Anthropic Chief Product Officer: Why AI Model Development Is Accelerating

Channel: Alex Kantrowitz

Published at: 2025-10-08

YouTube video id: GmcTq0Zo8kM

Source: https://www.youtube.com/watch?v=GmcTq0Zo8kM

Anthropic product head Mike Kger joins
us to talk about how AI model
development is accelerating and what we
should look out for as things continue
to move faster. That's coming up right
after this. Welcome to Big Technology
Podcast, a show for coolheaded and
nuanced conversation of the tech world
and beyond. Well, Anthropic has a new
model out, Sonnet 4.5, just months after
the series of uh Claude 4 models came
out. So things are moving fast and we're
going to figure out why they're moving
much faster and what the implications
are for the AI industry and businesses
as a whole. And we're joined today by
the perfect guest to do it. Anthropic
product head Mike Kger is here with us.
Mike, it's good to see you again.
Welcome to the show.
>> It's good to be here. Thanks, Alex.
>> So I remember sitting in the audience
for Anthropic's first developer day. And
it's funny because in the AI world, you
sort of you go and what is it? cat years
or dog years, I don't even know. Every
every month feels like a year. And uh
this was in May, May 2025. And I
remember yourself and Dario were on
stage saying, "Yes, we're we releasing
Claude 4 uh but you know, we're going to
release the next iterations much faster
than we ever have uh previously and
we're already at 4.5. How is it
happening?"
>> I think there's a couple of things that
we're seeing. I mean even just thinking
about I mean May again feels like a year
ago I think doggeears is about right. I
think there's a couple of things. One is
um we've been working much more with
sort of enduser sort of customers of for
example of our platform. Um and with
that we can hear like a much faster
feedback loop of hey signet 4 is great
in these ways. We wish it was better in
these ways. And you're starting to get
customers that really push the models in
really interesting ways. And that ends
up being very helpful for us on the
research side because then we can say
all right these are problems to be
tackled in the next uh version of cloud.
So for example uh one of them was uh you
know claude you know sonet 4 and even
opus for opus is our biggest model um is
good at writing code but you know tends
to get sidetracked or lost if it's
working over longer time horizons. that
was a real emphasis uh of sonnet 4.5 or
you know we you know put a lot of data
into the context basically how the model
is what it's thinking about in a given
point but at some point that gets filled
up and how do you then manage you know
to keep working on those things so
having that feedback loop really helps
and also gives us a lot of urgency
because it means that there's sort of
sort of almost like bugs in some ways
out you know that you want to go fix or
at least like feature requests that you
want to go fix. So that's that's one
piece. The other one is we've just
streamlined a lot more of our model
release story. So um I think uh having
now seen you know I joined shortly
before sonnet 3.5 which was back in like
May of last year. So really long time
ago in AI years um uh from then to now
just the sort of operational upleveling
that I think we've seen in terms of you
know how do we do how do we get uh early
access feedback from customers? How do
we give like the remainder of customers
uh like a good heads up so they can la
they can co-launch on launch day? What
does even that morning look like on roll
out? I was talking to a customer. He's
like, I've seen a lot of lab rollouts of
of models and this was like the
smoothest I've seen which I like took as
like a big endorsement of how much we've
like streamlined that model release
process. That just makes it so that like
every release doesn't feel like, you
know, this very, you know, bespoke, very
difficult process. It can be much more a
great like we know what we're doing.
Here's the date. to the extent that
research can be predictable which it
can't be but within that domain uh how
do we actually make that as smooth as
possible
>> right and maybe um I'm looking at this
from a dumb outsers perspective but the
one thing that I didn't hear you mention
was scale and you know hearing so much
about the scaling laws especially from
anthropic you know part of me believe
that like okay four is you know cloud 4
is x number of GPUs and 4.5 is I number
of GPUs and five will be Z number of
GPUs. So does the numbers in your model
release you know um uh rubric correlate
at all with the scale of the data
centers that you're trading on and the
scale of the data. I mean I think what's
been interesting is um at different
points and if you talk to to Jared
Cavlin who's our chief scientist he'll
he'll I think tell you much the same is
um the scaling laws I think paint a
picture of what is possible but is not
predetermined like to actually get there
there's a lot of actually really
difficult both machine learning and
engineering work. So I think one thing
that's been notable to your question
about scale over the last you know 6
months is how much it's been really
engineering like if you're going to do
both pre-training and post-training on
an increasingly large number of uh of
accelerators how do you make that
reliable how do you keep that um you
know how you keep that run as we call it
like going even if you know some portion
of it uh has an issue so a lot of the uh
I think to your question a lot of the
the improvement in our ability to
deliver these models really has come
from our ability to run these large
training runs at scale which you know
again fundamentally an engineering and
machine learning problem I think both
have improved uh I think if I pointed at
something between sonnet 4 and 45 a lot
of it really has been on the engineering
side to just be able to scale up um
especially a lot of the post- training
work
>> if I'm reading you're right it's not
necessarily gains that anthropic is
seeing from scaling up the data centers
it is algorithmic work that is being
done by your teams to make the models
better
>> they really come together I think it's
the algorithmic work and then the
ability to maximize the amount of uh
compute that we can use like on those
algorithmic improvements. So they really
kind of go hand inand sometimes directly
hand inand in that um you know either uh
an idea that works at small scale when
you scale it up doesn't work as well and
then other times an idea only works when
you get enough data and scale in there
as well. So it really becomes, you know,
when we when I think about our our team,
we actually just brought in um a new
CTO. Um and uh a lot of I think his
remit will be how do you really partner
research and our like kind of core
engineering teams together to achieve
that kind of scale.
