Why Meta Wants To Build Artificial General Intelligence — With Joelle Pineau, VP of Meta AI Research

Channel: Alex Kantrowitz

Published at: 2024-01-24

YouTube video id: pjYGuH8pFXA

Source: https://www.youtube.com/watch?v=pjYGuH8pFXA

the head of meta's AI research division
joins us today to discuss the company's
pursuit of human level artificial
intelligence The Cutting Edge of AI why
its open sourcing its large language
models and plenty more in the only
podcast interview the company is giving
about its recent news all that and more
coming up right after this welcome to
Big technology podcast a show for
cool-headed nuance conversation of the
tech world and Beyond Boyd we have a
show for you today we're recording uh to
the minute on Wednesday here right
before we drop this episode because
there's breaking news coming out of meta
all about the moves uh that they're
making with their AI division their
pursuit of human level intelligence and
we have none other than Joel Pino here
to talk to us about it she's the head of
meta's AI research division formerly
called Fair now I guess it's called
mayor and um still Fair fundamental AI
research we love the name okay well keep
it fair uh keep it running we spoke
actually in in October 2022 before chat
GPT so this is going to be a really cool
moment to talk a little bit about where
we've come from there and where we're
going Joel welcome to the show great to
see you thank you Alex great to be here
so if you recall in October 2022 and we
spoke a couple times at the world Summit
AI one of the it's kind of funny because
like the big storyline then was whether
AI is sentient and this was kind of a
moment where like all the big research
houses had big large language uh model
chat Bots uh internal and they hadn't
released it yet and it's kind of
interesting how Society starts to talk
about a breakthrough right it sometimes
goes in a weird Direction before we're
actually refocused on what matters and
and now I think we are refocused on what
matters right there's been much more
talk Beyond sentience in terms of like
the near-term viability of this
technology I'm curious just to start
what has surprised you in the research
since that discussion not necessarily
about okay we all know that it's you
know now taken off and it's been hyped
but has there been anything that's made
you sit back and be like wow we can
actually do more than we thought we
could you know a year or a year and a
half
ago so many times this year honestly and
it's you know it's great to to think
back to that to that point in time I I
hope you didn't ask me for any very
specific predictions even for someone
who's deeply in the space of AI just
predicting how this this is unfolding um
continues to be full of surprises um I I
will say you know it's also been
interesting it's the faster we progress
uh the more we have a sense of how much
more is left to do and so though you
know back you mentioned back in October
2022 we were worried about sentience and
and we don't we hardly talk about it now
um and yet we are so much further along
on the map in terms of our ability to
have models that deeply understand
information and process multimodal data
so we're getting further along and yet
we worry about some of the the more um
more concrete problems we've talked a
lot this year about safety for example
about how to make sure that we have
models that are performing well but also
our our aligning with the values of of
peoples and the needs of people um which
I consider sort of a much more grounded
uh problem that we can tackle with
research so that that's I think the the
major change that I see sign ific
progress but that means we also have a
much better view of what are the real
problems we need to solve yeah it's
funny because back then we also had a
discussion about whether we should be
focusing on like the short-term or the
long-term problems and obviously those
are both worthy of attention and it's
kind of wild that the focus on the
long-term problems it seems like blew up
open AI over a weekend and maybe it's
been put back together now but the talk
from meta now is actually focused on
some of the more Big Ideas that people
might have thought were more long-term U
but now it actually seems like you know
it might be closer than we think at
least According to some of uh what we
hear from open Ai and others so um this
is a quote from Mark Zuckerberg that
just came out fairly recently he says uh
as as recently as last week we've come
to view that in order to build the
products that we want to build we need
to build for general intelligence so I
mean Yan Lon in our discussions I've
been speaking with him since 2015 one of
your colleagues he's always talked about
how the goal is building for artificial
general intelligence so when I saw Mark
come out with that last week I was like
yeah yeah that's been the focus for meta
but all of a sudden it almost feels like
there's there's a a more pragmatic or it
feels more real now than it did before
am I reading that right like what is
leading us to now start to like talk
about this as something that's not pie
in the sky you know 20 30 years down the
road but something that might be
achievable in the you know nearer term
yeah I mean
Yan and I and the team in Fair have been
talking those terms for for many years
it's been clear we've been putting in
place sort of a portfolio of projects
that are trying to build the building
blocks towards general intelligence um
in in the last year uh Mark as well as
many others has taken a deeper interest
in what's going on in AI I think he was
