OpenAI researcher on why soft skills are the future of work | Karina Nguyen (Research at OpenAI, ex-Anthropic)

Transcript from https://podmenti.com/t/e43e0066df5b05c9

Not only are you working at the cutting edge of AI and LMs, you're actually building the cutting edge. When I first came to Andarbine, I was like, Oh no, I really love front and engineering. And then the reason why I switched to research is because I realized, oh my god, Cloud is getting better at front ends. Cloud is getting better at like coding. I think Cobb can like
What skills do you think will be most valuable?
going forward for product teams in particular. Creative, thinking. And you kind of want to like generate a bunch of ideas and like filter through them and just build the best product experience. I think it's actually really, really hard to teach the model how to be aesthetic, a really good visual design, or like how to be extremely creative.
in the way they write. What do you think people m most misunderstand about how models are created? When you taught the model some of the self-knowledge of you actually don't have a physical to operate in the physical world. The model would get like extremely confused.

Today my guest is Karina Nguyen.
Karina is an AI researcher at OpenAI, where she helped build Canvas, tasks, the O1 chain of thought model, and more.
Prior to OpenAI, she was at Anthropic, where she led work on post training and evaluation for the Clot 3 models, built a document upload feature with 100k context windows, and so much more. She was also an engineer at New York Times, was a designer at Dropbox and at Square,
It's very rare to get a glimpse into how someone working on the bleeding edge of AI and LLMs operates, and how they think about where things are heading.
In our conversation, we talk about how teams at OpenAI operate and build product, what skills she thinks you should be building as AI gets smarter, how models are created, why synthetic data will allow models to keep getting smarter, and why she moved from engineering to research after realizing how good LMs are gonna be at coding.
If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. It's the best way to avoid missing future episodes, and it helps the podcast tremendously.
With that, I bring you
Parina.
Nueva.
This episode is brought to you by Interpret. Interpret unifies all your customer interactions, from gone calls to Zendesk tickets, to Twitter threads, to app store reviews.
And it makes it available for analysis. It's trusted by leading product orgs like Canva, Notion, Loom, Linear, Monday.com, and Straba.
To bring the voice of the customer into the product development process, helping you build best in class products faster. What makes Interpret special is its ability to build and update customer specific AI models that provide the most granular and accurate insights into your business.
Connect customer insights to revenue and operational data in your CRM or data warehouse to map the business impact of each customer need and prioritize confidently.
and empower your entire team to easily take action on use cases like win loss analysis.
Critical bug detection, and identifying drivers of churn, with interpret's AI assistant wisdom.
Looking to automate your feedback loops and prioritize your roadmap with confidence, like Notion, Canva, and Linear, visit ENTERPRET.com slash Lenny to connect with the team and to get two free months.
When you sign up for an annual plan. This is a limited time offer. That's interpret.com slash Lenny.
This episode is brought to you by Vanta, and I am very excited to have Christina Cassiope, CEO and co-founder of Vanta, joining me for this.
Very short conversation.
Great to be here. Big fan of the podcast and the newsletter. Vanta is a longtime sponsor of the show.
But for some of our newer listeners
What does Vanta do and who is it for?
Sure.
So we started Vanta in 2018 focused on founders.
Helping them start to build out their security programs and get credit for all of that hard security work with compliance certifications like
Stock two or ISO twenty seven oh one.
Today, we currently help over 9,000 companies, including some startup household names like Atlassian, Ramp, and Langchain, start and scale their security programs, and ultimately build trust by automating compliance, centralizing GRC, and accelerating security reviews.
That is awesome. I know from experience that these things take a lot of time and a lot of resources.
And nobody wants to spend time doing this.
That is very much our experience, but before the company to some extent during it. But the idea is with automation, with AI, with software, we are helping customers build trust with prospects and customers in an efficient way.
And you know our joke, we started this compliance company, so you don't have to.
We appreciate you for doing that, and you have a special discount for listeners, they can get$1,000 off Vamta.
at banta.com slash Lenny, that's V A N T A dot com slash Lenny.
For one thousand dollars off Anta.
Thanks for that, Christina.
Thank you.

Karina, thank you so much for being here. Welcome to the podcast.
Thank you so much, Slenny, for inviting me.
I'm very excited to have you here because
Not only are you working at the cutting edge of AI and LMs, you're actually
Building the cutting edge of
AI and LMs. We recently launched this feature which basically
uh the first agent feature of open AI.
I also just did this survey. I don't know if you know about this. I did this I did a survey with my readers and asked them what tools do you use every day in your work and most use.
And chat GPT was number one above Gmail, above Slack, above anything else. Ninety percent of people said they use Chat GPT regularly.
It's so it's absurd. It wasn't around two years ago.
Yeah.
Uh also we're recording this the week that OpenAI announced Stargate, which is this half trillion dollar investment in AI infrastructure. So there's just like a lot happening.
uh constantly in AI and you have a really unique glimpse into
How things are working, where things are going, how thing how work gets done.
So
I have a lot of questions for you I want to talk about.
how you operate and how you work at OpenAI.
Where do you think things are going? What skills are gonna matter more and less in the future.
And also just where things are going broadly. So how does that sound?
Sounds great. Thank you so much. Um
Yeah.
I was extremely lucky.
To join early days on topic and kinda learned a lot of things.
Uh there and and I joined OpenAI around like eight months ago. So yeah, I'm excited to have more.
Okay, I'm gonna definitely ask you about the differences between those, but I wanna start
More technical.
And just dive right in. I want to talk about model training.
People always hear about models being trained, this these big models, how much data it takes, how long it takes, how much money it tosses.
It takes how uh how we're
Running out of data, which I want to talk about.
Let me just ask you this question. What do you think people
most misunderstand about how models are created.
Moral training is more an art than a science.
And in a lot of
Ways like V S like model trainers.
Think a lot about like
data quality is I like it's one of the most important things in model team is like
Uh, how do you ensure
the highest quality data for certain like interaction
uh model behavior that you want to create. But the way you debug
models is actually very similar the way you debug software.
Um so one of the things
that I've learned early days and ontopic was like
We've discovered, especially with like
Clock three training.
When you taught the model some of the self knowledge
of like, hey, like, you actually don't have a physical body to operate like in the physical world.
But then at the same time we had theta
They kinda taught the model.
Um some of the function calls.
Which is like
This is how you set the alarm.
As a model would get like extremely confused.
Uh about like
Whether
It can set an alarm in the f but it doesn't have a body in the physical b uh role. So it's like the model gets confused and sometimes it'll like overrefuse. So sometimes it says like I don't know like uh sorry, I cannot help you.
And so there is always like a balanced
trade off between
Uh, how do you make the model to be more helpful for users, but also
Um not being harmful.
uh in other scenarios. So it's always about like
How do you make the model like more robust and like
operate across like variety of diverse scenarios.
That is so funny. I never thought about that. Most of the data that it's trained on is kind of like assuming
It's like a human describing the world and how they operate and there's
Well assumes there's a body and you could do things and the model told you don't have a body.
Yeah. Uh
Okay.
I wanna talk a little bit about data.