>> Okay. And another thing I was expecting
you to say, which I'm not sure if I've
heard yet, is that teams within
Anthropic have used the uh coding
capabilities of your AI models to be
able to ship faster. Is that a sort of
supporting character here or is it is it
the star?
>> I think it's um it's a good I'll have to
think about that for a sec. I think it's
a little bit of both. I would actually
say there's there's a thing that is
emergent even beyond the coding
capabilities which is the ability of
Claude to be a really active participant
in the process. And here's what I mean
by that. Um you know I think about the
way Claude was being used around even
sonnet 4. um was uh you know help write
code you know to to launch these models
help write the product code for sure
contribute really strongly to cloud code
you can imagine cloud code itself is
like a very sort of uh we use cloud code
to develop cloud code very much in a
loop I think that the biggest delta
between four and four five is that now
we have much more of uh claude as an
agent or almost like a co-orker in for
example our slack channels so for
example we have um uh something we built
that's clawed on call. So if you've been
an engineer uh one of the things you
have to do is you know you take the
metaphorical pager uh which is basically
you're on call for a week or two to
manage a system and you know if you get
paged um you'll show up and say like all
right there's certain number of things
that could be wrong. I got to go check
these graphs. I got to maybe try this
out. And uh one thing that we've built
using uh the cloud agent SDK, which we
also released alongside its sonet 4.5
publicly, but we've been using
internally for a while, is the ability
for for cloud to basically show up first
in those incident channels and already
have a sense of what might be going on
and be able to answer really quickly,
hey, can you do some data diving while I
work on something else? Um and so we've
increasingly had Claude play these sort
of um yeah, these really collaborative
roles within our company even beyond the
ability to code. And it's again using
the same technology as cloud code under
the hood but it's accelerating the
company in being more efficient or
better able to scale up or better able
to understand it. So I think the answer
to your question is it's support it's a
supporting role on the sort of building
side but it's playing a much more
fundamental role in terms of the actual
operational side. So let me see if I can
zero into it. So instead of basically
being autocomplete for coding this is
actually going out and being proactive
examining things and then coming back
with insights. Exactly. And we have
similar sort of um you know agents is
the I guess the industry term of art
now. But um I feel like agents can mean
so many things to different people right
now.
>> What does agents mean to you if you're
going to if we're going to start talking
about agents? I I need a definition of
this word because I'm struggling to
figure it out.
>> I think the purest definition and this
is not so pure because I'm probably
going to use like 20 words to do it. So
maybe we can edit it down together. But
it's going
>> Yeah. AI systems um that can um plan um
and and and sort of run actions over
long time horizons using a variety of
tools where the kind of steps are not
predetermined. They're able to um solve
problems dynamically based on um what
information emerges on it. So there's,
you know, I I I I end up having this
sort of um agent um uh kind of scorecard
that I've been using internally as we
think about our own products. And
there's a bunch of characteristics that
I look at. This is way more than 20
words, Alex. So, uh, as attributes I
look at are things like autonomy. So,
how long can the can the agent run
unconstrained? So, Sonic 4.5 is a big
leap there. Proactivity, like is the
agent able to not just react to
questions, but actually start to sort of
suggest either ideas or or or interject.
Um, ability to use tools. Um, and often
a variety of tools. Some of them might
be research tools. Some of them might be
uh, you know, being able to write to a
database. Um, memory. So can the agent
sort of learn over time and and improve
its ability to uh to perform a task? I
always say like the hundth task with an
agent should be much better than the
first because that should be the case
for uh human employees as well. Um and
then communication is it showing up in
all the right places, right? And so for
us we think you know these these
entities these agents are going to start
showing up in all places where you do
work whether that's your um uh your
Slack or your teams. For example, we
launched a research preview of cloud in
Chrome. Like we think of cloud, you need
to be in all of these places where
you're doing work so that you can
actually bring it to work rather than
having to bring work to it. So I even
have this like spider, you know, spider
chart of like attributes. So for any
given agent that we're building
internally, we sort of like grade it on
all these different attributes and we
can say, "All right, great. For our the
next quarter, our investment is going to
be on autonomy or it's going to be on
memory." And we can kind of kind of pick
our pick our attributes that we're
working on.
>> No, that that was a good definition of
agents actually. I think that's the most
complete definition I've heard. So
here's like an overriding question
that's coming up as we talk. Is the
improvement that much most of the
improvement that we're going to see at
least in the near term in AI is it just
going to be coming on the like or back
of the orchestration of these models
getting them to be able to take multiple
steps as opposed to I think what was
sort of the defining characteristic of
the earlier days of LM LMS which was
basically just make it bigger make it
generally smarter maybe get some PhDs to
feed some information to it in post
training and then you'll just you know
see what happens as you
I think that there's going to be some uh
fields or disciplines where that sort of
um extremely
sort of precise depth in a particular
task or domain will continue to be
important. Um but I think I'm much more
excited and just like overall I think
we're spending a lot more of our time
even from the product side around that.
I think it's actually two pieces. one is
that orchestration. Um, and then two is
how do you take the work that Claude is
doing from like pretty good to great.