always aware of a lot of the good work
we were doing um but uh didn't dig in
quite as deeply and and in the last year
definitely has um and through a lot of
conversations you know I think has come
to to see how in many ways the the path
even to Bringing AI to to the products
that that people use and and love from
the company the path to making those AI
systems better goes by through building
general intelligence not narrow
intelligence and and we've done a ton of
work on AI on the platform over the last
few years that was what I would call
more narrow specialized models um we can
continue to do that but the the bigger
step change are going to come through
the more General model building
Foundation models building up to World
models that essentially can capture a
much richer version of the information
um so I think that's that's what you're
hearing from Mark it's it's things that
you've been hearing from from Yan myself
and others uh Through The Years we're
we're working together to to connect
these pieces together uh both the the
research road map as well the product
road map and and make sure that we have
um the ability to to connect these
together so so the ability to have our
research um quickly diffus in the best
way possible through the product and the
ability to learn the thing about general
intelligence is you have to solve many
different problems to have you know the
ability to claim general intelligence
and
unfortunately there are a lot of use
cases across meta across our family of
products and so that's giving us
wonderful material with which to work so
why so let's go back to you know that
October 2022 discussion that we had
before chat GP PT come out like the idea
of me asking you this question about
like why is human level intelligence now
in Focus I never would have asked it it
just didn't seem like it would be
something that would be relevant to ask
but now it does seem more relevant and
we're hearing it more and more in the
discussion so you mentioned World models
um foundational models but what about AI
research now is allowing us to ask those
questions um I think it's because you
know the the models are are getting
increasingly General if you look at a
model like Chad DPT the Llama family of
models that we've been
releasing you know they started just as
word prediction models all they would do
is take in sentences and predict what
comes next um and what we're seeing is
we can use them now through through many
other uses whether it's to predict
things that are not just words but
they're actually code and some of that
code is actually executable um or you
can predict you know the components of
an image and then you can plug in a
diffusion model or or other kind of
synth synthesizer to to realize the
information so what's started as just
language model has become much more
General on its own it gives us a path it
may not be the path but it gives us at
least a path to move towards general
intelligence um and it's an exciting one
it's one that that we're exploring it
doesn't mean that we've stopped
exploring other paths towards General
intellig but but that is definitely the
one that has has proven to to make the
fastest progress and what would you what
would you say the path is uh the path is
to essentially capture a lot of human
information through this this uh
representation that we call language and
so the the hypothesis that you know even
things that are not necessarily text
based originally if you describe them
through these discrete tokens and
sometimes these discret tokens are the
words that we use to express but but
sometimes discrete tokens are for
example code numbers um essentially like
chunks of images and these discrete
tokens are a path to representing all of
human information there was a study that
came out uh last year basically saying
that these models can't generalize
outside of their training set you know I
think that was like a lot of the hype
around these models were people saying
that they were really able to have these
um capabilities that you wouldn't expect
emergent capabilities and the study
basically pushed back on it and was like
listen they're not going to generalized
beyond their training set and your
evaluation of that study basically you
know leads you to believe either a if
you believe that study then you're a lot
less optimistic about this wave if you
don't believe that study you're you can
be you can really use your imagination
and believe that what we're hitting on
now these foundational models that you
talk talk about can lead us in
directions that we never could have
dreamed of so I'm curious what your
evaluation of is of that study is and
and how we should be thinking about this
I I tend to really like be quite
balanced on a lot of these questions I I
think it's very easy to kind of you know
pull opinions to one side or another but
but the truth is like machine learning
algorithms can generalize that is a
property of of how we build these
algorithms even the simplest just linear
models they do linearize they just your
eyes along a line so you know the the
fact of the matter is though when you
project that into very very high
dimension so some of these models have
hundreds of billions of parameters you
have to think of like you're you're
learning a function in that really high
dimensional
space the directions in which you can
generalize are so many that it's hard to
know which are the good directions to
generalize and which are the poor
directions to generalize the more data
you have the more that constrains that