While we're on this topic, I know you have strong opinions here, there's kinda this meme that
Models are gonna stop getting smarter because they're running out of data.
They're trained in um
large part on the internet.
And there's only one internet.
And they've already been trained on it. What more can you
Show them about the world.
And there's this trend of synthetic data, this term synthetic data.
What is synthetic data? Why do you think that's important? Do you think it's gonna work?
I think there are two questions here. Um we can unpack.
Or one other time.
But uh people say we're hitting the data wall.
Uh, I think people think more in the terms of like
pre trained large models that are trained on the bi well on the entire internet.
To protect the next
Talking.
But what actually the model is learning during that process is actually
How do you compress?
The compression algorithm here. The model learns to compress a lot of knowledge.
And it learns how to model the world.
Um
So the next prediction of the word like
Teach me how to
Drive.
Basically, and you only have like
a few words that will match that, a car. So the model actually learns
Um
about the world a in in itself.
So it's like it's modeling human behavior sometimes. It's modeling and when you talk to like pre train models.
Which are very, very large. They're actually extremely
diverse and extremely creative.
Because you can
Talk to almost any Reddit user through Progene model.
But I think what's happening right now is like new uh paradigm of like oh one serious is that like
The scaling
And the chaining itself.
Is not hitting the wall.
And that's because
Basically we we went from like
Raw data sets.
From from featuring models.
To
Infinite amount of tasks.
That you can teach the model in the post training world.
via reinforcement learning.
So any task, for example, like
How to search the web.
how to use the computer, how to write uh well, like
All sorts of tasks that you like trying to teach the model.
All the different skills.
And that's why I'm be saying like there's no
data wall or whatever, because there will be infinite
amount of tasks. And that's how the model becomes extremely super intelligent.
And we are absolutely getting saturated on all benchmarks.
So I think the bottleneck is actually in evaluations. Uh, but you don't have
Um
All the frontier
Like Eva's like
I don't know, um, GPTA.
Which is like
A Google proof.
Fashion answering like PhD level.
And Dollar J the benchmark is like getting to like, I don't know, more than like
Sixty, seventy percent, which is what G H D gets.
Um so it's like they're h literally hitting the wall and like evolve.
I wanna follow both those threads. So the first is on this idea of synthetic data.
Is a simple way to understand it that the models are generating the data.
That
future models are trained on.
And you ask it to generate all these
Ways of doing stuff, all these tasks as you described, and then
the newer models trained on this data that the previous generated.
Some tasks are synthetically curated.
So this is like an active like research area is like how do can you constru uh synthetically construct like
new tasks model to like learn.
Sometimes, you know, like when you develop products
You get a lot of like
data from the product and like user feedback and you can use that data too in like uh this like post-changing world.
Um
Sometimes you still want to like use like human
Um
human data because
Uh actually s some of the tasks can be like really, really hard.
uh to a teach. Um
Like like like experts like
Only no like sort of knowledge about like some chemicals.
like biological knowledge, so like you actually need to tap into
uh the expert uh knowledge a lot so
Yeah, I think
To me like synthetic data training is more um
Well like
Product is like a rapid model iteration for similar product outcomes.
And
more into but uh
The way we made canvas and tasks and like new like product features for G Biki was
Mostly done with by synthetic chaining.
Let's actually get into that. That's really interesting. I want to talk about evals, but let's follow that thread. So talk about how this helped you create Canvas.
So when the first came
to open AI I really
How this idea of like
Okay, like it would be real cool.
for Chapity to
Actually like change the
Visual interface but also sh
Change like the way
It is with people, so
Going from like
In a chad boss.
to more of a collaborative agent.
And the collaborator.
It is like a is like a stop towards like
Мой чай такси сенс.
Um they become like innovators ultimately.
And so the the entire team of like applied engineers, designers, product
Like research kind of like God.
Like formed.
uh in the air. Oh almost out of like nothing. It's just like a collection of people who just like got together.
And we rapidly started each waiting with each other.
Actually like caves is like like one of the
I would say like the first project of Open AI were researchers and applied engineers
started working together from the very beginning of the
Product development cycle.
And I I think like there's
A lot of things that we have learned.
On the way?
But um
I definitely came to with the mindset of like we need to do like a really rapid model iteration such that like it would be much easier
Four.
Engineers
to you know ripe with the latest model.
possible, but also learn from like
User feedback or like early like internal dog food.
uh how do we improve the model very rapidly.
And
Yeah, it's really hard to like um
kind of like figure out like how people when you deploy a product how people would be able to like use it.
And so like
The way
you synthetically train the model is basically figuring out like what are the most core behaviors
That you want that this product feature.
To do.
And for Canvas, for example
Uh it was
If Game Donzo likes to be mean.
He versus it was
How do you trigger canvas?
Full promise like
Write me a long essay.
when the user intention is mostly like
Iterating of our long documents.
Or write me a piece of code.
Or when to not trigger canvas for prompts like
Can you tell me more about
President like
I don't know. Um so some of the general questions. So you don't want to like trigger campus because the user intention is mostly getting answered, not necessarily like iterate uh always a long document.
The second behavior is
Um
How do you how do we teach the model to update the documents when the user ask?
So one of the behaviors of the
Uh the model is actually
have like the a t some agency and autonomy to
literally go to the document and like select s uh specific sections and
Either deleted or edited. So highlighted and rewrite certain sections.
So sometimes the model
Sometimes the user would just like say
Change the second paragraph to be something friendlier.
And you would have to like teach the model to
Literally find the second paragraph in the document and change it.
To a friendly tone. So basically you teach
Both like how to trigger like
Uh edit itself.
But also how do you teach the model to get higher quality at it?
uh for the document.
In case of like coding, for example.
Uh there's also like the question of like how good the bottle is that like
completely rewriting the document.
versus like v having like very specific targeted ads. So that's like another like
layer of decision boundary within like edit itself.
It's like select.
the entire document then like rewrite completely, or you want to like have like very talking custom keyboard. And you know, like when we first launched the model,
We would bias the model towards like more rewrites?
Because we saw the quality of the much higher.
But over time you're like kinda shifting based on like user feedback and what they're learning from iterative deployment.
Lastly, this the third behavior that we taught synthetically the model is how to make comments on any document.
So
The way we use that is like
We would use a one model.
to produce
to like seem away of like music conversation. Let's say like
Write me a document about XYZ.
But then we used O One to like produce the document.
And then the
kind of injective like user prompt to be like
Oh
make some comments, critique my piece of writing.
uh or critique this piece of writing that you just named
Um
And then the tell the model to like make comments on the document.
On like very specific talking to the document. And so it's like also like what kind of comments
you want the model to make, like do do they make sense or not? Like how do you teach
The quality of that, um
And
It all came down to like measuring progress.
Why very robust evolves.
But yeah, this is how you like used like a wan and the like kinda like synthetic data condition for like the staining. Okay. This is so interesting.
Uh, so you talk about this idea of teaching the model and you mention how it's
Using synthetic data to teach the model different behaviors.
