back to indexDan Kokotov: Speech Recognition with AI and Humans | Lex Fridman Podcast #151
link |
The following is a conversation with Dan Kokotov, VP of engineering at rev.ai,
link |
which is by many metrics, the best speech to text AI engine in the world.
link |
Rev in general is a company that does captioning and transcription
link |
of audio by humans and by AI.
link |
I've been using their services for a couple of years now and I'm planning
link |
to use Rev to add both captions and transcripts to some of the previous
link |
and future episodes of this podcast to make it easier for people to read
link |
through the conversation or reference various parts of the episode, since
link |
that's something that quite a few people requested.
link |
I'll probably do a separate video on that with links on the podcast website
link |
so people can provide suggestions and improvements there.
link |
Quick mention of our sponsors, Athletic Greens, All in One Nutrition Drink,
link |
Blinkist app that summarizes books, Business Wars podcast, and Cash App.
link |
So the choice is health, wisdom, or money.
link |
Choose wisely my friends, and if you wish, click the sponsor links
link |
below to get a discount and to support this podcast.
link |
As a side note, let me say that I reached out to Dan and the Rev
link |
team for a conversation because I've been using and genuinely loving
link |
their service and really curious about how it works.
link |
I previously talked to the head of Adobe research for the same reason.
link |
For me, there's a bunch of products, usually software, that comes along
link |
and just makes my life way easier.
link |
Examples are Adobe Premiere for video editing, iZotope RX for cleaning up audio,
link |
AutoHotKey on Windows for automating keyboard and mouse tasks, Emacs as an
link |
IDE for everything, including the universe itself.
link |
I can keep on going, but you get the idea.
link |
I just like talking to people who create things I'm a big fan of.
link |
That said, after doing this conversation, the folks at Rev.ai offered to sponsor
link |
this podcast in the coming months.
link |
This conversation is not sponsored by the guest.
link |
It probably goes without saying, but I should say it anyway, that you
link |
can not buy your way onto this podcast.
link |
I don't know why you would want to.
link |
I wanted to bring this up to make a specific point that no sponsor
link |
will ever influence what I do on this podcast, or to the best of
link |
my ability, influence what I think.
link |
I wasn't really thinking about this.
link |
For example, when I interviewed Jack Dorsey, who is the CEO of Square that
link |
happens to be sponsoring this podcast, but I should really make it explicit.
link |
I will never take money for bringing a guest on.
link |
Every guest on this podcast is someone I genuinely am curious to talk to or just
link |
genuinely love something they've created.
link |
As I sometimes get criticized for, I'm just a fan of people.
link |
And that's who I talk to.
link |
As I also talk about way too much, money is really never a consideration.
link |
In general, no amount of money can buy my integrity.
link |
That's true for this podcast, and that's true for anything else I do.
link |
If you enjoy this thing, subscribe on YouTube, review on the Apple podcast,
link |
follow on Spotify, support on Patreon, a podcast on YouTube, and
link |
support on Patreon, or connect with me on Twitter at Lex Friedman.
link |
And now here's my conversation with Dan Kokotov.
link |
You mentioned science fiction on the phone.
link |
So let's go with the ridiculous first.
link |
What's the greatest sci fi novel of all time in your view?
link |
And maybe what ideas do you find philosophically fascinating about it?
link |
The greatest sci fi novel of all time is Dune.
link |
And the second greatest is The Children of Dune.
link |
And the third greatest is The God Emperor of Dune.
link |
So I'm a huge fan of the whole series.
link |
I mean, it's just an incredible world that he created.
link |
And I don't know if you've read the book or not.
link |
It's one of my biggest regrets, especially because a new movie is coming out.
link |
Everyone's super excited about it.
link |
I used to, it's ridiculous to say, and sorry to interrupt, is that I
link |
used to play the video game.
link |
It used to be Dune.
link |
I guess you would call that real time strategy.
link |
I think I remember that game.
link |
Yeah, it was kind of awesome.
link |
Nineties or something.
link |
I think I played it actually when I was in Russia.
link |
I definitely remember it.
link |
I was not in Russia anymore.
link |
I think at the time that I used to live in Russia, I think video games
link |
were about like the suspicion of Pong.
link |
I think Pong was pretty much like the greatest game I ever got to play in Russia,
link |
which was still a privilege right in that age.
link |
So you didn't get color?
link |
You didn't get like, uh, so I left Russia.
link |
I left Russia in 1991, right?
link |
So I was one of the few lucky kids because my mom was a programmer.
link |
So I would go to her work, right?