question so I I do believe they can
generalize I think they generalize
relatively narrowly or at least you know
as long as you stay close you get a good
manifold of information when you start
to go really far a field from your data
because the dimensions are so large you
get all sorts of all sorts of noise so
the advantage and the one of the reasons
you know a lot of the progress has been
through better and bigger data sets
bigger but also cleaner data um is
because that really defines which parts
of this really high-dimensional space
are the most interesting one and when
there's not a lot of data to populate
that space then the models tend to
regurgitate the things that they've been
trained on so let's go back to that
Zuckerberg quote that I read earlier
we've come to view that in order to
build the products that we want to build
we need to build for general
intelligence now we talked a little bit
about why that is now relevant the path
towards general intelligence but now I
I'm kind of left with another question
which is why does meta need to build
general intelligence in order to build
the products that you want to build I
mean yeah I mean ju just looking at like
a couple of the AI products we've we've
released this year you know one of them
is the m meta AI assistant uh people who
are in the US have been able to to try
this out on some of our platforms where
you can essentially ask for questions
and and ask for for assistance in that
case you know there's a sense that that
it has to understand a very large
spectrum of information to to be able to
to do well um and as we incorporate more
data and as we perfect this this ass
instant the more it's going to have
essentially World Knowledge the the
better it's going to be um another
example is um for those who've been
following our work on on AR devices the
the smart glasses that we released uh
earlier this year also come now uh with
an AI model also accessible mostly in
the US at this time um there too you
know you have essentially a a more
embodied version of of this meta AI
assistant that sees the world as you see
it that is able to take on some action
in this case the actions are not just
words it can take pictures it can
provide information it can record
information um and so to be able to do
well in a wide set of different tasks
with a wide set of different people
different
environments you need to have to move
towards more general intelligence
um that's really the where that where
that connects you know the research work
we're doing and already what we're
seeing in terms of the the applications
that that meta is putting out there now
let's say you do achieve this and you
open source it is that kind of like the
the end like is human reaching human
level intelligence or general
intelligence kind of like the end of AI
research or is there more to do after
that
happens there is no end to this journey
I I mean I hope there's no end to this
journey Journey right like do do we as
adults sort of say okay I'm going to
keep on growing my knowledge and at some
point in time I don't know for some of
us maybe 25 some of us maybe you know 75
you decide like okay now I'm done like I
have reached where I am in terms of
human intelligence I don't think that's
how it works for humans the the world is
always evolving there is always more to
be curious about and and so I think
that's the path that we are on with our
AI algorithm similarly need to stay
curious about the world that they evolve
in and over time they need to figure out
you know how to integrate that that
information um in and sort of rise to
the challenge of the world that they're
building but because the environment is
not static I don't see us coming to an
end that's so interesting because it's
always described as the finish line and
actually there's people who would argue
that there's no such thing as as human
level intelligence that the second you
hit that you're basically left with
super intelligence and game over but
yeah I mean I I don't really ascribe to
that that scenario I have to say um and
and the the other Nuance I will add to
this you know often this notion of
general intelligence is is articulated
in in the context of like a single agent
a single Uber intelligent agent um and I
don't I I don't think that's really
where we will move towards either um
there's there's clear evidence that that
as a species humans animals we we learn
so much more through interactions and so
much of our culture and our intelligence
is derived from our ability to interact
to collaborate um so I think that's also
going to be a super interesting door to
open as we as we are on this journey to
think about how do we build AI agents
that are not just you know pushing for
single entity intelligence but are
connected to a network of other
intelligent agents whether synthetic
silicon agents or or human agents well
it's so interesting because language of
course like speaking of types of
intelligence language is only one type
yes and Jan and I spoke about this on a
recent show not so recent anymore but
your interactions in the world teach you
so much that you never learn with
language your understanding of gravity
for instance is not something that like
you can implicitly understand from
language so are you doing research now
to help metas research division to
figure out stuff beyond words and images
absolutely um and I would say you know
that may be one of the distinguishing
factors compared to other research group
out there there's a strong belief that
um having AI agents that are deployed in
the physical world where the notion of