Is a simple way to think about it. Basically that's where you you do that by uh showing it what success looks like using basically evals. Is that the simple way to think about it? Like
Here's what
you doing this successfully would look like and that teaches it Okay, I see this is what I should do. Yeah. Amazing. Yeah, you got it. Okay, got it.
Um I wanna start unpacking what your day to day looks like as you're building these sort of things. Is it like you sitting there
Uh talking to
Some version of chat GPT.
Uh crafting these evals.
Sometimes I do that, but sometimes I do set I think I learned this so much.
From
It's like
People spend so much time just like prompting models and like quality delivery by bash all the time, and you actually get
A lot of new ideas.
Oh, how do you make the model?
uh better. It was like oh like this is this response was kinda weird. Like why is it doing this? And you start like debugging or something or like
You you start like thinking out like new methods or like how do you teach?
the model to respond in a different way, like have better personality, let's say.
So it's the same thing of like
How
Personality.
mm is made like in the models of the goals?
Like very similar methods.
But yes, I
I think my time
I don't think I have changed. I think when they first came, I was like
mostly like research I see work.
Uh so I was like
Building a lot of like
I was like writing codes, like you know, changing models.
Writing evolves, working with PMs and like designers to like
learn teach them how to like even think about like evaluations.
I think it was like
Really cool experience.
And I think it was just like an a a
an adoption of like
How do we like do this like
product management of like yeah features or like um
AI models. Um
Yeah, but now it's like mostly like
you know, like management and like
mentorship, um, I'm still like
doing I see like research codes.
after like four PM although but um
Yeah, I just kinda like changed.
All right, don't talk too much about being a manager'cause everyone's firing their managers. Who needs managers anymore? That's the what I hear now.
Just kidding.
It's interesting that so much of your time
was spent on teaching product teams how evals integrate and how important
That is and I've heard this a few times and I
Haven't personally experienced it yet, so I think it's an important thread to follow is just
How writing
These evaluations is gonna become
increasingly an important part of the job.
of product teams, especially when they're building AI features and working well on
So
Can you just talk a bit more about what that looks like? Is it like sitting there with an Excel spreadsheet?
Basically showing like here's the input, here's the output, here's how good the
result was. Talk about what that actually looks like very practically. It certainly depends a little on the wood you're developing.
But uh there are various types of like immunities. So sometimes I do ask
product managers or uh there's also like new role that we have like model designers.
to um
Kinda like go through some of the user feedback maybe or like think of like various like user conversations.
That should have triggered
Like under this circumstances, it should trigger canvas.
And then you have this like ground truth label of like, okay, with this conversation it should look trigger candles, and with this conversation it should not trigger candless. And you have this like very bind deterministic kind of like evolution.
That for like
This is modern behaviors is like this.
uh when we were launching tasks, for example, like
How do you make correct schedules?
is like actually really hard for the model.
But we both out like some of the deterministic evaluations.
Then it's like, okay, like if the USA says like seven PM it's like
the model should say seven PM. So if you can like have a defeministic evaluation, whether it's like
Pass or fail.
Um so yeah, and like the way it dresses, I was like
Sometimes I ask
Pragmatist just like
go create like a rule sheet, like have different tabs and like
Um
What's the current behavior, what's like the ideal behavior.
And like why like some notes.
And sometimes we usually use it for eval, sometimes uh we use it for training.
Because like if you give the spreadsheet to like a one model, it can probably figure out like how to
uh teach itself.
A good behavior.
And I think there are second type of like evolves.
There is
kind of more prevalent is like uh human dominations.
And you can have
Specific trainers, or you can have like internal people.
to um
When when you have like a conversation of the prompt and then you have like
various completion of models, you kinda choose the win rate.
Which model is the best?
Which model produced the highest quality?
comment or
Edit.
And then you can have like continuous win rates.
And as you develop new models, it should always like win over the
previous models. So
Um it depends on what you want to measure.
So interesting. Like basically what I'm hearing in
This is something I'm learning about as I talk to people.
Is product development start might move from this like
Here's a spec PRD. Let's build it together.
And then cool, let's review it. Are we happy with this two from that to
Hey, AI
Build this thing for me.
And here's what correct looks like, and I'm spending all my time on what does correct look like.
Any valves essentially.
You definitely want to like
measure progress uh the model and this is where evasive is because like you can have
Prom Third's model as a baseline.
Already and of the most robust evolves is the one where
Pronto baselines uh
get the lowest score or something. And then because then you know like, oh if you're trained a good model, then it should like just like hill climb and that you out.
all the time while not like also like regressing on like other intelligence evolves. So it's like I think it's more what that's
That's what I'm saying, like it's more of an R than a science. It's like, okay, like if you optimize the model for this behavior, like you kinda don't want to like bring damage in like other areas of intelligence or
This is happening like
all the time in every lab and every like research team. Um
I would say
Like prompting is like also a way to like prototype like new
On the good news?
Oh, the early days of Andar wina.
was working like file uploads feature.
I remember I was just like
Yeah, prompting the model.
Um
To just like
I mean when we were like launching like Honda Ky contacts.
I was just like prototyping this in the local local parson where sh I did the demo and like people really, really loved it.
And they just like wanted like API for like file uploads or something. And then that's when it clicked to me like
I also like run the blog post.
On temphe, like it clearly to me like prompting is like a new way of like
product development or prototyping for designers.
And for like privilegers.
For example, one of the features that I want to do is like have like personalized
uh recommended personalized starter prompts.
So when ever you come to like Cloud.
Like
It should like
recommend you like
starter prompts based on what your interests are.
And so like you can
Literally do it like
Haunting.
For that.
Another feature was like
generating titles for the conversations.
It's a s very small like micro experience, but I'm really proud of
Uh
The way we did that was because we we took like five later conversation from the user and like asked the model like what's the style of the user.
And then like for the next kind of new conversation, the generated title will be
of the same like style.
Micro experience is like those.
Oh, that's so cool.
Did you do that at Ethropic or at OpenAI?
One topic.
Okay, cool.
I love the file upload feature that Claude has, by the way. Oh chat GPT doesn't have that yet, is that right?
Um, I think it has. I think like the way it's implemented is like very different though. Okay, maybe it's the PDF feature because I use it all the time with colour. Okay. That's cool. Someone needs to get on that.
Uh man, it's wild how many features you built that I use every day and that many people use every day.
This prototyping point you made is really important. It's something that comes up a ton on this podcast also of how that
is maybe the way that AI has most impacted the job of product builders.
Recently is just prototyping.
Instead of going from
Showing just like here's a PRD, here's a design.
PM's more and more just here's the prototype with the idea that I have and it's working. You can play with it.
Yeah.
Yeah.
Okay.
I wanna spend a little more time on how you operate. So you talked about
You built this in launch this task.
Feel is that is that the way you describe your tasks?
Yeah.
So talk about how that emerged and let's better understand just how you collaborate with product teams and how open A works in that way. Whatever you can share there.
I think Canvas and tasks are
Uh going into the bucket of
Ducks where it's like more like
Short like medium terms?
And um
Actually the way cameras and tasks

About to be it was like
It started was like one person prototyping.
And
Creating like
A spec?
It's kinda like PRD. It's like creating a spec of like the behavior of the model.