link |
I would take the, the Metro.
link |
I've got our work and play like on, I guess the equivalent of like a
link |
286 PC, you know, nice floppy disks.
link |
Put back to doing what you get back to doing.
link |
And by the way, the new movie I'm pretty interested in, but the
link |
skeptical, I'm a little skeptical.
link |
I'm a little skeptical.
link |
I saw the trailer.
link |
So there's, there's a David Lynch movie doing as you may know, I'm
link |
a huge David Lynch fan, by the way.
link |
So the movie is somewhat controversial, but it's a little confusing, but it
link |
captures kind of the mood of the book better than I would say like most any
link |
adaptation and like doing so much about kind of mood and the world, right.
link |
But back to the philosophical point.
link |
So in the fourth book, God, emperor of doing, there's a sort of setting where
link |
Leto, one of the characters, he's become this weird sort of God emperor.
link |
He's turned into a gigantic worm.
link |
I mean, you kind of have to read the book to understand what that means.
link |
So the worms are involved, the worms are involved.
link |
You probably saw the worms in the trailer, right.
link |
And in the video, you kind of like merges with the swarm, um, and becomes
link |
this tyrant of the world and like oppresses the people for a long time.
link |
But he has a purpose and the purpose is to kind of, uh, break through kind of
link |
a stagnation period in civilization.
link |
Um, but people have gotten too comfortable, right.
link |
And so you kind of oppresses them so that they explode and like go on to
link |
colonize new worlds and kind of renew the forward momentum of humanity.
link |
And so to me, that's kind of fascinating, right.
link |
You need a little bit of pressure and suffering, right.
link |
To kind of like make progress, not, not, not get too comfortable.
link |
Maybe that's a bit of a cruel philosophy to take away, but that seems to be
link |
the case, unfortunately, obviously, I'm a huge fan of, uh, suffering.
link |
So one of the reasons we're talking today is that a bunch of people requested
link |
that I do transcripts for this podcast and do captioning.
link |
I used to make all kinds of YouTube videos and I would go on up work, I
link |
think, and I would hire folks to do transcription and it was always a pain
link |
in the ass, if I'm being honest, and then I don't know how I discovered Rev.
link |
But when I did, it was this feeling of like, Holy shit, somebody figured
link |
out how to do it just really easily.
link |
I I'm, I'm such a fan of just when people take a problem and they just make it easy.
link |
You know, like just, uh, there's so many, uh, there's so many,
link |
it's like, there's so many things in life that you might not even
link |
be aware of that are painful.
link |
Then Rev, you just like give the audio, give the video, you can
link |
actually give a YouTube link.
link |
And then it comes back like a day later or, uh, two days later, whatever
link |
the hell it is with the captions, you know, all in a standardized format.
link |
It was, I dunno, it was, it was, it was, it was truly a joy.
link |
So I thought I had, you know, just for the hell of it, uh, talk to you
link |
that one other product just made my soul feel good.
link |
One other product that I've used like that is, uh, for people who might
link |
be familiar is called isotope RX.
link |
It's for audio editing and like, and that's another one where it was
link |
like, you just drop it.
link |
I dropped into the audio and it just cleans everything up really nicely.
link |
All the stupid, like the mouth sounds and sometimes there's a background
link |
like sounds due to the malfunction of the equipment.
link |
It can clean that stuff up.
link |
It can, it has a general voice denoising.
link |
It has like automation capabilities where you can do batch processing
link |
and you can put a bunch of effects.
link |
I mean, it just, I dunno, everything else sucked for like voice based
link |
cleanup that I've ever used.
link |
They've used audition, Adobe audition, and he's all kinds of other things
link |
with plugins and you have to kind of figure it all out.
link |
You have to do it manually here.
link |
It's just, it just worked.
link |
So that's another one in this whole pipeline.
link |
It just brought joy to my, to my heart.
link |
Anyway, all that to say is, uh, uh, Rev put a smile to my face.
link |
So can you maybe take a step back and say, what is Rev and how does it work?
link |
And Rev or Rev.com?
link |
Rev, Rev.com, the same thing, I guess, uh, that way we do have Rev.ai now as
link |
well, which we can talk about later.
link |
Like, do you have the actual domain or is it just, uh, the actual domain,
link |
but we also use it kind of as a, as a sub brand.
link |
Oh, so we've, so we use Rev.ai to denote our ASR services, right?
link |
And Rev.com is kind of our more human and to the end user services.
link |
So it's like wordpress.com and wordpress.org, they actually have separate
link |
brands that like, I don't know if you're familiar with what those are.