embodiment is important um is something
we should be pursuing we have a research
team that that's dedicated to this they
they do some work in robotics in
particular because that's the the best
agents we have to to consider physical
embod embodiment spatial constraints um
it's not necessarily because meta
intends to commercialize robots it's
because by going through these
essentially devices we have a lot to
learn about how to build AI models that
that live in the physical world and in
in the work that we've done recently
with the with the smart glasses the the
models that proved to be useful for that
use case came out of the work that this
group was doing people who were looking
at Robotics and devices physical devices
living into the world and building AI
models specialized for that um was
incredibly useful to inform the work
going into the classes of course we also
leverage the work we were doing on
language and our Lama family of models
but Lama on its own doesn't make um for
for the best uh the best assistant on on
glasses because it doesn't have enough
of an understanding of the physical
world of images uh and so on now there
there are some people saying that the
reason reason why met is now speaking
about AGI is because open AI is speaking
so much about AGI and other research
houses are and getting the talent to
work on these projects is really
difficult this is something that Mark
actually said in that Verge interview I
think that it's important to convey uh
because a lot of the best researchers
want to work on the more ambitious
problems so I got to ask you straight up
like is is the talk about AGI more of a
recruiting
thing know I mean like of course we love
to have great talent and of course this
is a competitive market for talent but
but we don't talk about anything just
because someone else talks about it like
we we genuinely are doing the work and
we've been doing it for a number of
years um it's there's no major shift in
terms of our ambition to solve AI that's
been inscribed in our mission and our
goals for fair for many years now um
Mark is talking about it now I think
he's excited about the work it's
wonderful to have have his support to do
it but but it doesn't necessarily
fundamentally change the the the
problems that we have to solve the work
that we're doing I think there's also a
sense that we are you know we are being
more explicitly ambitious about this
work which goes along with some of our
investments um on the compute side uh
which are necessary to to fuel that work
and so that's why it's coming out um
maybe more from what you're hearing from
Mark but I think if you go back and and
listen to what Yan or I or some of our
other senior researchers have been
saying for for a number of years
there's not a departure there right and
briefly on the talent Market what does
the talent market look like right now is
there a real scarcity in the type of
people that can do this type of work and
what is it like uh recruiting against
fast growing and especially in terms of
valuation competitors like open AI
anthropic
Etc yeah it's always been a very
competitive market I would say going
back to about um 2016 2017 um um since
then I I don't really remember a year
where where it was like an easy slow
Market in Ai and so it continues to be
one of the things that has changed in
the last uh year or so is um mostly on
the startup scene I would say you know
three years ago we didn't feel much
competition with a startup scene now we
do a lot more I tend to view this as
relatively positively to be honest with
you and that's one of the reasons we
open source our work we we genuinely
believe that more people working on this
is good and and so when we we open
source our work we get to Leverage The
the creativity of of a greater number of
people and there's many more than than
we can hire um so I think that the very
very top talent that can train these
models continues to be incredibly
valuable to meta as well as to to other
organization fortunately there's also a
good pipeline of of students you know I
do have an affiliation with Mila the
Montreal Institute for learning
algorithms there hundreds of amazing
grad students coming out of that
Institute as well as others we've set up
some joint PHD programs in some cases so
that these students have an opportunity
to come uh work at least uh part-time or
through internships with us and so we're
both you know sharing with them the work
that we do as well as having an ability
for for us to to see whether they're
good fit for our work so I feel like we
have a great talent pipeline but it
continues to be a competitive market um
got to ask you about open source uh Brad
Smith from Microsoft has talked about
how open AI is the most open and I'm
kind of curious from your position are
they living up to that open name is
there real open sourcing there and uh
what is the state of of open sourcing I
mean why is meta open sourcing outside
of like I mean from like a a you know
met as a business so from a business
perspective why open source yeah there's
there's different levels of open
sourcing right I do do think you know
having an AI model where you provide an
API is sort of one one layer of that
which is something that open AI has done
um but there's a lot more that goes on
and so you know from just providing an
API you can make available the code that
was used to train it you can make
available the trained model weights
which enable someone to run the models
and then there's a number of other
artifacts that come across uh from this