I don't think like
Tasks is like extremely like
Grand baby.
B ground breaking feature necessarily.
What makes it like
Really cool is
Because the models are so general
model can now search, they can like write sci fi stories, they can like
Search for stocks they can like summarize your news every day.
Because the models are so general
Like giving something familiar to people that like, you know, notifications like very familiar. Like
having reminders is like very familiar. So like
Feeling like a
Form Focta for the people who like
Very familiar. Same as like Camus Moses like a bulldogs are very familiar.
Oh, but then you add like magical ammo and then it's like it becomes like very powerful.
But the way comes like operationally, like
Yeah, it says this like a prototype, like literally prompted.
prototype or like how you would want like the model to behave. For like tasks, for example, like
You kinda like need to design.
A little bit like design design systems design thinking is like
Okay, like well
Is the more is this the user says like um
Remind me to go to lunch like
eight a M tomorrow.
Okay, what kind of information does a model need to
extract from that prompt in order to create a reminder.
And so this is how you like
Like design like a spuck.
For um a new feature.
Like a tool.
Canvas and tasks are all tools. So it's like how do you like create the tool stock? And then it's like mostly mostly like like
Uh developing JSON schema is like
Okay, like from this problem, maybe the model should extract like
The time
To the user request it.
And then you're thinking about like which which form right do you want the time to be?
And then like
How do you
One the model to like
Not a five you is like
Basically
I the user should give instruction to the model.
Uh, and then this instructions would like fire off like every day or something at the that particular time.
Uh, so for example if you say like
Search like every day I want to like learn
No about the um
Latest AI news.
Um
The model should derive into like
Okay, like search for the latest AI news.
And this will will this task will get fired at that go type that the model that the user requested.
And then you know, your design was like Tull Spark and then
Actually I don't know, like I feel like sometimes like
It's like through conversations I
Like
I thought like people asked me to like join the listen like
Genuinely like, Oh my god, like we need to be so charged. And we need like some support, like
You've like to train the models, or sometimes like Tango was like most of like
I just pitched the idea of like
It got staffed quite immediately during the break.
Um, so I feel like it it's like depending on the project. And usually with staffing it's like mostly like a product manager.
Um
model designer um
Actual product designer.
A couple of researchers and buy to like applied engineers.
Depends on the complexity of the project.
And then like
Yeah, it takes it to for tasks it took like.
I don't like
Two months or so?
To go from like zero to one, basically. Oh wow. Um
For canvases was like
Four five months, I guess. Um
To go find out the each one.
But uh yeah, and then like you know, you teach
Part of mine is just how to like build evals and like maybe
You know.
How how do we
Not really like ship.
uh the better feature, but how do we think like more longer term? Like what kind of cool features that you want tasks to have? Like I think it would be nice for tasks to be like extreme a little bit more personalized
It'd be nice to have like
to create tasks via voice in on a mobile, right? Like so you kinda need to like this is how you get like research mo roadmap right here is like thinking like how the feature will be developed in the future.
And then from there on it's like

You like start?
Creating data sets like
Uh with e was you want to make sure that goes
Well, and then like
You
Mm, need to have like a trade off between like what methods you want to use.
And the reason why I really love like synthetic, like relying purely and synthetic data instead of like collecting.
data from humans is because it's like much more scalable.
It's cheap, let's have like you literally sample from the model.
And you teach the core behaviors of the model, then that will generalize um to all sorts of
the the worse coverage.
And when you launch the better feature, you learn so much from the users.
That you can like
All your synthetic.
That's
Can be
can be shifted in the distribution of how the users behave in the on the private behavior, and this is how you improve.
Uh this is what happens kind of stuff.
When we learn from better to
Sh.
This episode is brought to you by Loom.
Loom lets you record your screen, your camera.
And your voice to share video messages easily. Record a loom and send it out with just a link to gather feedback, add context, or share an update.
So now you can delete that novel linked email that you were writing. Instead, you can record your screen and share your message faster.
Loom can help you have fewer meetings and make the meetings that you do have much more productive.
Meetings start with everyone on the same page and end early.
Problem solved.
Time saved.
We know that everyone isn't a one-take wonder when it comes to recording videos. Saloon comes with easy editing and AI features to help you record once and get back to the work that counts.
Save time.
Alig your team, stay connected, and get more done with Loom. Now part of Atlassian, the makers of Jira. Try a Loom for free today.
at loom.com slash Lenny
That's L O O M
dot com slash Lenny.
Something that I wanna help people understand and I don't even a hundred percent understand this is what's the simplest way to understand the job of a researcher?
versus, say, a model designer and other folks involved. Like what's the simplest way to understand what researchers
do at open air.
So the project that I described that mostly like product oriented, like research is mostly like product research.
Another part component of my team is actually more like longer term exploratory projects.
And it's more about like
Developing new methods.
Understanding those methods.
Under variety of circumstances.
So
Ли бисикли девалні мати
You kinda like me to follow
very similar kind of like recipe of like building e balls, but it's much more sophisticated evaluation, like
you kinda want to have like outer distribution or like if you want to like measure generalization
Um
If I don't need to like capture thoughts.
Uh, but it's basically more science y in a way where
You know, it's it's the
Talk about synthetic data. Like one of the hardest things about synthetic data is like how do you make it like more diverse?
Diversity and subject data is like one of the most important questions. Uh right now. And so it's like
exploring like ways to ench like diversity
as a general method that will work for all is like a one of the like research explorations.
other ones is like more like about making new capabilities. I feel like it's all just about like you know, like
You
You
Work on this like new method.
And you have like signs of life that it's working.
I did just think of like how do you make it more general
Or you think of like
How do you
make it very useful or like
And this is how like longer term projects become more like medium like short term projects. That makes sense. Essentially working on developing ways to make the model smarter, oh four, or five, oh six.
Yeah.
Like O one was a big breakthrough, right? The way it um operates where it's not just here's your answer, it actually
Things and has
Right. Takes time to think through the process of coming up with an answer. Okay. Yeah.
Very helpful.
Speaking of that, of thinking about the future, where things are going.
I wanna spend some time on
Just this. I
insight that basically you are building the cutting edge of AI. Like at the very bleeding edge of where
AI is going and where it is.
And so
Uh I'm very curious to hear just your
Take on
How you think things are gonna change?
In the world.
And how people work.
Based on where you see things are going and I know it's a broad question, but let's say like in the next three years.
How do you see the world changing? How do you see people's way of working changing?
It's a very humbling experience to be in both labs, I guess, like
To me when they first came to and daring I was like
Oh no, I really love from then engineering. And then like the reason why I switched to like research
It's because I realized at that time is like, oh my god, like Cloth is getting better at like front ends. Like Cloth is getting better at like coding. I think Cloth can like
develop new apps or something. And so like it can like
There are new features for the thing that I'm working.
It was kinda like this meta realization where it's like
Oh my god, like
The world is actually changing and they're like
When the first like
launched Hundred K context at that time.