link |
Yeah, they provide almost like a separate branch of a little bit.
link |
I think with that, it's like wordpress.org is kind of their open source, right?
link |
And, uh, wordpress.com is sort of their hosted commercial offering.
link |
Um, and with us, the differential is a little bit different,
link |
but maybe a similar idea.
link |
Before I launch into, uh, what is Rev?
link |
I was going to say, you know, like you, you were talking about like
link |
Rev was music to your ears, your, your, your field was music to my ears.
link |
To us, the founders of Rev, because, um, Rev was kind of founded to improve
link |
on the model of Upwork that was kind of the original, um, or part of their
link |
original impetus, like our CEO, Jason, was a early employee of Upwork.
link |
So he was very familiar with their work, the company Upwork company.
link |
Um, and so he was very familiar with that model and he wanted to make the whole
link |
experience better because he knew like, when you go at that time, Upwork was
link |
primarily programmers, so the main thing they offered us, if you want to hire,
link |
you know, someone to help you code a little site, right.
link |
Um, you could go on Upwork, um, you could like browse through a list of freelancers,
link |
pick a programmer, you know, have a contract with them and have them do some
link |
work, but it was kind of a difficult experience because, uh, for the, for you,
link |
you would kind of have to browse through all these people, right.
link |
And you have to decide, okay, like, well, is this guy good as, um, or somebody
link |
else better and naturally, you know, you're going to Upwork because you're not
link |
If you're an expert, you probably wouldn't be like getting a programmer
link |
from Upwork, uh, so, so how can you really tell?
link |
So there's a kind of like a lot of potential regret, right?
link |
What if I choose a bad person, they're like, going to be late on the work.
link |
It's going to be a painful experience.
link |
And for the freelancer, it was also painful because, you know, half the time
link |
they spent not on actually doing the work, but kind of figuring out how can I make
link |
my profile most attractive to the buyer, right?
link |
And they're not an expert on that either.
link |
So like Grav's idea was let's remove the barrier, right?
link |
Like, let's make it simple where we'll pick a few, uh, verticals
link |
that are fairly standardizable.
link |
Now we actually started with translation, um, and then we added
link |
audio transcription a bit later and we'll just make it a website.
link |
You go give us your files.
link |
We'll give you back, uh, the results, you know, as soon as possible.
link |
You know, originally maybe it was 48 hours.
link |
Then we made it shorter and shorter and shorter.
link |
Um, yeah, there's a rush processing too.
link |
There's a rush processing now, uh, and, uh, we'll hide all the details from you.
link |
And like, that's kind of exactly what you're experiencing, right?
link |
You don't, you don't need to worry about the details of how the sausage is made.
link |
That's really cool.
link |
The, so you picked like a vertical by vertically, you mean basically a
link |
service, a service category.
link |
Why translation is Rev thinking of potentially going into other verticals
link |
in the future, or is this like the focus now is a translation transcription, like
link |
language, the focus now is, is language or, uh, speech services, generally speech
link |
to text language services, you can kind of group them however you want.
link |
Um, so, but we, uh, originally the categorization was work from home.
link |
So when, uh, work that was done by people on a computer, you know, we weren't trying
link |
to get into, you know, uh, task rabbit type of things and something that could
link |
be relatively standard, not a lot of options.
link |
So we could kind of present the simplified interface, right?
link |
So programming wasn't like a good fit because each programming
link |
project is kind of unique, right?
link |
We're looking for something that, uh, Transcription is, you know, you have five
link |
hours of idea, it's five hours of audio, right?
link |
Translation is somewhat similar.
link |
In that, you know, you can have a five page document, you know, and then you
link |
just can price it by that and then you pick the language you want and that
link |
that's mostly all that is to it.
link |
So those were a few criteria.
link |
We started with translation because we saw the need, um, and, uh, we picked up
link |
kind of a specialty of translation, um, where we would translate things like
link |
board certificates, uh, uh, immigration documents, things like that.
link |
And so they were fairly, um, even more well defined and easy to
link |
kind of tell if we did a good job.
link |
So you can literally charge per type of document.
link |
Was that, was, was that the, so what, what is it now?
link |
Is it per word or something like that?
link |
Like, how do you, like, how do you measure the effort
link |
involved in a particular thing?
link |
So now it looks like for audio translation, it's like,
link |
so now it looks a for audio transcription, right?
link |
It's a per audio minute.
link |
Well, that, that yes, for, for, for our translation, we don't really,
link |
uh, actually focus it on anymore.