we've been focusing on making available
like model cards that give a better
understanding and transparency about our
models um good use guides uh tools for
safety and so on so there's like a whole
ecosystem of AR depths I think the
purists would say like everything has to
be to be out there um in an open way so
we even have some some people who are
coming from the software op Source
Community who who feel like we're not
you know living up to the the full view
again there's a Continuum on this it is
clear that meta has taken a much more
open view than other big players in the
space and in particular
we've been releasing some of our code
and model weights for some of our larger
models including llama um it comes from
a lot of of deep discussions uh in in
doing that um and so I think there there
maybe a
misperception that that um we're you
know we would um do this without you
know the without any process or
reflection and there a little bit of a
religion that's really not the case we
have quite a thoughtful process that's
been put in place you have to remember
we've been doing open sourcing work for
10 years since the first day of this
organization so we've built up a lot of
muscle of how to do that in a in a
responsible way we do it in consultation
with a wide set of people who have deep
understanding of safety ethics and and
so on um who get brought into the into
the process and what's been wonderful to
see in the last year is as the
conversation has been moving and as the
models have been going better and
getting bigger we've invested a lot more
into being thoughtful about our release
process and so I would say now we have a
much more mature process than we did a
year and a half ago um that involves a
much more diverse group of stakeholders
we have a really rigorous process in
terms of measuring the risks of these
models across different different
categories of risk um so it's been
exciting to see how much our commitment
to open sourcing has driven us to
innovate and we've open sourc a lot of
those Innovations on the on the safety
side I think the the purple llama tools
is an example of that which we released
in December and so it's been great to
see that um I I do hear a lot of people
who are concerned about open sourcing um
I and I have many conversations with
them including at other other large
organizations
my my worry about closing the doors down
now is that the models are only getting
better and so if we don't really them
now we really miss an opportunity to
develop the muscle we need to make these
models safer um and I don't think
today's model are the ones that are
going to you know bring to the front the
hardest questions these models are yet
to be trained is there anything is there
anything standout that you've seen being
built on top of uh the open- source
llama model that meta has put out there
anything stand out in terms of like a
cool product that you've seen and
anything concerning that you can talk
about um there's definitely been dozens
of of of product that are that are
coming out of that um how about naming
yeah do you want to name one yeah let
let me take an as an example our segment
anything model which is a little bit
different than our than our llama model
um but but I think it's been the one
that has been just incredibly impactful
in terms of people quickly building on
it our our segment anything model is one
where you take an image and it gives you
a detailed segmentation of that uh of
that we really St back in April
including a lot of um tools and data to
to go along with it and and within days
we had people who had built up
applications essentially for um
conservation applications so being able
to track down some species who may be
endangered using that to to follow them
we had people use it for uh the
treatment of medical images so
segmenting cells from some some these
images and it's been wonderful to see
that that explosion that explosion of
work um on the language side we also saw
many people build up all sorts of um
different tools and in particularly the
the work that we're most excited about
is uh the work on efficiency to be
honest with you um there is so much that
we can do to make these model more
Compact and and efficient and and and
running um really really fast with low
energy and I think that's one of the
things that I've been most excited about
seeing there's lots of other
applications too um anything that stood
out and made you say oh that's that's
not good that's not what we want um
there are definitely some that are that
are getting flagged that we discuss
internally um probably not going to go
into the details of them right now but
there there definitely a number of them
that that we are tracking I will say in
a number of the cases that we are most
concerned uh people are not respecting
the terms of use of these models so we
release these models with very clear
terms of use and and people may not be
respecting those those terms do you have
recourse once they disrespect those
terms yeah I think that's I'm going to
go into the the details of that today
but this is you know this is definitely
part of the conversation we we we we are
thoughtful about the conditions under
which we release and and so we're
thoughtful about the follow through as
well before we go to break I want to ask
you about this move toward getting these
models to reason uh there was like this
momentary freak out uh around this qar
model thing that open AI apparently has
has uh developed internally which gets
people to to re can gets the model to
reason what's your perspective on this