Um, and so, you know, we launched uh
ability for Claude to create Word and
PowerPoint and Excel files that you can
then download and and bring into those
apps. Um, and if you get to like 50% as
good as you would have done yourself, I
don't think that's good enough and it
won't speed you up. And in fact, it's
like I don't know, I could have just
done this myself and now now then at
least I would have known what it's done.
When you start clearing this sort of
like 75 to 80% threshold, of course, is
not scientific, but it's kind of like a
little bit of a vibes based uh thing. Um
then it starts actually being able to
really accelerate work. And so that's
the other emphasis too. It's and that's
interesting that that some of that is
post-trading. Some of that's actually
also giving um a lot of really good
examples to Claude and really working
closely to with how the um model uh is
producing outputs that are what we think
of as like professional quality.
>> Right. And I look I know we're 15
minutes in, so I think we should
probably take a minute to talk about the
concrete things that you've improved uh
in Claude between 4 and 4.5. Do you want
to just give us briefly uh a little bit
of a list of the things that get better
with the new model? Yeah, I think the
the ones that I think are are highlights
um maybe I'll I'll I'll buck it into
three. One is from a price performance
uh perspective. So 4.5 s 4.5 basically
outdoes Opus, our largest model in
effectively every category but does so
while running faster and at a fifth the
cost. So if you think about where we
were in May at you know code with
Claude, we were announcing announcing
Opus 4. you now have a model that is
better than that and even its successor
Opus 4.1 but does so at a flip the cost
which is very like you know opens up a
whole new set of of use cases for that
kind of intelligence that's that's one
on the price performance piece. Um the
second one is um on its ability not just
to to code for for longer but just
execute agentically for longer. We
talked a little bit about agents, but um
what we saw was um actually I put a fun
video of this uh on my ex account, which
is we asked every claude from Claude 1
to Claude, you know, 4.5 to recreate
claw.ai, so like our flagship AI
product. And um 4.5 was really the first
one that was able to do it and to and it
actually produced something of, you
know, quality. It actually works. You
can log in, you can use an API key, all
of those things as well. Um, and so that
ability to like execute agentically,
work for long time horizons. We had one
customer had it work for 30 hours. Of
course, that's not going to be every
task, but like that's the kind of upper
bound that we're starting to see um is
another big improvement. And then the
third one is moving some of those um
post- training wins beyond just code to
other domains we think are really
important. So, for example, um financial
um analysis is an area that we've been
really interested in. And we launched
cloud for financial services a couple of
months ago. Um and we incorporated that
into the model training in Son 4.5 as
well. So when you look at things like
benchmarks like finance agent um
different domains like the legal domain
as well the model is improving not just
on code which is obviously important but
also these other domains that are um you
might actually use code to solve these
challenges but uh the point is not to
write code the point is to solve a
financial analysis for example.
>> Okay. And I definitely want to get into
these various agents in a moment but let
me ask you this. Uh you mentioned that
the new Sonnet 4.5 model uh is more
performant than Opus the big model in
the last release or the four release and
uh and it's cheaper. So how how do you
do that? I think it's I mean we talked a
little bit about scale that's one piece
which is you know that just really being
training Sonnet 45 um on like
significant scale. Um another one is
improvements in the post-training work
that we've done um as well. On the third
third one is um uh really sort of
closing the loop on what we hear from
customers around what are the things
that they wished either Opus or Sonnet
were better at and then getting that
right. So um one we hear all the time is
instruction following like if I tell
Sonnet to do this thing I need it to do
the thing very reliably even if it's AI
even if it tries to be creative like
there's times where you really want it
to uh to be more prescriptive and we put
a bunch of work into instruction
following for for this sonnet too.
>> So I want to talk about these agents.
So, I've got a list of four different
types uh that you highlighted upon
release. Finance, personal assistant
agents, customer support, and deep
research. And I just want to talk about
who they're for. So, the finance agents
are interesting. So, it says you say you
could build agents that can understand
your portfolio and goals as well as help
you evaluate investments by accessing
external APIs. personal assistant
agents. Build agents that can help you
book travel and manage your calendar as
well as schedule appointments, put
together briefs, and more by connecting
your internal data sources and tracking
context across applications. I think to
set these up, it looks like it's a
decent amount of work. Like you'd have
to, for instance, with the finance agent
understand what an API is. So, it's not
going to be something that I think most
people would take off the shelf. So, who
is this set of agents for? And do you
have plans to make this technology more
accessible? So let's say, you know, I'm
a finance, not even a finance
professional. Let's just say I'm someone
that wants to have AI run through my
portfolio. Can I eventually be able to
easily set that up and run it without
having to know any of this fancy tech
stuff?
>> Yeah, I mean that that is absolutely the
goal. So there's agents that we'll build
ourselves and kind of deploy end to end.
And I'll talk a little bit on the
personal assistance side next, but I
think by and large these will be agents
that we can help power for, you know,
companies that have, you know, that
particular domain expertise that they're
bringing it to bear. One of the first
companies I ever worked with at
Anthropic, uh, was intuitit. We were
powering their sort of tax advisory
service. And, you know, Enthropic, we're
never going to build a tax product. Um,
but in it has the largest one. And so,
being able to power their sort of tax
Q&A was really powerful. Now, you can
imagine all these other places, too.