link |
Uh, but you know, back when it was still a main business of Revit was per page,
link |
Or per word, depending on the kind of, uh, cause you can also do translation
link |
now on the audio, right?
link |
So like subtitles.
link |
So it would be both, uh, transcription and translation.
link |
I wanted to test the system to see how good it is to see like how, how, uh,
link |
well, is Russian supported?
link |
And it'd be interesting to try it out.
link |
I mean, one of the, now it's only in like the one direction, right?
link |
So you start with English and then you can have subtitles in Russian.
link |
Not really, not really the other way.
link |
Because it's, um, I'm deeply curious about this.
link |
Um, when COVID opens up a little bit, when the economy, when the
link |
world opens up a little bit, you want to build your brand in Russia?
link |
First of all, I'm allergic to the word brand.
link |
All right, I'm definitely not building, uh, any brands in Russia, but I'm going to
link |
Paris to talk to the translators of Dostoevsky and Tolstoy, there's this
link |
famous couple that does translation.
link |
And, you know, I'm more and more thinking of how is it possible to have a
link |
conversation with a Russian speaker?
link |
Cause I have just some number of famous Russian speakers that
link |
I'm interested in talking to, and my Russian is not strong
link |
enough to be witty and funny.
link |
I'm already an idiot in English.
link |
I'm an extra level of like awkward idiot in Russian, but I can understand it.
link |
And I also like wonder how can I create a compelling English Russian
link |
experience for an English speaker?
link |
Like if I, there's a guy named Grigori Perlman, who's a mathematician who,
link |
uh, obviously doesn't speak any English.
link |
So I would probably incorporate like a Russian translator into the picture.
link |
And then it would be like a, not to use a weird term, but like a three, like a
link |
three, three person thing where it's like a dance of, like, I understand it one way.
link |
They don't understand the other way, but I'll be asking questions in English.
link |
I don't know the right way.
link |
It's complicated, but I feel like it's worth the effort for certain kinds of
link |
people, one of whom I'm confident of Vladimir Putin, I'm for sure talking to.
link |
I really want to make it happen.
link |
Cause I think I could do a good job with, but the, the right, you know,
link |
understanding the fundamentals of translation is something I'm really
link |
So that's why I'm starting with, um, the actual translators of like Russian
link |
literature, because they understand the nuance and the beauty of the
link |
language and how it goes back and forth.
link |
But I also want to see, like in speech, how can we do it in real time?
link |
So that's, that's like a little bit of a baby project that I hope to push forward.
link |
But anyway, it's a challenging thing.
link |
So just to share, uh, my dad, um, actually does translation, um, not, not
link |
professional, he's a, uh, he writes poetry.
link |
That was kind of always his, uh, not a hobby, but he's, uh, he, you know, he
link |
had a job, like a day job, but his passion was always writing poetry.
link |
Uh, and then when I got to America, like he started also translating, um, first
link |
he was translating English poetry to Russia.
link |
Now he also like goes the other, uh, the other way, you kind of gain some small
link |
fame in that world anyways, because, uh, recently this poet like Lewis
link |
clock, I don't know if you know of, uh, some American poet, um, she was
link |
awarded the Nobel prize for literature.
link |
Uh, and so my dad had translated, uh, one of her books of poetry in
link |
to Russian, and he was like one of the few.
link |
So he kind of like, they asked him and gave an interview to Radiosvoboda,
link |
if you know what that is.
link |
And he kind of talked about some of the intricacies of translating poetry.
link |
So that's like an extra level of difficulty, right?
link |
Because translating poetry is even more challenging than
link |
translating just, you know, interviews.
link |
Do you remember any, any experiences and challenges to having to do the
link |
translation that, that's the God to you, like something he's talked about?
link |
I mean, a lot of it, I think is word choice, right?
link |
It's the way Russian is structured is first of all, quite different
link |
than, um, the way English is structured, right?
link |
Just there is inflections in Russian and genders, and they don't exist in English.
link |
That's just one of the reasons actually why, um, machine translation is quite
link |
difficult for English to Russian and Russian to English, because there's
link |
such different languages, but then English has like a huge number of words.
link |
Um, many more than Russian, actually, I think.
link |
So it's often difficult to find the right word to convey the same emotional
link |
meaning, yeah, Russian language.
link |
They play with words much more.
link |
So you, you're mentioning that, uh, Rev was kind of born out of, um, trying to
link |
take a vertical on the upwork and then standardize it.
link |
So we're just trying to make the, the freelancer marketplace idea better, right?
link |
Um, better for both customers and better for the freelancers themselves.
link |
Is there something else to the story of Rev finding the right word?