technology moving towards like the
ability to reason and how should how
should we think about it when we think
when we see stories like the one about
qar I I I mean I think the number one
thing is just like don't get too worked
up about it um the the the amount of you
know speculation probably far outweighed
what what was going on there I don't
have firsthand information on on on qar
we have a lot of you know a lot of
speculation of our own of of what it is
what I will say though is to some degree
people shouldn't be too surprised you
know a while ago um we shared a model
that could play the game of diplomacy at
the level of human player I don't know
what people thought that Cicero exactly
right Cicero was having conversations
with other players and it was reasoning
about the game strategy and so this was
an example of a model that had language
and that could reason than arguably in
the hardest game uh out there so I I
don't think people should be surprised
that a language models have the ability
to be to be effective in reasoning task
especially paired with mechanisms in the
case of ciso we were using some search
mechanisms inside to be able to to
achieve reasoning it's you know it's
it's a different architecture than than
what we have in in llama um but a lot of
the ingredients of how to do reasoning
have been explored in AI for 40 years um
and and are published and well known to
to anyone who's taken even an undergrad
level course in AI um so I'm not saying
that there's not any innovation in the
work that that open AI is doing or in
the work that's happening across the
community I'm just saying it's not like
a magic ingredient I or I'd be extremely
surprised that so what could The Next
Level jump there be I mean there's a lot
of theories of how to achieve reasoning
in these models one of them is to
incorporate search um as part of the
model and another one is uh
incorporating for example a lot more
coding abilities coding is executable
coding allows us to essentially dig in
through a a a sequence of operations um
another another you know direction that
many groups are exploring is the use of
retrieval based uh techniques so you're
retrieving information some of that
retrieval can can make use of
information where reasoning is present
in the information so lots of different
ways to go about it um all we're
exploring many of them any respectable
AI research group probably is too um
what's really going to make the
difference is is how do we bring this
together right how do we make sure to
have the right way for these components
to integrate and in some ways that's
still the hardest question in AI um how
do we have different components working
together in a very coordinated way is
there anything that you could see in the
sort of research or production that
would freak you out or are you sort of
com cool collected about where we're
heading there's stuff I I mean I don't
tend to freak out a lot um there's stuff
that concerns me every day um you know
we we we review you know rigorously the
performance of our model for for for
different aspects you know there's many
cases where I see a model and the
performance for example on safety
benchmarks isn't what I would expect it
to be and then we go back and we keep on
on working on it and so it's it's not
that there I don't think there's a ton
of work to do I just don't feel that
like you know freaking out or being
fearful about it is the best way to go
about it I think you just have to to
look at the the data in a collected way
in many cases we don't even have the the
right way to analyze the properties of
our model you know are do the model safe
or unsafe does it have you know toxic
Behavior does it have IAS there's a lot
of work to do to even develop the tools
to assess this so we can look at it in a
in a rational way so we invest a lot in
that
also we're here with Joel Pino the head
of meta AI research division still
called Fair fundamental AI research uh
we've talked a lot about re uh the
research side on uh this side of the
break on the other side of the break
we're going to talk about product
because Joel's division has recently
moved uh toward the product side of meta
and we're going to talk about what that
means right after this and we're back
here on big technology podcast with Joel
Pino the head of meta's AI research
division um your division just moved
toward the products or under the product
division within meta um let me start
this segment with this question it's a
broad question I I don't think I've ever
seen a disconnect as much as I'm seeing
now where the discussion of where this
technology can lead and what it does
today is so so I would say even divorced
from the products that we've seen I mean
yes chat GPT was was groundbreaking and
still incredible to use and so is like
some of the competitors um but beyond
that have we really seen the product
momentum when it comes to building on
large language models and the you know
we we heard so much about an Enterprise
yeah we've seen some co-pilots from
Microsoft stuff like that the bots in in
the messaging apps that met is created
but you know for all the talk of
Revolution it seems somewhat like an
evolution so what do you think about
that and what am I missing
here I I I do see it as a bigger step
change I think than than you're
articulating it um I think we have seen
the birth of what I would call an AI
research product um and so if I take you
know for example the GPT family of of
models I I do think there there there is
a a real product there people are using
using it some people are using it every