We've um been working more closely with
Microsoft even for some agents even
within um their office suite. So being
able to take the financial analysis
capability and the financial planning
capability and bring it closer to um an
Excel user for example, I think that's
the way you unlock the maximal value um
of some of these as well. And I think
you'll see us sort of demonstrate these
capabilities, but in terms of the first
party products we build, we're pretty
thoughtful about which ones we end up
going deep on cuz to your point, um it's
to reach the scale that we I think
these, you know, products deserve to
reach. Um you want somebody who's really
thinking through the whole endto-end
user experience and probably has some of
the pre-existing connectors already kind
of set up as well. But I think it's
important also to build some of these
ourselves. So um you talked about the
the personal assistant case. One of the
things that we've had a lot of fun with
um on our mobile apps is using ondevice
capabilities as well. And so um I
actually just saw that Apple featured us
today as our you know like uh like new
features like for for Sonic 4.5 and one
of the things that they were fe
featuring was um on on iOS and Android
now. Um it can uh Claude can sort of
read your calendar, read your reminders,
like compose text messages without
really any setup at all. So that's
ideal, right? which is like you got
those pre-existing connectors, you're
not sort of spending a lot of time uh
sort of initializing the the just
getting it set up to even get any work
done as well. But I guess to be more
succinct and answer your question,
there's some that we'll build ourselves
and in those we'll try to, you know, do
our best to sort of simplify the setup
process. But I'm also very excited for
embedding these agents in existing
products that are out there that then
have all that data built in. And so as I
read through your blog post, I also
started to think a little bit about
Dario's prediction about the white
collar blood bath. methods like
impossible not to um where he says you
know within a few years you might see
50% of white collar work uh automated by
uh these AI bots u looking at it being
able to do these finance tasks or
customer support tasks or even be a
personal assistant um I'm just curious
from your perspective as the person
running product here is this something
that you're like merily running towards
trying to automate human work or like
how do you think about it in your role
>> we have um you know kind of like product
principles we try to work kind of
towards um and it's actually interesting
like I think we had very or different
not entirely different but kind of a
different set of product principles even
at Instagram I think it's important to
sort of like figure out what like who
you're building for and how and and and
and how you go about it and um one of
the uh like principles that we operate
you know with is if you can build things
that are complimentary or augmentative
like bias towards those first um And
it's not to say that in the long run
like overall these products like might
not or probably will be doing more um
sort of uh automation or even
replacement of work but um we think that
two things happen if you can build more
augmentative products right um so it's
like not a finance agent that like takes
all the work you know and does it all
for you but it becomes more of a back
and forth one is I think it helps people
develop an intuition of what the AI is
good at today and not good at so that
kind of helps people position even their
own sort of skills um against that. So I
think there's the intuition building. Um
and then the second part is um it I
think extends the timeline by which
people are making that adaptation. So I
think if you see Daario out there
talking about the you know likely labor
impacts, it's not to sort of um uh try
to accelerate towards those but more
around like hey we think this is coming.
Let's start this conversation now. And I
think in the products that we build um
can we sort of show that this is likely
to come but still build a bridge between
here and there by building more
augmentative products. It's definitely a
like a an there's art and science here.
is I think we debate a lot within the
product team as well like had a a great
conversation with our head of design
where he's like if we had a product
where you hit a button and it did all
your work for you that day like would
that be a good product and would that be
like an anthropicy product and we both
came to conclusion like no like one of
our kind of core brand tenants that
we've like come out is like keep
thinking and like we want it to be much
more of this uh collaborative sort of
accelerator of human thought rather than
replacement for human thought um and
would like to keep that the case for as
long as possible.
Yeah, I'm still trying to figure out how
I feel about this stuff. Uh, but I do
think that the conversation around
augmentation versus automation uh is
still like so elementary and honestly
like it's a fairly dumb way uh to look
at. I'm not saying what you're saying is
I'm just saying this the industry's you
know perspective on this like are you
automating or augmenting tasks because
let me give you an example. If you
automate, you know, some if you automate
a job within your company, you've
automated a job. The question is what
happens next? And if you put that person
who was doing that job on something
leading a new project, for instance, or
something higher value, you've now
augmented it in a way that the word
augment doesn't even come close to
describing. So, it's really tough, I
think, to to measure this stuff. And uh
and I don't know, I just sort of feel
conflicted about the way that the
conversation has gone so far. What do
you think? I think that that there's a
lot to what you're saying, which is
there's um there's the uh point in time
task, right? Like, oh, you know,
managing my calendar or, you know, uh
doing some research out about something
that I'm I'm talking about. And then
there's the broader context of what is
the sort of role uh that that person
even has within the company. And you
know a lot of the things that we think
about is um people end up I think people
end up feeling more like managers of AI
than just users of AI. And we think a
lot of it about this even with it's
happening in engineering right where our
best engineers are managing three or
four cloud code instances running at
once. Um and all of a sudden you've like
had to think higher level like right
what is the unit of task that I want
each of these sort of subcloud codes to
be doing. I think the same will be the
case for how we interact with AI systems
and there's going to be some blend of
automation and augmentation there as
well. The the way I think about this uh
sort of the bullc case for this is
twofold. One um can you bring to bear
sort of world expert level thinking of a
particular discipline into companies
that might not have had that before,
right? either because that talent isn't
present in that local market or because
the company's just getting off the
ground and they can't afford a like
worldclass CPO somewhere, you know, or
CTO. Can you like elevate the the kind
of baseline there? So, I think that's
one piece too. Um, and then the second
one is um having companies that will uh
I think emerge and be able to scale and
maintain that sort of small team
cohesion. And I think we did this really
well with Instagram um without having to
like you know build a huge workforce
from day one. And I think the kinds of
companies that get built will change but
I still think there's like a tremendous
amount of economic opportunity
throughout. It just might be you know
more smaller companies rather than fewer
bigger monolithic companies.