Obviously
You know, I'm thinking about like
From talk there, so it's like
Yeah, like
File uploads were like very natural, very familiar to people. But you could imagine we could just like make like
Infinite chats in the
called that AI app, right? Like as if like it's like a ha in a hundred key context. But because like file uploads
It's like
for and follow his function, it's like
The form factor, the file uploads kinda enable people to just like
literally upload anything, the books, like any reports financial, and like ask any task to the model.
And then I remember it was like
Yeah, enterprise customers like
Um
like financial customers are like really interested in that as like
Oh wow, like actually the
It's actually one of the
very common task that people do.
Uh
In that setting it was like
Kinda crazy to like see uh how some of the redundant tasks are getting like
Oh, to me, basically. Buy this like smart models.
And we're entering the the era where
I actually don't know, for example, sometimes like if a one
Gives me the correct answer.
or not because I'm not an expert in that field.
And it's like
I don't even know how to verify the outputs.
Um
Other models is because like
Oh, my experts know in life. They can like verify those.
So
Yes. So basically there are
Trends that were going on.
The first trend is the cost of reasoning.
and intelligence.
Is drastically going down.
I had a blog post about this.
Maybe I should update them like
Latest benchmarks because at that time like MMO everybody was like
Do I um
Not one like one benchmark and then be like quickly saturated the benchmarks and like now be you need to like do the same flop but was with another like frontier evil.
But the cost of intelligence is like going down because it's it becomes like much cheaper.
S smart small models are becoming
It was smarter than like large models. And that's because of like
The
Distillation research.
This happened was like Claude High Cool.
I was like working on like post annual clothes high cool.
And I realized it was much smarter than like Claudio, which was like
way, you know, bigger or something like that. Um, but like the power of like small models become very intelligent.
And fast and cheap.
We are moving towards the world.
Dale
has like multiple implications
But
The news that like
People who have more access to AI, and that's really good. Like builders and developers.
will have much better access
Two
AI
But also it means like all the vertical that has been like bottlenecked by the intelligence.
will be kind of like unblocked.
So
Anyone
Like in I'm thinking about like healthcare, right? Like
If I have
Instead of going to the doctor, I can like
Аск чачі пітії лаків чачі пітії
A list of symptoms and ask me like
Uh which
Like
would I have like a cold flu or like something else like
I can literally got
The access to like
Uh
doctor almost and there's like been some like research studies around that.
Yeah, there was a New York Times story about that where they compared doctors
Two doctors using chat GPT to just chat GPT and
Chess Chat GPT was the best.
Yeah. Like doctors made it worse. Yeah.
Yeah, that's crazy, like right, like education, I think
Uh I will have drands if
Like I had the tool like ChatchPT and when I was like young and like
would learn so much, but it's like
People can now learn almost anything.
From these models so they can learn
New language. They can learn how to build new
look ups like I don't know, anything that you want in like
And so
Oh, like it's humbling to like have like launch canvas and like bring that thing to the people, enable them to do something else that they couldn't have
ever before and I think this is there's something like magical around this experience.
Uh so education has will have massive implications like
I guess like scientific research, right? Like I I think it's like the dream of like any AI research is like upmate AI research.
Uh it's kinda scary, I'd say. Um
Which makes me think that like people management
Well stay, you know, it's like
One of the hardest things to is like emotional intelligence.
for the models or like creative we creativity in itself is like one of the hardest things. So
Writers I I don't think like
People should be worried as much. I think it's like
I think it's L Eviate for a lot of like redundant tasks.
Um for people.
This is awesome. Okay, I wanna follow this story for sure. And it's funny that what you described is like you were an engineer at Anthropic and you're like
Okay. Claude is gonna be very good at engineering. This isn't gonna be a
Potentially career long term. So I'm gonna move into research.
And
The AI's gonna need me for a long time to build it, to make it smarter.
I would say we still have like
I think
Canvas team has still have like a really cool
Like from the engineers that I
Really like
Yeah, people who like really care about like
Interaction design, like interaction spirits. Like I don't think like models are there yet, but like I think
If y but if you can get the models to like this top one percent of like front end or something.
Um
Sure.
So what I wanna move on to next along these lines is just uh and this is just speculation, but uh
What skills do you think will be most
Valuable
Going forward.
for product teams in particular. So folks are listening and they're like, Okay, this is
Scary.
What should I be
building now to help me
Stay ahead and not.
Damn.
in trouble down the road.
What skills do you think are gonna be most more and more important to build?
Yeah, I think like create a
Thinking
Like you kinda want to like
Um
come up like generate a bunch of ideas and like filter through them.
And it's like build the best product experience.
Listening
You know, you want to like build something that like
The most general model will not replace you. And oftentimes you
You build something and you
Make it really, really good.
For
Like
Specific set of users.
And
I should emote.
is now in like
Your user feedback the mode is like
More in life.
Y whether you listen to them, like whether you you can like rapidly trade, like
The mode is like
in here. I I don't think like V yet to like
There are so many ideas. I think there's an abundance of like
Ideas that you shouldn't look recalls like.
I wouldn't be worried. I feel like in fact I just think like people
NAI fields are like
I w I wish they were like a little more creative and like connecting the dots across like different like fields or something like that to like develop really cool new
Like
Generation.
A new paradigms of interactions with this AI. Like I don't think we've cracked this problem at all.
Um
A couple of years ago I was like telling some people I was like, you know, you kinda want to like
built for the future. So it's like
It doesn't
Necessarily
Matter whether the model is good or not.
But you can build
Product ideas.
such sounds like by the time the models will be really good, it will work really well.
Um
I think it just like happened naturally. Like for example like an anthropic light, right? Like
Um
The clawed artifacts. And I feel like
early days of Canvas was like
back in like twenty twenty two, like before Chai Chi P T, like writing I D was like on all Ch P but I feel like
Claude
one point three model itself.
was like not there to like mate like
really extreme good, like high quality edits, for example, like coding. Um
And I feel like I I feel like
Startup is like Cursor and it's like
doing super well. Like
I guess because you like
Iterates so fast, they like invent like new ways or like training models.
Day
move really fast. They listen to like users like
Massive distribution is like
Yeah.

That's really helpful actually. So what I'm hearing is
that soft skills essentially are gonna be more and more important powerful. You s talked about management, leading people, being
creative and coming up with innovative insights, listening.
There's a post I wrote that I'll link to where I look I I try to analyze what AI how AI will impact product management.
And we're actually very aligned.
And my sense was the same thing that soft skills are gonna become more and more important.
And the things that are gonna be replaced is the hard skills, which is interesting'cause usually people value the hard skills like
Coding design.
Writing really well.
And it's interesting that AI is actually
Really good at that'cause it's
Taking a bunch of data, synthesizing it, and
Writing, creating a thing versus
All these fuzzy things around of
What influences convinces people to do things and aligning and
Listening, like you said, creativity. Anything along those l along those lines come up as I say that?
I think it's actually really, really hard to teach the model how to be
aesthetic or like uh do like visual really good like visual design or like how to be
Extremely creative in the way they write.
I think like
I still think like Chad GP kinda sucks at like writing. And that's because it's like it's like bottled mouth by this like creative reasoning. I think like privatization is like one of the most important like I think like
Um
For Matters, I feel like
I actually like AI research progress is unlocked by like
Management
like research manager is because you have like
Has change set of computers?