link |
Rev, finding Rev, like what, what did it take to bring it to actually to life?
link |
Was there any pain points?
link |
Um, plenty of, plenty of pain points.
link |
I mean, uh, as, as often the case it's with scaling it up, right?
link |
Um, and in this case, you know, the scaling is kind of scaling the,
link |
the marketplace, so to speak, right?
link |
Rev is essentially a two sided marketplace, right?
link |
Because, you know, there's the customers and then there's the reverse.
link |
Um, if there's not enough Revers, Revers are world color freelancers.
link |
So if there's not enough Revers, then customers have a bad experience, right?
link |
You know, it takes longer to get your work done.
link |
Um, things like that, you know, if there's too many done, the Revers have
link |
a bad experience because they might log on to see like what work is available
link |
and there's not very much work, right?
link |
Uh, so kind of keeping that balance, um, is, is, is a quite challenging problem.
link |
And, you know, that's, that's like a problem we've been working on for many
link |
years and we're still like refining our methods, right?
link |
If you can kind of talk to this gig economy idea, I did a bunch of different
link |
psychology experiments on mechanical Turk, for example, I've asked to do
link |
different kinds of very tricky computer vision annotation on mechanical Turk.
link |
And it's connecting, connecting people in a more systematized way.
link |
I would say, you know, between task and, and, uh, what would you call that worker
link |
is what mechanical Turk calls it.
link |
What do you think about this world of gig economies, of there being a service
link |
that connects customers to workers in a way that's like massively distributed,
link |
like potentially scaling to, it could be, it could be scaled to like
link |
tens of thousands of people, right?
link |
Is there something interesting about that world that you can speak to?
link |
Well, we don't think of it as kind of gig economy, like to some degree,
link |
I don't like the word gig that much, right?
link |
Because to some degree it diminishes the words being done, right?
link |
It sounds kind of like almost amateurish.
link |
Well, maybe in like music industry, like gig is a standard term, but in work, it
link |
kind of sounds like, oh, it's, it's, it's frivolous, um, to us it's, um, improving
link |
the nature of working from home on your own time and on your own terms, right?
link |
And kind of taking away geographical limitations and time limitations, right?
link |
Uh, so, you know, many of our freelancers are maybe work from home moms, right?
link |
And, you know, they don't want the traditional nine to five job, but they
link |
want to make some income and rough kind of like allows them to do that and decide
link |
like exactly how much to work and when to work or by the same token, maybe someone
link |
is, you know, someone wants to live the mountain top, you know, life, right?
link |
You know, cabin in the woods, but they still want to make some money.
link |
Um, and like generally that wouldn't be compatible before, before this new world,
link |
you kind of had to choose, uh, but like with Rev, like, if you like, you don't
link |
have to choose, can you speak to like, what's the demographics like distribution?
link |
Like where do rivers live?
link |
Is there a way to, to, to, to, to, to, to, to, to, to, to, to, to, to, to, to, to,
link |
to, to, to, but you really want to teach it how to, how to run, how to track
link |
Once you're out of the bush, 들어가, things like that, you know,
link |
but like in the back of you know, like hard, but you
link |
just as you approach, there's a lot more control over you.
link |
Like you, you may be Oh, like you know, one day you might go to the
link |
For some years now, we've been doing these little meetings
link |
where the management team will go to some place
link |
and we'll try to meet Revers.
link |
And pretty much wherever we go, it's
link |
pretty easy to find a large number of Revers.
link |
The most recent one we did is in Utah.
link |
But anyway, really.
link |
Are they from all walks of life?
link |
Are these young folks, older folks?
link |
Yeah, all walks of life, really.
link |
Like I said, one category is the work from home.
link |
Students who want to make some extra income.
link |
There are some people who maybe have some social anxiety,
link |
so they don't want to be in the office.
link |
And this is one way for them to make a living.
link |
So it's really pretty wide variety.
link |
But on the flip side, for example,
link |
one Rever we were talking to was a person
link |
who had a fairly high powered career before
link |
and was kind of like taking a break.
link |
And she was almost doing this just to explore and learn
link |
about the gig economy, quote unquote.
link |
So it really is a pretty wide variety of folks.
link |
Yeah, it's kind of interesting through the captioning
link |
process for me to learn about the Revers
link |
because like some are clearly like weirdly knowledgeable
link |
about technical concepts.
link |
Like you can tell by how good they are
link |
at like capitalizing stuff, like technical terms,
link |
like a machine learning or deep learning.