>> Interesting. I mean coming from a guy
where you were at Instagram was what 16
people when you sold it for a billion
dollars and people said that was crazy.
>> Exactly. So,
>> I got a question once that was like,
"What when do you think the first single
person billiond dollar company will
emerge?" I was like, "Well, we had 13."
It was like, you know, we were getting
close. We're 13 at at uh sale and 16 at
close. So, basically, it was yeah, just
around then. So, yeah, I mean, we got a
lot done with a little and I think a lot
of that came from focus. Um, and uh, you
know, there's probably work that we
could have done even more efficiently,
>> right? And I mean, I think if you sold
waited a couple years to sell, it might
have been worth double or triple that.
So folks, if you're just uh by the way,
if you don't know about Mike's previous
work, he's the co-founder of Instagram.
So we are going to get to some of the
social media elements of of this or the
the comparisons uh to social media
building in the second half. But uh two
more questions for you as we round out
our first half. Uh you mentioned memory.
I think it's one of the most interesting
parts uh of this work that's sort of I
think underrated and underappreciated in
the common conversation. um can you talk
about how building better memory within
these bots is um how important that is
and how that's actually happening?
>> I think the biggest sort of breakthrough
or really key piece of what we've done
on memory is um rather than treat it as
a sort of substitute for how the model
might otherwise access information or
sort of a system built on top of the
model, we actually have trained it
deeply into the model. And so u the
model knows about the concept of memory
which I know sounds kind of funny but
you can really see it as you talk to it.
And um you can even
>> wait you have to what does that mean?
You have to talk about what that means
model.
>> So basically uh in training we give the
model uh effectively a series of tools
to let it both read from update you know
write memory. And what that means is it
understands the concept that it like uh
is capable of managing its own memory.
And then in our platform, we actually
now have that as a sort of, you know,
basic building block that you can use.
And what that means is as you're talking
to to Claude with access memory tool,
you can say, "Hey, Claude, can you
update your your memory about this?" And
it's it knows what that means. It'll
say, "Great, I'm going to update the
memory." Or when it's uh you performing
an action, if it thinks there's a good
chance that uh it has some memory
related to that, you know, action, it
will retrieve that memory before doing
the action. And previous systems, you
would have to either build that yourself
on top of it, um, or Claude or any of
these systems wouldn't be as good at
using it. And so, um, effectively, in
the same way that, you know, we might
have the thought, hey, I think I think I
did this before or like I think this
happened before. I'm going to go, you
know, either like think about it for a
sec or maybe even search my email. Um,
how do you we've basically given Claude
that same um, uh, that same ability. And
that can be um uh sort of memory that's
very uh like factbased like who are you
interacting with? What should you do?
But it can also be more task based like
whenever I'm doing X make sure I
remember to do Y.
That's pretty amazing. And so what what
will the memory get get you when you're
using this? like better memory will it
start to remember uh many more aspects
or like the so I'll give you one example
and this is so rudimentary but like if I
ever use claude to do a podcast
description I have a format prompt that
I drop in first sentence should be this
second sentence should be this and every
single time I you know write that prompt
uh I I have every time I ask for a
description I have to use that exact
prompt or else it will do whatever it
wants in freelances when are we going to
get to the point where these bots are
going to be smart enough where when I
tell it, remember this is the way that
we do things here, it knows. And I'm
sure that my problem is something that
people have all across the board when
they're trying to get these bots to work
on the same things for them.
>> Yeah. Um, very soon. So, we have a
launch coming up in the next like week
or so that's going to really like uh
there's both memory and then also the
idea of, you know, what are the
repeatable ways in which you want work
to get done. Um, and so we'll have
something really exciting there very
soon. But from the memory perspective,
um, beyond the sort of like very sort of
basic fact-based things like I'm Alex, I
run a podcast and a and a newsletter and
a site and, uh, that's somewhat helpful,
but I think not not sufficient. Like
getting to the point of, um, hey, have I
interacted with this person before? Like
what happens last time I chatted with
Mike? Can I like search over my memories
there? or it can be hey um whenever you
generate these summaries like make sure
that you always you know cite this piece
or lead with a punchier sort of thing
and it's able to sort of update and and
learn over time the results. So that's
the goal is again like um if claud is
like a very competent new hire, we
wanted to get to the point where as you
use it over time either on our platform
or using uh our kind of firstparty
products, it is improving and it just
feels like a companion that you've
actually helped train to your
preferences.
>> Where on the list of priorities is that
capability? It sounds like it's probably
very high for you. I know that it's high
for OpenAI.
>> Yeah, I think it's very it's really high
for us. I think for us it's both it's
high on the first party side but it's
also very high on the on the platform
piece as well.