And you need to like allocate the computers to
The research path that you feel the most
convinced about it was like you need to like really you need to have like a really high conviction in the research paths.
to put the compute and like it's more like return on investment, um, kind of situation. As like, okay, yeah, like I I'm thinking a lot about like okay, like
How do I across all my projects which projects a higher priority is like
prioritization and also like on the lower level like which experiments are really important to run right now and which are not and like cut through the line. So I think you're like
prioritization, communication, like
um management, um people's skills like
Another thing. Like
Understanding people like
kinda like collaboration, like I think like
Canvas wouldn't be
Like an amazing launch.
If the it wasn't like about like
People.
And I think it's a it's a wonderful glad group of people I like.
I get a chance to look at Rugges, like people like Lee Byron, who's like a co creator like Ralph Kiel, and like some of the best like Apple designers. And it's like
So cool.
To like see
And like how do you create this like collaboration between people is just like
Something that's still humane, I think.
Let me just follow throughout a little bit because I imagine people listening are like, Okay, but once we have AGI or SGI, it's like it'll do all this. It'll you know, it's like there's a world where like why isn't all this done?
I think it's easy to just assume all that.
I'm curious
this idea of creativity and listening.
Why you think AI m isn't good at it?
Other than it's just very hard to
Train it.
to do this well. Is there anything there of just like why this is especially difficult for AI
Now lamps to get good at.
I think currently it's difficult
For
Many reasons I think it's still like a now I feel like research area and is something that like I think my team is like working on is like
Okay, how like how do we teach the model to be like more creative in like the writing? And so like
I'm thinking like
Doesn't you paradigm always that the morals
think more should actually lead to like better writing.
In itself.
But like when it comes down to like
Idea generation
Oh like
Um
discriminating of like what is a good
Like visual design and odd. I feel like if Hasn't had learned
Like
Examples from like people
To discriminate it by well.
I do think it's because like
You know, there are not that many people who are like actually like really like
Oh, it's not like accessible to like models to learn from these people, I guess. Um
So I definitely that's why it's talking. Yeah, that makes sense. Basically there's not enough of you yet.
Uh researchers.
T teaching it to do these things slash.
People that have incredible taste and
Creativity that can teach these things. You could argue this will come.
But I'm not going to
We don't need to keep going down that thread. Let me ask you a specific question. In this post I wrote, I I made this argument that a lot of people disagreed with that strategy.
is something that AI tooling will become increasingly
Great at and take over.
There's the sense that
That's the thing that
People will continue to be much better at and you
can't offload AI basically developing your strategy.
Telling you what to do to win.
My case is
Isn't strategy just take all the inputs, all the data you have available.
Understand the world around you and come up with a plan to win.
Yeah, I would be
Like an LM would be incredibly smart at this.
What's your take?
I think so too. I think like
Again, like we g we you teach the model all sorts of like tools and like capabilities and like reasoning, right? And it's like when it comes down to like
This is for Canvas right now, it would be very cool to like for the model to just like
aggregate all the if you've bought from users like summarize me like
the top five like most painful flow flows like user experiences.
And then like
The model itself is like very capable of like
Like thinking of like knowing how it's
Being made?
Uh
figure out like how to like
treat a data sets for itself to like train on it.
And then they look we have far away from
that kinda like self improvement.
Models becoming like self improved.
buyer
Wait.
Then like the part the Wildman is basically kinda like
self improving like s it's kinda like its own like organism or something. Um
Yeah, like I like strategies like
It's more like data analysis and like
Um
Um
Coming up was like
Like I think what models are really good at is like um
Well connecting the dots, I think.
It's like
Okay, if you have
Usual feedback from the
Sores.
But you also have like internal like dashboards with metrics.
And then we have
You know, like other kind of like feebal.
Um
Oh like
Input.
And then like it can co create like a a a plan for you, like recommendations, even.
I think this is like one of the
most common one use cases for Chaptwood is like
Coming up with like this sorts of things. That makes sense. Like essentially
A human can only comprehend so much information at once and look at so much data at once.
to synthesize takeaways and
As you said, these context windows are huge now. Here's all the information. What's the most important thing I should do?
Yeah, same as like scientific research, is because like you like ideally the model would be able to like suggest like
ideas, like new ideas or like iterate on the experimental like
Given the empirical results of the previous experiments, like how do you
Like
Come up with like new ideas or like the methods.
Yeah, no man.
Uh okay, so just to close the loop on this conversation, this part of the threat is
The skills you're suggesting people
Focus on
Building and leaning into soft skills.
Like creativity.
Managing influence collaboration.
Looking for patterns.
Is that generally
Where your mind is at?
Yeah, I'm thinking a lot about like how do we make organizations more effectively.
And I think this is m most of like management, I guess. It's like
How do you organise like
research teams like
generally teams like combine
c compose teams such that they will be at their maximally
succeed or like at the maximum like performance of
What can possibly
Like if you can like literally create like
The next generation of computers it's just like
the matter of conviction and like
The way you manage through that.
Organizations, all of scaling
Product research does.
Yeah, I think what like you're basically building
this thing and
not efficiently doing it is like
limiting the potential of the human species right now is mismanagement within the research team and open and anthropic and some of these other models.
Yeah.
Yeah, kinda crazy to think about it.
Okay, so speaking of anthropic and open AI, you've worked at both, very few people have worked at both companies and seen how they operate.
I'm curious just what you've noticed about the differences between these two, how they operate, how they think, how they approach stuff. What can you share along those lines?
It's
more similar than different. Uh obviously there was a lot of like
There are some like differences also comes to like nuances.
I'll take culture.
I really love Anthropic and they have a lot of friends there. And I also love Open AI and they still have a lot of friends, though. So it's like it's not about like enemies. I feel like there's like in the I was all like, yeah, the competitors and there's like enemies. This is actually like one big community and like
of people like doing
The same thing.
What they would have learned from Antarctica is
This like
Real Karen Croft?
Towards like moral
behavior, model crossed, model teething.
And
I've been thinking a lot about like okay, like what makes cloth cloth and what makes chip chip hands like
I suppose sometimes sounds like
Operational processes.
That
kind of leads to the outputs.
To to the model.
uh is the odd quoted model. And it's like the reason why Claude has so much more personality and like
Uh
is more like a librarian, I don't know, like I don't know, I am like visualizing
A claw being like a librarian.
Like a um very like naughty or something. Um
It's because I feel like
It's a reflection of
the creators who like making this novel and like
A lot of like
details around like the character and the personality and like whether the model should follow up on this question or like not like
was the correct like ethical behaviour for the model to like in this scenario is like
A lot of like craft, um
And like trouble you did like the stats?
And where I learned.
That part of like
Art I guess.
Uh
I don't know.
I say on through is like much smaller. Like when it joined it was like what, like six seventy people?
when I last the most comfortable people and like obviously the culture changed so much
I really enjoyed the like early days startup like vibes.
And like people
knew each other as a family, but like the culture shifted. I would say like Underwick I learned from Undartic that like
They're much better at like
Focusing and like preaching like
very, very hard like very hardcore priorities, I guess.
