2026 will be the year of AI software after intense focus on hardware: Bessemer's Byron Deeter
2025-12-29_16-25 • 3m 29s
? (Anchor)
00:00.000
Let's
bring
in
Byron
Dieder.
He
is
the
partner
at
Bessemer
Venture
Partners.
Byron,
it's
great
to
have
you.
So,
what
do
you
think
changes
in
2026?
Byron Deeter (Partner)
00:08.760
Wonderful
to
be
back.
And
uh
this
is
the
extension
now
of
the
AI
build
and
we
can
just
look
back
at
the
cloud
story
to
see
how
this
plays
out,
which
is
the
fundamental
pieces
go
in
first,
which
is
the
data
centers,
the
uh
compute,
and
then
the
core
infrastructure
uh
foundational
Byron Deeter (Partner)
00:25.880
layers,
in
this
case
the
foundation
models,
and
then
1,000
flowers
blue
on
top.
It's
the
upstack
applications.
And
so
as
you
look
at
the
coming
trillions
of
dollars
of
data
center
spend
going
from
about
550
billion
a
year
to
1
trillion
a
year
shortly,
there's
about
a
9x
Byron Deeter (Partner)
00:41.840
multiplier
that
comes
for
implied
enterprise
value
of
software
because
they
are
the
downstream
customers
of
the
data
centers
ultimately.
That's
the
bet
we're
making,
which
is
the
build
that's
coming.
And
ultimately,
if
you
roll
it
forward
to
the
5
trillion
of
combined
spend
over
Byron Deeter (Partner)
00:56.840
about
the
next
half
dozen
years,
that
implies
about
40
$35
trillion
of
software
and
application
layer
value
that
should
be
coming
down
the
next
several
years,
and
that's
the
bet
that
we
think
is
the
next
most
exciting
bet
to
make.
? (Anchor)
01:10.440
I
wonder
if
that
that
does
that
lift
some
of
the
worst
performers
that
we've
seen
this
year.
Adobe
is
down
20%,
sales
force
is
down
22%,
Workday
down
about
18%
all
on
this,
you
know,
software
versus
AI
trade.
Byron Deeter (Partner)
01:27.960
So,
I
do
think
that
first
and
cloud
and
software
as
a
service
is
oversold,
but
it's
not
obsolete,
and
so
there's
going
to
be
a
separation
there.
Who
are
the
beneficiaries
of
AI
that
are
going
to
lead
the
next
cycle
versus
those
that
aren't
going
to
make
the
transition?
And
so
if
Byron Deeter (Partner)
01:42.160
you
look
at
a
service
now
or
a
Viva,
I
think
those
are
names
that
are
really
interesting
to
watch
to
see
how
they
incorporate
AI,
and
I
think
that
there's
a
little
bit
of
value
left
there
that
they're
under
their
peer
group.
You
look
at
some
companies
like
a
a
cloud
flare
or
an
Byron Deeter (Partner)
01:54.960
app
loving,
um
even
a
snowflake,
and
at
their
current
values,
um
they're
the
high
suppliers
in
this
kind
of
genuine
ecosystem
and
I
have
to
argue
the
opposite
which
is
that
perhaps
they
are
over
valued
relative
to
their
their
point
in
time
and
the
challenges
that
are
coming
and
Byron Deeter (Partner)
02:11.080
the
data
bricks
and
the
click
houses
may
be
more
interesting
bets
for
this
next
cycle.
? (Anchor)
02:16.440
But
I
I
I
would
assume
there
there's
companies
that
we've
yet
to
even
hear
from
conceivably.
I
mean
if
they're
going
to
be
hundreds
of
applications
and
thousands
of
agents,
I
would
assume
it's
not
just
the
names
that
you
just
mentioned.
Byron Deeter (Partner)
02:26.880
Very
much
so
and
that's
the
interesting
bet
that
we'll
start
to
reveal
it
so.
So,
everyone
knows
the
open
AI,
the
anthropic,
the
XAI,
uh
Google
Gemini.
That's
wave
one
and
step
one.
These
foundation
models
are
now
in
place.
We
do
think
there's
continued
growth
there.
They
could
Byron Deeter (Partner)
02:41.040
double,
triple,
4x
over
the
next
several
years
and
be
fantastic
outcomes.
However,
it
is
that
entire
layer
of
next
gen
companies
that
are
starting
to
be
be
be
built
on
top.
And
there's
this
narrative
in
the
market
that
the
foundation
models
are
going
to
do
everything
and
that
Byron Deeter (Partner)
02:54.120
their
agents
and
agentic
worlds
and
frameworks
are
going
to
take
over
all
of
software
and
eat
it
all.
I
don't
buy
that
at
all.
Nor
do
to
be
at
Best
of
the
Venture
Partners.
It's
not
going
to
be
a
four
vendor
ecosystem
for
all
of
software
in
eternity.
There
is
going
to
be
this
Byron Deeter (Partner)
03:06.720
wave
of
companies
being
built
and
we're
just
starting
to
see
those
now.
I
don't
think
those
are
going
to
be
your
IPOs
in
2026,
but
I
do
think
by
the
end
of
2026
some
of
those
transitional
companies
like
the
Canvas
of
the
World
etc.
maybe
candidates
and
then
for
2027
we're
going
Byron Deeter (Partner)
03:22.160
to
see
some
of
those
AI
native
application
companies
really
burst
on
the
scene
for
the
public's
investors
to
bank
on.
Autoscroll