>> Okay. All right. Let's go to break. I
want to ask you afterwards about what
the moment building AI uh has in common
and differs from uh in building social
media which of course we just mentioned
uh you were right at the center of. So
let's do that right after this. And
we're back here on Big Technology
Podcast with Mike Kger. He is the head
of product at Anthropic and the
co-founder of Instagram. All right,
let's let's talk about social media and
AI. Very interesting. I mean, when we
look around the AI industry, we see so
many uh folks who've come from places
like Facebook and Twitter uh now running
large parts of these AI companies. Of
course, yourself, co-founder of
Instagram, head of product at Anthropic,
Kevin Wheel, uh is running former head
of in well, head of Instagram product as
well, I think, uh is is running product
at OpenAI. Fiji Simo who came from
Facebook is running consumer
applications at OpenAI. I I mean I could
go on. Um what does what does building
these products have in common with
building social media and how does it
differ?
>> I think there's there's maybe the you
know abstracted from the actual product
itself like what does it take to build
good product? And I think that um I
think it's it's less that there's a lot
of social media um sort of oriented
folks that have now moved into an AI.
It's more that I think a lot of the best
product people were focused on that you
know even four years ago you know pre-
chat GPT uh you know pre the emergence
of a lot of these LLMs um so I think
it's a sort of like the most recent
place that concentrate I I find that
that often happens like the
concentration of talent among a
particular uh discipline and I think
that was that was social media
beforehand so that's partly one of them
and and there it's you know um all of
the pieces around understanding uh what
your data is telling you but also having
the intuition around like what bets you
want to place in terms of where you want
to move into next. How do you assemble a
great product team? How does product
engineering and design and marketing
work well together? All these different,
you know, sort of aspects of that. So
that there's that one and then I think
there's the the the separate question of
you know within social media like what
are the similarities and differences. Um
with claude it feels quite different in
that you know we have more of a business
audience like plenty of people use it
for their individual pieces but it has
less of that sort of um you know social
component right now. Um, it's definitely
more word of mouth. Like the the most
social thing that we've experienced is
how people got excited about all the the
merch and the pop-up we just did in New
York where that was like a real like
attractor moment where there was like
more of that. But in general less of the
sort of mechanics of uh of like capital
Growth, right? The you know uh how many
you know people did you bring in, what
who did they invite, all of those
different pieces. Um so maybe a little
bit different there at least for from
the the pieces that we're tackling with
Claude. But of course as a lot of these
uh or non-cloud tools move into more of
this uh generation of like images and
videos like there is much more of an
over a strong overlap with what folks
were were doing on the social media
front.
>> How important is engagement to you? I
mean I think the thing that really drove
Facebook decisions was engagement and of
course growth and maybe the two go hand
in hand. And we always wonder about AI
products. Like of course you want people
to use them, but you don't want
engagement for engagement sake cuz it's
pretty expensive to serve these use
cases. So where does engagement sit for
you in the terms of the metrics that
you're optimizing toward? We don't
really look at engagement, at least not
in the typical like at Instagram, we
spend a lot of time looking at things
like time spent, right? Um we do look at
things like um daily visitors as a proxy
for a utility. So I think that's that's
one piece that we look at as well. Um,
but it's interesting like um I I was
talking to our mobile team yesterday
like I think in the future people's
interactions with something like cloud
code might be much more mobile oriented
and ideally like we're right by
Salesforce Park or office. Like I would
love to be able to kick off some coding
task, go for a walk in Salesforce Park,
maybe with a co-orker. Um maybe it pings
me halfway through and has some
clarification question and get back to
my desk and it's done. Um it's a very
different discipline than being hands-
on keyboard. I also love that. that
feels like a different discipline than
what coding has, you know, evolved to to
primarily being nowadays. And now it's
more about like what are the creative
ideas that I have that I want to see
manifested. But in that world, I the
time spent was quite low, right? It was
maybe like kicking off the task and
resolving some questions, but the value
of what was produced was much higher.
And so I think the interaction paradigm
is just really really different in terms
of what we end up looking at. And so I
think much more about the the sort of
value of work getting done than the sort
of like interaction and and um sort of
uh yeah the long uh long sessions u that
you might see at social media.
>> I legitimately just had a founder that I
interviewed tell me that her favorite
use case is just using AI to get away
from her computer which is something
you've never really heard of before in
technology. So, um, I got to ask you,
what do you think Mark Zuckerberg is
trying to do with his very unique AI
strategy?
>> I I think I there's folks in there that
I've known for a long time like like Nat
who I really respect. So, I think um
what I suspect you'll see is sort of
more experimentation around what like AI
means for this kind of portfolio of
companies. I think the sort of uh
initial wave of well you know we've got
some chatbot type stuff in the search
bars was like not particularly
transformative and I think the teams
there likely know it and so um yeah I
think it'll or maybe what I hope we'll
see is more experimentation that can
kind of live outside of those uh of
those surfaces like in the same way that
with Instagram you know there was some
ideas we had that didn't really belong
in the app like hyperlapse or even
nobody remembers Bolt but Bolt was our
like very very fast messenger um uh you
know I think that experimentation once
you get a service as widespread as
Instagram or as widespread as Facebook
or WhatsApp it's hard to introduce a new
behavior there um you know we did it
with stories I think they've since done
it with reals uh but it's almost like
one you get one per generation and I
think you want to have more of an
experimentation kind of test bed beyond
that and I suspect just like given what
I know about how those folks think that
there'll be more of that sort of
experimentation
>> interesting So, as as the co-founder of
Instagram, I'm sure you've watched with
interest as AI generated images and
videos have filled social media feeds
and even propelled like was Sora the
Sora app to the number one spot on the
app store. Uh, do you think AI generated
content and video maybe in this Cameo
version where you can put yourself in
the videos, do you think it it threatens
or makes a run for replacing the human
generated content that we have today or
uh do you think that the human stuff is
going to stay on top and this is a flash
in the pan?