day um and so I I don't I don't think
we've seen anywhere near everything that
is possible but I think we have to have
a very open mind that the the product
that are AI first are going to look very
different than product we've we've seen
before that being said I I I will say
you know as much as we spent a lot of
time worrying about what is the path on
the research side I I do think we need
almost as much exploration on the
product side you know the research side
the space of hypothesis to build these
model is huge but on the product side
like the space of new things you could
build with this is huge and and we don't
yet have nearly enough information about
what are going to be those products and
those experiences that people are going
to actually use every day and love using
so I I'm you know as as I talked to
Partners across the company one of the
things I encourage them to do is to
really embrace the exploration that
comes out of having a completely new
tech stack compared to to what they had
before and not just take you know the
products that they know and like shove
AI into them but completely reimagine
what is possible so that's been a really
really fun conversation to have and one
of the things that is going on is Met
has brought a bunch of AI Bots into the
messaging apps can you tell us a little
bit about how that's going I mean I saw
like the was there's like 12 or 20
different Bots that are in these apps
and I played with them for a little bit
and then I kind of lost interest and I
haven't like seen any reminders that hey
they exist so how's adoption been there
what can you tell us about those are you
asking for more reminders that they're
there because we can do that honestly
yes honestly yes I think that would be
good okay um yes the Bots are there
they're they're available um H the the
the Bots are an example of exactly what
I mean right this is product exploration
to to some degree at at its best in
terms of like trying out different
things there's there's an intuition that
there's enough there that it's worth
putting it out in the hands of of people
there's enough conviction as well as
data to support uh releasing that for
for people to use but I I think it was
very much the kind of product we hadn't
done before and and we're going to learn
so much out of getting that into into
people's hands you can think of it as
really accelerating that that cycle of
of development and there's some Bots
that are doing quite well that are
seeing quite a bit of use some Bots that
are seeing a lot less use I don't have
the the numbers with me andbe for for
your listeners to understand you know fa
does the fundamental AI research and we
have a sister organization um that is
more connected to the product and is
releasing uh those uh those spots um
we're tracking that really closely
that's feeding back into the product
exploration conversation going on I
would say the the Bots as well as the
meta AI assistant are within a category
of things that we call AI agents and so
we have a a pretty wide exploration
within uh that space of AI agents
um that that you should expect to see
new things new things coming in in years
to come um and the the other example
that we explored a bit this year is uh
on the smart glasses where we also have
an AI system running on that which is a
very different very different experience
uh compared to the the desktop desktop
or Mobile cases yeah so adoption how is
adoption looking with those messaging
Bots we'd have to you know we'd have to
get someone else on your on your podcast
to give you more detail on that yeah
absolutely I'm sure we can find you
someone who can give you some of that
information so you mentioned that you
have the product teams and you have fair
but fair used to be in reality labs and
now it's like directly under the product
team within metal why did that move
happen um so I mean we we had a
wonderful set of uh colleagues and and
great work happening in in reality Labs
research the truth is right now ai is
moving so fast um that it's really
useful to be close to products that are
in the hands of billions of people to be
able to have that quick product
Innovation that quick signal back to the
back to the research we were already
working in close collaboration with the
with a family of app product teams um
but this just makes things go a little
bit faster and the the geni team that is
really putting out some of these uh AI
characters and meta AI assistant was
already in that product team so bringing
us together gives us the ability to to
be much more coordinated um in
particular from the research to you know
building up the products and then and
then releasing them um we're still going
to continue to do a lot of work on the
reality lab side you know we've been in
that work for for a few years we've
built up a lot of exciting projects
going to be maybe a few more years
between between now and when when some
of these um get on the market but these
projects are not slowing down in any way
I think there's a really good
understanding at the company level that
right now the more we Accel cerate the
AI road map it is going to benefit both
the existing products as well as the the
arvr and the reality lab side of the of
the company so um I think that's that's
really where we are with with this one
one thing that seems like it's really
going to be a thing that people talk
about this year is video generation
we've just seen a little bit come out
this week from Google I know that you
guys are working on it um tell us a
little bit about what that could look