And then you need to do it. Like but I think like opening eyes look much more um
You know what else?
and um much more like risk takers in terms of like
Product, or like we start actually
Yeah, we
I don't know.
your full time job can be just like teaching the model how to be like creator writers.
And it's like there's s some luxury in this like research freedom.

That comes to scale, maybe? I don't know. Um
But it gives you
It it's like you'll have I feel like I have much more creative like product freedom to do.
Almost anything, I guess.
within like opening eye like a Ross Chashi Pikine to like
The Ugand. It's like more like
Yeah.
Yeah, that's how I was I was thinking about it. It feels like opening eyes more
Mm-hmm.
Bottoms up.
Uh distributed people, bubble up ideas, try stuff.
There's more
Well and that in more leads to more th products launching, I imagine, more things just kinda being tried.
Versus more of a
Let's just make sure everything we do is awesome and create and craft and

Right.
That's really interesting. I've never heard it described this way.
Uh
Karina, we've covered so much ground.
This is gonna help a lot of people with so many
uh ways of thinking about where the future's going.
Before we get to our very exciting lighting round, I'm curious if there's anything else that you think might be helpful to share or get into
One of my regrets I guess when I was
Early days I don't know it was
That like
I think there was like some luxury of the time.
pre chai chippy to actually like
come in with like a bunch of ideas and like prototype like almost every day.
Um
And I think like we did a lot of cool ideas, like
Claude and Slock was actually one of the first like
uh tool usy like product
It's like
Yeah, cloth could operate in like your workplace now. It's like
But then you like
Add Claude to summarise the thread.
So maybe you have like a entire conversation with someone and then you want to like a summary of like what happened, like you can say like At the cloud summarizes.
Also it was really fun to like even like iterate on the model itself. It's like when you just like talk to the model in like slog forever.
Um
It created like a some social element that's kinda cool, it's kinda like me joining me and like
Um this discord.
Like people learned
So much about prompting and like how to work.
Was like Claude?
I should have the
Features that was like
Early tasks part of time was like, you know.
Every Monday Claude would just like summarise the entire channel.
Or like every Friday we just like summarise like
Bunch of channels.
Uh and give like
the news about the organization or something. So it's it's kinda like really cool, like
Phone factor? I think they're thinking about like phone factors like
A really important like question like in AI.
Especially if you have them like even thicker odds, like
How do we
create like an awesome like product
experience was like oh serious models.
It's like the paradigm between like synchronous real time.
Give an answer.
paradigm into like more asynchronous
Perdonal like.
Agents working on the background.
But then now the question is like the ancients should both trust with you, right? And trust both over time, which is like with humans.
And um
You know, you you saw this collaboration, which is why like a collabor like
This collaboration model was like
you and the model is like so important.
Because you both trust and the model learns from your preferences.
So that it can become like more personalized?
And it will start predicting the next like action that you want to take.
on the computer or something. And it was like
Kinda like more predictive, much more of a
We we could we went from like personal computer to like personal model uh basically here.
Yeah. That seems like such an obvious feature that every
L M should have is a Slack bot version of them. Is that is that a thing I can have you install or is that not a thing right now?
I know that Claude and Sug were sunseted in like twenty twenty three or something, but that's because like I think
Uh I think it was like after Chipt it was mostly like the focus on like
Consumer use cases or like enterprise use cases.
Uh I think it's the one like
I think the form five like clawed and slot'cause like
Um
Was kinda constrained a little bit.
Uh when you want to call it new features.
Never I want that.
I know that J B had like Slack parts, I don't know, like maybe it will come back.
All right. I would I would pay for that.
Uh any other m memories from that time of early days?'Cause that's a really special place to have been
is early days anthropic. Any other memories or stories from that time that might be interesting to share?
I think the very first launch when we felt like
When clips and use the game was like a hundred key context.
Um, launch is like when the models could input
The entire like book.
um give you like some read a book or something.
Um
or the entire financial or like have like multi files.
financial rewards and then like
give you an answer.
Um to the question.
To very specific question.
I think there was something in there that kinda like
Oh my god, this is like a real cool
New capability, not like model capability, but more like
The capabilities
came from the product form factor itself rather than like
The model capability as much. Um
I think like other prototypes that we m
Right.
Thinking about like
Yeah, like
Cl uh there was like one part of the clog work spaces and it's like kinda same like idea of like
Claude and I would have the shared workspace and that shared workspace is like a document and we can like it to another document. And I feel like sometimes the ideas like private ideas log.
And they'll last for like two years. Um, just like in this case.
It's interesting there are these milestones that kind of uh
open up our view of what is happening and where things are going.
Chat GPT, I think was the first of just like, wow, this is
Much better than I would have thought.
You talked about
One hundred K context windows where you could upload a book and ask you questions, have it summarized. I actually use that all the time when I have interview guests and they wrote a book. I sometimes don't have time to read the whole book, so
I use it to help me s understand what the most interesting parts are and then I actually dive into the book just to be clear.
Uh huh.
Uh
And then I don't know, maybe like voice
Was another one where you could talk to say chat GPT.
There any other moments there that you're like wow, this is much better than I thought it was gonna be.

Yeah, I think like
Uh
The computer use agents like
The model operating
The desktop.
And you can essentially think of like
You know, new kind of like
experience where
The model can learn the way you browse.
And from that preference it can just like brass as just like you and like simulation simulated
Persona.
And it's actually very similar to the idea of like
Okay, like
Maybe
Some album doesn't have a lot of like
Um
time maybe I want to like talk to like his simulators like his simulation and ask like or like for example like yeah like I I I really appreciate some of the technical measures that like Ya'cok like but he doesn't have a lot of time so it's like I really want to like ask him this question like how do you respond
Like simulated environments like those. Um
That's a great place to plug Lenny Bot. I have one of those. It's trained on all of my podcasts and newsletters.
And I it sits on many models, I don't know which one exactly.
They use, but it's exactly that. It's
Uh and it's not even me, it's
All the guests that have been on the podcast on the newsletters I wrote, and you could just ask it, How do I grow my product?
How do I develop a strategy and it's actually shockingly good.
Do you feel like it reflects
Yeah, like what is the again. The best part of it is you can talk to it. It's built there's an eleven labs voice.
Version that's trained on my.
voice on the films podcast and it's actually
Very good.
And people like have told me they sit there for hours talking to it.
And somebody
Uh.
told it, interview me like I am on Lenny's podcast, ask me questions about my career, and he did the half hour podcast episode with Lenny Buck. That's so fun.
It's incredible. Future is wild.
Yeah, I think like
Content transformation is like
You know, like
I would imagine sometime live, you know, um
When you generate a sci fi story.
In Canvas.
Like you can like transform this into like audiobooks like very have like very natural like
content transformation like one media to another medium.
I think like
One of my
In earless inspiration.
Um, is like
One of the last episodes of like Westworld.
Where uh
I wanna slow, but where Dolores comes to her work.
At the time when she comes to like
There's like
New workspace and she starts like
Writing a story.
And then if she writes a story, like a three D, like virtual reality stuff, like
Creating on the fly.