>> Yeah, I mean I think there's
here's what I'm not sure of yet. We saw
this with Instagram that there were
creative tools that would emerge and of
course these were at a much more um sort
of uh basic level than the kind of
capabilities that you're seeing with um
with VO and with Sora and with even some
of these these other models. Uh but you
would see an emergence of a creative
tool and whether they were able to sort
of transcend that to being a network
that you come back to was often not the
case. Um and I think that was for a
couple of reasons. one is um at least in
that generation of products like the
creative content or the created content
started getting a little bit sy over
time right especially if it was like a
very highly stylized tool um and the
second one it's like the dynamics that
make Instagram Instagram like the people
you already know on there the people
that you follow the creators that you
know and of course this has shifted now
and more of like a uh like pure
algorithmic realsoriented piece or maybe
I'm talking more about like the the the
previous Instagram that was still had a
heavy kind of follow component um ends
up being a thing that feels like, oh, I
know what I'm I know who I'm interacting
with here. And of course, Tik Tok has a
very different take on things. So, to
the extent that it's replacing um I
think the the things that would have to
be true is one that there's the the
content feels varied over time and not
just sort of like, yeah, okay, I've kind
of seen this this before. It's really
interesting, but I've seen it before. Um
and then two, is there value to being uh
in that network over time? And do you
find yourself um uh opening it because
there's like not just content that
you're interested in, but maybe people
that you care about um or there's um uh
sort of communities that form within it.
Cuz I think that's actually what
Instagram got right is that you started
seeing these emergent communities that
maybe were just around taking
photography, maybe they were oriented
around living in a particular city. Um
and they were very self-organizing. The
only tool we gave people was hashtags.
Um and that was enough to sort of spur
the these communities. So, I think
that's like the fundamental uh question
to be to be answered still.
>> Yeah, it's a great point and I think the
cameo aspect where you can put yourself
in the video may go some degree towards
making that happen in these apps. But I
also I'll tell you on Friday uh I
couldn't put Sora down and we're at
Wednesday and uh I don't I don't really
feel compelled to open it right now. I
think you're right that maybe all the AI
content creation can have that level of
sameness to it where you watch one
video, you feel like you've watched them
all and then maybe people come with
creative come up with creative prompts
and you know you see a new trend. But um
I think that's a spot-on uh point there
that that's the challenge.
>> Yeah. Well, uh, well, I mean, I think it
it's it's all happening very quickly in
terms of the the experimentation and so
I think there's also this like um
ability for these tools to adapt as well
and um whether it'll sort of uh open the
door to sort of a new Cambrian explosion
of social products is going to be
another thing that I'm tracking really
well. It feels like it's been very quiet
on the social front for the last couple
of years. because you know we've sort of
like stabilized among you know a couple
of really big players not a lot of new
experimentation and um yeah I I miss the
you know 2010s of you know what if what
if social products were like this and
what if we took this differentiated take
on things and not all most of them
aren't going to work but at least like
there is that value of like hell yeah I
want to try that like that that is a
different experience like you know even
if it's uh again things that feel like
maybe novelties like oh it's a photo app
that takes the front and the back camera
together like is that lasting network.
No, but it painted the way towards
something.
>> Yeah, that was fun. And I I missed that,
too. By the way, that happened, I think,
in the 2020s, but I definitely missed
the 2010s when I was uh you know, doing
social media reporting at BuzzFeed and
there would be a new app every every
week and it was like, "All right, well,
what's Peach? Let's try this out." And
then be gone, but there would be
something new.
>> Oh my god.
>> Peach was classic. All time classic.
>> It was.
>> So, I want to talk to you about about
community briefly. Um, where do you look
to find, I guess, a community of of
users and get your feedback? Um, and how
important is Reddit to you? Because I've
seen so much of the activity in the AI
space move on to Reddit and I'm curious,
are are you reading like our/Sarity or
how deep into it are you?
>> Um, that's a great question. So, it's
interesting being um I would say that
there's like sort of somewhat
overlapping but distinct communities
that we look at. one is like being a
platform, we have um like a strong kind
of customer base that often has a very
sort of cleareyed perspective about
where the models could continue to
improve or what we could be doing better
as a platform. And so um this is very
different than my time at Instagram
where, you know, there were people that
we talked to a bunch about how they
they're using Instagram, but we didn't
have this like more permanent notion of
like an advisory board that we have here
at Anthropic. And we just brought in uh
a couple months ago Paul Smith as our
new chief commercial officer and he's
brought also this sort of community of
more enterprise folks as well that we've
been talking to. So that's one kind of
big delta which is uh like more stable
uh sort of set of people that community
that we're we're talking to. So that's
one. Um we actually have a phenomenal
user experience research team and that's
a place where we end up being able to
stay connected to how um more of the
power users that I think of as like our
core demographic for something like
cloud.ai AI uh are using the product and
I love it. Like every month we do a
product all hands and then my favorite
chunk of that product all hands is
basically the UXR team doing like a
voice of the user piece and it's
surprising sometimes because you know
it's not necessarily who you might
expect being like the software engineer
archetype who for sure are using our
products but there's also hey I'm you
know a marketing manager I need to
produce 20 decks a week and now I
finally found a tool that I'm like using
to to cut it out but here's my uh here's
my feeling about it. Here are my fears
about AI. Here's the promise of AI. So
just a very humanizing uh kind of aspect
of it. Um and then for sure you know I
think still today I think like Twitterx
and Reddit have a like strong pocket of
uh of that AI community and we um you
know I think we've we've gotten better
at um engaging in that community than
before. I think there was a time period
where we were like well like there's a
lot of volume. How do we react? And then
like you don't want to be showing up
only when there's like something that
you want to like correct or something
cuz then it feels very like corpo and
not like authentic. And so I think we
have found a better like ability to
participate uh in some of those those
communities. And you know it's good.
