what
we
call
gut
instinct
doesn’t
solely
stem
from
pattern
recognition
that
somehow
develop
through
repetitive
tasks.
It
involves
a
broader
mix
of
cognitive
and
emotional
skills
like
empathy,
or
ethical
judgment.
A
lawyer’s
ability
to
read
a
client’s
emotional
state
or
anticipate
a
judge’s
reaction,
for
example,
might
rely
on
interpersonal
skills
or
situational
awareness,
not
just
patterns
from
grunt
work.
These
softer
skills,
developed
through
mentorship,
courtroom
observation,
or
client
interactions,
could
contribute
significantly
to
what’s
perceived
as
“gut
instinct.”
AI
taking
over
grunt
work
wouldn’t
erode
this
instinct
although
it
might
shift
its
foundation
to
other
forms
of
experience.
Practical
Solutions
So,
I
still
think
we
will
survive
AI
and
that
future
lawyers
will
turn
out
fine.
But
we
do
need
to
at
least
consider
that
the
RPD
theory
may
have
some
merit.
While
it
makes
little
sense
to
make
younger
lawyers
do
what
AI
can
now
do,
we
may
need
to
think
about
things
that
would
build
pattern
recognition
and
be
more
purposeful
about
it.
Things
like:
-
Adopting
the
notion
of
deliberate
pattern
exposure
by
creating
structured
programs
that
expose
junior
lawyers
to
diverse
scenarios
even
if
not
billable. -
Embracing
simulation-based
learning:
Use
case
studies,
mock
transactions,
and
scenario
planning
much
like
what
Alta
Clara
is
doing
and
I
have
discussed.
The
idea
is
to
present
young
lawyers
with
a
simulated
legal
scenario
and
then
have
senior
lawyers
critique
their
solutions -
Thinking
of
AI
as
teaching
tool:
Teach
younger
lawyers
how
to
effectively
use
AI
and
then
mentor
them
to
the
separate
the
wheat
from
the
chaff
in
the
results. -
Embracing
AI
for
younger
lawyers
and
recognizing
that
it
may
enable
younger
lawyers
to
do
more
sooner
so
that
the
development
of
pattern
recognition
will
come
more
through
actual
experience
instead
of
grunt
work
in
a
dusty
warehouse. -
Evolving
mentorship
programs
to
ensure
senior
lawyers
actively
teach
pattern
recognition
rather
than
assume
it
develops
naturally -
Understanding
that
what
makes
a
good
lawyer
is
not
only
how
well
they
do
mundane
tasks
but
how
well
they
adopt
and
learn
the
key
skills
that
document
review
may
in
the
past
have
constituted
only
a
small
part
of. -
Engaging
in
client
education
to
help
clients
understand
the
value
to
them
of
investing
in
junior
lawyer
development
A
Path
Forward
Clearly,
AI
will
alter
not
only
the
substantive
practice
of
law
but
how
we
need
to
train
younger
lawyers.
Whether
you
buy
RPD
or
not,
AI
and
the
work
it
replaces
isn’t
going
away.
And
the
need
to
maximize
its
use
across
the
board,
including
training,
is
imperative.
So,
we
need
to
think
carefully
about
training
our
lawyers
for
the
future
and
what
skills
they
need
to
have.
Wringing
our
hands
over
what
used
to
work
doesn’t
solve
the
training
dilemma.
Let’s
be
thoughtful.
And
purposeful
about
training.
The
legal
profession
stands
at
a
crossroads.
We
can
either
thoughtfully
redesign
legal
training
for
the
AI
age,
or
watch
as
tomorrow’s
lawyers
lack
the
pattern
recognition
that
makes
today’s
best
attorneys
so
effective.
Stephen
Embry
is
a
lawyer,
speaker,
blogger
and
writer.
He
publishes TechLaw
Crossroads,
a
blog
devoted
to
the
examination
of
the
tension
between
technology,
the
law,
and
the
practice
of
law.