So I kinda want to
Um
Kind of cool.
Wow.
Speaking of medium uh uh
I guess I I was I was wondering if I should go in the direction or not, but real quick.
Uh Kevin Wyl slash Kevin Weil. I don't know exactly how to pronounce his last name.
The CPO of uh of Open AI.
Uh.
Is a while or wheel?
I think real. Real. Okay, okay. Let's just say that.
Okay, we open uh He was uh he did a panel at Millennium Friends Summit last year and he made this really infascinating point that
Chat is a really interesting interface.
for these tools because
They're just getting smarter and smarter and smarter and smarter and smarter, and chad continues to work.
As a paradigm to just interact with them.
Similar to a human. You could talk to Albert Einstein, you could talk to someone not very smart and it's all conversation still.
And so it's a really flexible way to interact with increasingly
Good intelligence.
At some point it'll not be so great and
You're talking about all these ways that
You're adding additional ways to interact.
But it's interesting chat proved to be a really powerful
layer on top of all the stuff.
Yeah, that's really cool. I feel like Chad also has like social element, which is like very
Yeah, actually
Yeah, you sometimes want to like get into a group shot and like
Yeah, having conversations with AI is kinda like a group channel in itself, is a message thing. I actually think this this idea of like
How do you build like features like this? Like I see tasks as like
This like
Um, general kinda like
feature that will scale very nicely as the models will develop like new capabilities and so it's like
Like the model will be able to like do better like searches and like
Yeah, create new like
Come up with like a more creative like
writing on like render, you know, React apps and like HTML
prev like apps and like you can have like
Every day and you puzzle for you, like every day, like
continue the story from the few days. It's like
It it scales very nicely.
You mentioned something as we were
Getting into this extra section that we ended up going down is
This idea of
Uh your the agents using a computer.
I know this is actually something you are gonna launch today, the day we're recording it, which will be out by the time this comes out.
Call operator.
Very cool feature that people will have access to.
Yeah, so uh I unfortunately did not work enough, but
I'm really really excited about like
This launch. Um
It's basically
Imagine that can
Complete.
the task in its own like virtual computer. Like in its own virtual environment. You can do any literally task, like
Order me.
A book on Amazon.
And then
Ideally is a moral.
Well, I don't like follow up with you like which books do you want
Or like know you so well that it like start recommending like oh here's the five books that I might recommend to you
To buy and then like you
It's like, yeah, help me
Help me by and then
uh the model goes off uh into its own like virtual
Little browser and like
complete the task and buy the book in my Amazon and then
If you give the model like credentials.
credit cards. Obviously it comes with like a lot of trust and like safety.
Um
Then it will just complete
The thing for youth. There's a virtual
Assistance.
It's interesting how this just sounds like obviously this should happen. Like why is this not a yet a thing, which is also mind blowing that

Should exist. Like just some
AI doing things for you on a computer.
You just ask it to do.
Like it's absurd.
really hard. And I think like
Um
You're still cracking this, but I feel like
I don't know if you're used like top all, it's like a pair programming
No.
But um
I don't know if you love their product. Oh yeah, Shopify uses this. I remember came up on a podcast episode. Oh nice.
Yeah, so it's it's a very cool
product where you can just like call anyone.
At any time.
and then like share screen and the other person can like have access to the screen and like start like
Literally operating your computer.
And it's very like real time like
the Legions is like very
Um
It's like h very high quality. Um
And it's just like
I kinda want the same, it's like I wanna like Pair Program with like my model.
And like the model should
You wanna talk to me or like draw like very specific like section in my code and VS code and like
Tell me like I mean teach me and we can have like different modes. It's like right here. It does like
A product right here for you. I don't know. Um
People some people should both build up.
It sounds like a startup just gotten birthed. Yes. From someone listening to this.
You mentioned that it's very hard to do this
Uh agent.
controlling a computer as you and helping out what makes it so hard.
For whatever however much you can explain briefly.
Much of it is like
Uh because right now the models
Yeah.
Operating on like pixels?
Instead of like
Language or whatnot, like
Pixels is actually really, really hard for the models because like perception or visual perception.
I think there's a still a like a lot of a lot of like multi model like research that's going on.
Um
But I think like
language scale so much like easier compared to like multi models because of that.
Another s like thing that
I guess like my chimney store and I is like
How do you derive human intent?
Um
very correctly.
It's like
Sometimes like
Does the model know enough information to ask a follow up question or like
To complete the task.
You kinda don't want like an agent to like go off for like ten minutes and then come up with like
An answer that you didn't even want that actually creates like much
more verse user experience.
And this is comes with like teaching the model like
Like
People skills. It's like
You know, like
What did people like? Like
kinda like cre creating like the mental model of
the user and like
care about the user in order to ask certain questions like
Actually
That part is like hard.
Two for the morals.
That relates to what we talked about earlier, where the kind of the soft skill people skills pieces.
Not where these models are strong yet.
Okay.
I'm gonna skip the lightning round. I wanna ask just one question from Lightning Round something fun.
Uh
Okay, so when AI replaces your job, Karina.
I'm curious what you're getting and it gives you a stipend. Gives you a monthly stipend. Here's your here's your salary for the month.
What w what would you want to do? What do you want to spend your time on? What will you be doing?
In a future world.
I've been thinking about this
Oh, because I have I feel like I have a lot of
Jobs Options
I would love to be a rider, I think. I think that would be super cool. Um, just like a ride like
short stories, like sci-fi stories. Um
No
I really like Arn T studio. So you know it's like um
Contribution of this to like in the museums.
We just like
Try to preserve like art paintings, but just like painting through
A little bit.
I think that would be really cool. Um
To do.
Um
Yeah.
That sounds beautiful.
I don't know.
Uh what I'm hearing is you need to nerf these models to not get very good at writing.
So that you can continue.
Although at that point you don't need to do it from like you don't need people to buy. You're just doing it for fun. So it doesn't even matter if they're incredibly good at writing.
Or art art conservation.
Oh man, what an episode, what a conversation. What a wild time we're living in.
Karina, thank you.
So much for being here. Two final questions. Where can folks find you online if they wanna reach out and follow up on anything?
And how can listeners be useful to you?
You can find me on Twitter.
Amen.
Um you can also shoot me at email.
on my website um
And
I'm my team is hiring and so like I'm looking for
Research engineers, research scientists, as well as like
machine learning engineers like people who come from like part engineers who want to like learn like model training.
Um actually hiring for like
My team, my team is called Crontier Product Research. And
The tree models we develop new
My first but for product oriented outcomes.
What a place to work. Holy moly.
Uh what's the best way for people to
apply for these uh very lucrative roles.
I think you can shoot me a DM on Twitter. Okay.
Or um
I'm yet to create a job description. Okay. This is the job description.
Or you can apply under like post training.
Team. Yeah.
Okay, I was you're gonna get a flood of DMs. I hope you're prepared.
Kurina, thank you so much for being here. This was incredible.
Thank you so much, Lenny.
Bye, everyone.
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.
Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast.
You can find all past episodes or learn more about the show.
at Lenny's podcast dot com.
See you in the next episode.