It's they're they're often the like
power users extreme users that are
telling you something about the edge of
what's possible and then you can kind of
try to generalize it more more broadly
too but less like r/s singularity and
maybe more like r/cloudi.
It's very mundane, but it's where a lot
of folks are are hanging out.
>> All right, cool. uh we spoke this whole
conversation we haven't brought up the
fact that well I haven't and and maybe
this oversight on my end that that
openai basically has seen how well
anthropic has done on coding and said
that this is you know a number one
priority for the company and uh I mean
every day you can look at openai leaders
on X speaking of X trumpeting their
codeex product and talking about how uh
how advanced they are on on coding
skills so can you talk about how you
assess openai challenge and what it's
going to be like uh to sort of go
headto-head with them on what has been
anthropic bread and butter.
>> Yeah, I mean I think it's definitely
there was a maybe a window in the summer
where it was surprising to me, I guess,
in general how um uh sort of alone we
were in in sort of both paying attention
and having a product out there. It's
definitely gotten more um uh sort of
interesting and competitive, which I
love. I think that that my favorite
times at Instagram were also like when
we had interesting competitors that we I
think it pushes you forward in terms of
like what like what is the product we
want to build? What are the capabilities
that we're we're going to need to have.
Um so you know it's it's kind of like a
game on an interesting moment as well
for us. The coding piece beyond just the
fact that coding is a really high value
economic activity. I really see the
model's ability to plan, write code,
solve problems as not just being useful
for software engineering, but being
really critical path to the kind of like
agentic behavior we want to build long
term. So there's no way that would never
be anything but like one of our, you
know, top two or three priorities. And
then it's a matter of how do we make
sure that we're showing up with the
right products that like deeply solve
the right problems for people, right?
Like maybe this ties all the way back to
your question about like good product
design and how I think about products.
Um, it's one thing to score well on
software engineering bench. That's
important as like a benchmark, but it's
way more important, I think, to get the
feedback from people like great, those
are really hard tasks that I was doing
with Rust and in set 4 I couldn't do it.
Open four could barely do it. Sonet 4.5
can do it. Like that I get very excited
about because it means we're actually
having uh like real world impact. So I
think you'll see us um if we're doing
our jobs right on coding uh you know
even in the in the presence of other
players enter the space try to stay
really focused on um listening to how
people are using these products in the
wild and then uh ensuring that future
model versions are sort of meeting
people where they are in that you know
high utility space.
>> All right last one for you Mike. Uh
enterprises they're all interested in
generative AI they're not great at
implementing it. They'll admit it the
studies show it. Are they going to get
it together?
>> I think they will. I'm actually, you
know, we're we're um from from after our
our our conversation, I'm I'm having a
off-site with our product team and a lot
of the focus for next year is continue
to to go into the enterprise side of
things. And I think there's a few
things. There's the um and we could
probably do another whole like hour on
this. I get very excited about this as
well. There's a whole range from how do
you take a product like cloud for
enterprise that you know enterprises are
already adopting, but make it really
really useful. And we talked a little
bit before about like output quality and
just like how um how much it's actually
helping you. And like there's I think
part of the valley of disillusionment or
the trough of disillusionment that you
might be seeing around enterprise AI
adoption is the promise of these tools
around they're going to save you time,
they're going to like make like make
your work better just wasn't fulfilled
by the previous generation of products
and they need to be if we're going to
actually get like sticky adoption in the
enterprise, right? And so that's a lot
of what we're pushing on is like it's
not like it's not like AI produced
document slop. It's AI produced quality
stuff that you can then uh iterate on
and use and feel proud that you created
just like I think people can feel proud
about like here I built this thing using
using cloud on the coding side and then
there's all the way to you know beyond
the cloud for enterprise piece like
deeper integrations like internal
transformation and what we're learning
there to your question about how
enterprises are are thinking about and
adopting is um at least for the
foreseeable future we need to lean in
much more in terms of uh helping
enterprises get there and so we're doing
much more of model now where either uh
with our own engineers embedded in
enterprises in partnership with Deote
which we just announced this week can we
actually like take our technology meet
companies where they like what their
highest needs are and then just
co-develop and just you know lock
ourselves in the building until we've
solved their problem and then like learn
from that experience and move on to the
next enterprise. But I think uh it's
very different than sort of the lean
back sort of like we're just going to
have enterprise products and hope that
enterprises figure it out. I don't think
that that's the the reality. We just
need to lean in way harder on on both
ends of that spectrum.
>> Bud.ai is the website. Mike Kger is the
chief product officer at Anthropic.
Mike, always great to speak with you.
Thanks for coming on the show.
>> Thanks for having me, Alex. All right,
everybody. Thank you so much for
listening and watching. We will see you
on Friday to break down the week's news
and we will see you then on Big
Technology