link |
Like I've used Rev to annotate, to caption
link |
the deep learning lectures or machine learning lectures
link |
And it's funny, like a large number of them were like,
link |
I don't know if they looked it up
link |
or were already knowledgeable,
link |
but they do a really good job at like, I don't know.
link |
They invest time into these things.
link |
They will like do research, they will Google things,
link |
you know, to kind of make sure that they got it right.
link |
But to some of them, it's like,
link |
it's actually part of the enjoyment of the work.
link |
Like they'll tell us, you know,
link |
I love doing this because I get paid
link |
to watch a documentary on something, right?
link |
And I learn something while I'm transcribing, right?
link |
Yeah, so what's that captioning transcription process
link |
look like for the Revers?
link |
Can you maybe speak to that to give people a sense,
link |
like how much is automated, how much is manual?
link |
What's the actual interface look like?
link |
All that kind of stuff.
link |
Yeah, so, you know, we've invested a pretty good amount
link |
of time to give like our Revers the best tools possible.
link |
So typical day of forever,
link |
they might log into their workspace,
link |
they'll see a list of audios that need to be transcribed.
link |
And we try to give them tools to pick specifically
link |
the ones they want to do, you know?
link |
So maybe some people like to do longer audios
link |
or shorter audios, people have their preferences.
link |
Some people like to do audios in a particular subject
link |
or from a particular country.
link |
So we try to give people the tools to control,
link |
And then when they pick what they want to do,
link |
we'll launch a specialized editor that we've built
link |
to make transcription as efficient as possible.
link |
They'll start with a speech drag draft.
link |
So, you know, we have our machine learning model
link |
for automated speech recognition, they'll start with that.
link |
And then our tools are optimized to help them correct that.
link |
So it's basically a process of correction.
link |
Yeah, it depends on, you know, I would say the audio.
link |
If the audio itself is pretty good,
link |
like probably like our podcast right now
link |
would be quite good.
link |
So the ASR would do a fairly good job.
link |
But if you imagine someone recorded a lecture,
link |
you know, in the back of a auditorium, right?
link |
Where like the speaker is really far away
link |
and there's maybe a lot of cross talk and things like that,
link |
then maybe the ASR wouldn't do a good job.
link |
So the person might say like, you know what,
link |
I'm just gonna do it from scratch.
link |
Do it from scratch, yeah.
link |
So it kind of really depends.
link |
What would you say is the speed that you can possibly get?
link |
Like, what's the fastest?
link |
Can you get, is it possible to get real time or no?
link |
As you're like listening, can you write as fast as?
link |
Real time would be pretty difficult.
link |
It's actually a pretty, it's not an easy job, you know.
link |
We actually encourage everyone at the company
link |
to try to be a transcriber for a day,
link |
transcriptionist for a day.
link |
And it's way harder than you might think it is, right?
link |
Because people talk fast and people have accents
link |
and all this kind of stuff.
link |
So real time is pretty difficult.
link |
Like there's somebody, we're probably gonna use Rev
link |
to caption this, they're listening to this right now.
link |
What do you think is the fastest
link |
you could possibly get on this right now?
link |
I think on a good audio, maybe two to three X,
link |
I would say, real time.
link |
Meaning it takes two to three times longer
link |
than the actual audio of the podcast.
link |
This is so meta, I could just imagine the Revvers
link |
working on this right now.
link |
You're like, you're way wrong.
link |
You're way wrong, this takes way longer.
link |
But yeah, it definitely works.
link |
Or you doubted me, I could do real time.
link |
Okay, so you mentioned ASR.
link |
Can you speak to what is ASR, automatic speech recognition?
link |
How much, like what is the gap
link |
between perfect human performance
link |
and perfect or pretty damn good ASR?
link |
Yeah, so ASR, automatic speech recognition,
link |
it's a class of machine learning problem, right?
link |
So take speech like we're talking
link |
and transform it into a sequence of words, essentially.
link |
Audio of people talking.
link |
Audio, audio to words.
link |
And there's a variety of different approaches
link |
and techniques, which we could talk about later if you want.
link |
So, we think we have pretty much the world's best ASR
link |
for this kind of speech, right?
link |
So there's different kinds of domains, right, for ASR.
link |
Like one domain might be voice assistance, right?
link |
So Siri, very different than what we're doing, right?
link |
Because Siri, there's fairly limited vocabulary.
link |
You might ask Siri to play a song
link |
or order a pizza or whatever.
link |
And it's very good at doing that.
link |
Very different from when we start talking
link |
in a very unstructured way.
link |
And Siri will also generally adapt to your voice
link |
and stuff like this.