like I mean it's thing to sort of typee
in draw me a picture and you get one out
from dolly or or meta has one an image
generator as well but the video
generation seems pretty
wild yeah it's been it's been great to
see that it it's not surprising as soon
as you you know you have good image
generation every time we've had progress
in terms of image generation the next
step is how do we do 3D images and how
do we do videos like these are the two
two dimension in which um people quickly
extend any progress in in uh image
generation on the video side I would say
we've seen we've seen much much better
models coming out um but we haven't
totally cracked the problem of
generating long form videos um the the
temporal coherence is is quite tough and
I think you know for those of you who
who know a little bit more about video
you know there's a piece of spatial
coherence that you need to be thoughtful
of and that's the piece that image
generation has has to some degree solved
uh in the last year but the temporal
correlation is something that right now
is it's harder to do we get really good
quality video generation if you can you
know intervene and kind of set a lot of
the frames and then you kind of use a
diffusion model to to interpolate in
between but to go from a really high
level for example a script written in
words um and to have like a full you
know full length feature film um is
still going to take us a little while
one of the biggest problems there is to
think about how to do generation in sort
of a hierarchical way not just do frame
after frame after frame um but actually
think of how do you generate globally
some properties of your video and then
go through more and more granular
resolution over space and over time um
this is something that Yan has been
thinking about a lot he's working
closely with some of our research teams
in New York in Montreal and Paris to
make progress on that and so I'm you
know I'm leaving a lot of that on on him
to to drive but I know he has a lot of
ideas on this on this topic and how also
to achieve that in a way that um isn't
too intensive in terms of data and
compute right I think that that's when
you sort of get into like kind of
predict and plan and sort of and really
understand what reality is that's some
fascinating stuff okay we're coming
close to a landing here very quickly uh
we also spoke we really had some fun
conversations when we met the first time
we spoke with the I think Chief
technology officer of Nvidia and Mark
just announced that you have 350,000
Nvidia h100 chips and we'll end up with
650,000 by by the end of the year Nvidia
h100 or equivalent and just curious from
your perspective as a customer of Nvidia
what makes those chips so effective for
you now it's obviously a technology
component but there's a software side of
it as well right so can you talk us
through exactly what makes them so
appealing and do you think they're
they're they are going to just be the
Imp the unparalleled developer of these
chips forever or are you starting to
look at others like um arm Etc Intel you
tell us yeah I mean it's honestly it
it's it's clear to everyone that a lot
of the progress in AI has been fueled by
the availability of gpus uh built by
Nvidia It's Not the Only Solution Google
uses a lot their own tpus uh as an
example so there's a few there's a few
others but but overall I think nvidia's
gpus have been essential to to the
progress and we've been uh fortunate to
have many of them to to power our own
research there's a couple things that
make them great one you know the gpus on
their own have the ability to to parize
a lot of the computation which is
essential for training these models and
we also have the ability to build them
into systems you know networked with
very fast interconnection between them
to allow information to be passed around
uh very very quickly and when you do
that at scale with with a few thousand
gpus you can train some of these larger
models so that's really the the
essential ingredients in terms of the
trajectory there of of course you know
as as all responsible organizations
we're looking at all at all options that
could accelerate our work we keep a
close eye on the development of of
Hardware um right now as Mark has has
shared you know I think the the betting
on the dpus from Nvidia is a is a sound
bet for for our research but we're
always interested to see in a in all
aspects of the stack are you going to
build your own
ships um we will definitely be exploring
some of that yes yes I mean we built a
lot of hardware for reality Labs we have
some specific needs and you know as much
as we you know look at that for the the
arvr devices there's also a great group
doing some of that Innovation inside our
infra Team all right last question for
you we started the conversation talking
about AI re reaching uh human level
intelligence think that's going to
happen let's say five years over under
have a you have a perspective on that
one uh in five years we're going to see
a really strong systems across a broad
set of tasks I have some some strong
conviction that that we're on a path
there after that you know I don't want
to bend any intelligence into narrow box
whether human or AI but we will be
amazed by what gets done in the next
five years all right can't wait to watch
it Joel thank you so much for joining
thank you Alex all right everybody thank
you for listening uh we will be back on
Friday with a new show Breaking Down The
News until next time we will see you
then and uh we will see you for our
Friday show on big technology
podcast