link |
So for this kind of audio, we think we have the best.
link |
And our accuracy, right now I think it's maybe 14%
link |
word error rate on our test suite
link |
that we generally use to measure.
link |
So word error rate is like one way to measure accuracy
link |
So what's 14% word error rate?
link |
So 14% means across this test suite,
link |
of a variety of different audios,
link |
it would be, it would get in some way 14% of the words wrong.
link |
14% of the words wrong.
link |
So the way you kind of calculate it is,
link |
you might add up insertions, deletions, and substitutions,
link |
So insertions is like extra words.
link |
Deletions are words that we said,
link |
but weren't in the transcript, right?
link |
Substitutions is, you said Apple, but I said,
link |
but the ASR thought it was able, something like this.
link |
Human accuracy, most people think realistically,
link |
it's like 3%, 2%, word error rate would be like
link |
the max achievable.
link |
So there's still quite a gap, right?
link |
Would you say that, so YouTube, when I upload videos,
link |
often generates automatic captions.
link |
Are you sort of from a company perspective,
link |
from a company perspective, from a tech perspective,
link |
are you trying to beat YouTube, Google?
link |
It's a hell of a, Google, I mean,
link |
I don't know how seriously they take this task,
link |
but I imagine it's quite serious.
link |
And they, you know, Google is probably up there
link |
in terms of their teams on, on ASR or just NLP,
link |
natural language processing, different technologies.
link |
So do you think you can beat Google?
link |
On this kind of stuff, yeah, we think so.
link |
Google just woke up on my phone.
link |
This is hilarious, okay.
link |
Now Google is listening, sending it back to headquarters.
link |
Who are these rough people?
link |
But that's the goal?
link |
Yeah, I mean, we measure ourselves against like Google,
link |
Amazon, Microsoft, you know, some of the,
link |
some smaller competitors.
link |
And we use like our internal tests with it,
link |
we try to compose it of a pretty representative
link |
set of ideas, maybe it's some podcasts, some videos,
link |
some interviews, some lectures, things like that, right?
link |
And we beat them in our own testing.
link |
And actually Rev offers automated,
link |
like you can actually just do the automated captioning.
link |
So like, I guess it's like way cheaper, whatever it is,
link |
whatever the rates are.
link |
So it's a, by the way, it used to be a dollar per minute
link |
for captioning and transcription,
link |
I think it's like $1.15 or something like that.
link |
That was the other thing that was surprising to me,
link |
it was actually like the cheapest thing you could,
link |
one of the, I mean, I don't remember it being cheaper.
link |
You could on Upwork get cheaper,
link |
but it was clear to me that this,
link |
that's gonna be really shitty.
link |
So like, you're also competing on price.
link |
I think there were services that you can get,
link |
like similar to Rev kind of feel to it,
link |
but it wasn't as automated.
link |
Like the drag and drop, the entirety of the interface,
link |
it's like the thing we're talking about.
link |
I'm such a huge fan of like frictionless,
link |
like Amazon's single buy button, whatever.
link |
The one click, that's genius right there.
link |
Like that is so important for services.
link |
And simplicity and I mean, Rev is almost there.
link |
I mean, there's like some, I'm trying to think.
link |
So I think I've, I stopped using this pipeline,
link |
but Rev offers it and I like it,
link |
but it was causing me some issues on my side,
link |
which is you can connect it to like Dropbox
link |
and it generates the files in Dropbox.
link |
So like it closes the loop to where
link |
I don't have to go to Rev at all and I can download it.
link |
Sorry, I don't have to go to Rev at all
link |
and to download the files.
link |
It could just like automatically copy them.
link |
Right, you're putting your Dropbox in a day later
link |
or maybe a few hours later.
link |
Yeah, it just shows up.
link |
Just shows up, yeah.
link |
Yeah, I was trying to do it programmatically too.
link |
Is there an API interface you can,
link |
I was trying to through like through Python
link |
to download stuff automatically,
link |
but then I realized this is the programmer in me.
link |
Like, dude, you don't need to automate everything
link |
like in life, like flawlessly,
link |
because I wasn't doing enough captions
link |
to justify to myself the time investment
link |
into automating everything perfectly.
link |
Yeah, I would say if you're doing so many interviews
link |
that your biggest roadblock is clicking on the Rev download,
link |
but now you're talking about Elon Musk levels of business.
link |
But for sure, we have like a variety of ways
link |
You know, there's the integration.
link |
You mentioned, I think it's through a company called Zapier,
link |
which kind of can connect Dropbox to Rev and vice versa.
link |
We have an API if you want to really like customize it,
link |
you know, if you want to create the Lex Friedman,
link |
you know, CMS or whatever.
link |
For this whole thing.
link |
So can you speak to the ASR a little bit more?
link |
Like, what does it take?
link |
Like approach wise, machine learning wise,
link |
how hard is this problem?
link |
How do you get to the 3% error rate?
link |
Like, what's your vision of all of this?
link |
Yeah, well, the 3% error rate is definitely,
link |
that's the grand vision.
link |
We'll see what it takes to get there.
link |
But we believe, you know, in ASR,
link |
the biggest thing is the data, right?
link |
Like, that's true of like a lot of
link |
machine learning problems today, right?
link |
The more data you have and high quality of the data,
link |
the better label the data.
link |
Yeah, that's how you get good results.
link |
And we at Rev have kind of like the best data.
link |
Like you're literally,
link |
your business model is annotating the data.
link |
Our business model is being paid to annotate the data.
link |
Being paid to annotate the data.
link |
So it's kind of like a pretty magical flywheel.
link |
And so we've kind of like written this flywheel
link |
And we think we're still kind of in the early stages
link |
of figuring out all the parts of the flywheel to use,
link |
you know, because we have the final transcripts
link |
and we have the audios and we train on that.
link |
But we in principle also have all the edits
link |
that the Revvers make, right?
link |
Oh, that's interesting.
link |
How can you use that as data?
link |
Yeah, that's something for us to figure out in the future.
link |
But, you know, we feel like we're only
link |
in the early stages, right?
link |
So the data is there.
link |
That'd be interesting.
link |
Like almost like a recurrent neural net
link |
for fixing transcripts.
link |
I always remember we did a segmentation annotation
link |
So segmenting the scene, like visual data.
link |
And you can get all,
link |
so it was drawing, people were drawing polygons
link |
around different objects and so on.
link |
And it feels like it always felt like
link |
there was a lot of information in the clicking,
link |
the sequence of clicking that people do,
link |
the kind of fixing of the polygons that they do.
link |
Now there's a few papers written about
link |
how to draw polygons like with a recurrent neural nets
link |
to try to learn from the human clicking.
link |
But it was just like experimental,
link |
you know, it was one of those like CVPR type papers
link |
that people do like a really tiny data set.
link |
It didn't feel like people really tried to do it seriously.
link |
Yeah, I wonder if there's information in the fixing
link |
that provides deeper set of signal
link |
than just like the raw data.
link |
The intuition is for sure there must be, right?
link |
And in all kinds of signals and how long you took
link |
to make that edit and stuff like that.
link |
It's gonna be like up to us.
link |
That's why like the next couple of years
link |
is like super exciting for us, right?
link |
So that's what like the focus is now.
link |
You mentioned rev.ai, that's where you want to.
link |
Yeah, so rev.ai is kind of our way of bringing this ASR
link |
to the rest of the world, right?
link |
So when we started, we were human only.
link |
Then we kind of created this Temi service.
link |
I think you might've used it,
link |
which was kind of ASR for the consumer, right?
link |
So if you don't wanna pay $1.25, but you wanna pay,
link |
now it's 25 cents a minute, I think.
link |
And you get the transcript,
link |
the machine generated transcript and you get an editor
link |
and you can kind of fix it up yourself, right?
link |
Then we started using ASR
link |
for our own human transcriptionists.
link |
And then the kind of rev.ai is the final step
link |
of the journey, which is, you know,
link |
we have this amazing engine.
link |
What can people build with it, right?
link |
What kind of new applications could be enabled
link |
if you have SpeedTrack that's that accurate?
link |
Do you have ideas for this
link |
or is it just providing it as a service
link |
and seeing what people come up with?
link |
It's providing it as a service
link |
and seeing what people come up with
link |
and kind of learning from what people do with it.
link |
And we have ideas of our own as well, of course,
link |
but it's a little bit like, you know,
link |
when AWS provided the building blocks, right?
link |
And they saw what people built with it
link |
and they try to make it easier to build those things, right?
link |
And we kind of hope to do the same thing.
link |
Although AWS kind of does a shitty job of like,
link |
I'm continually surprised, like Mechanical Turk,
link |
for example, how shitty the interface is.
link |
We're talking about like Rev making me feel good.
link |
Like when I first discovered Mechanical Turk,
link |
the initial idea of it was like,
link |
it made me feel like Rev does,
link |
but then the interface is like, come on.
link |
Yeah, it's horrible.
link |
Why is it so painful?
link |
Does nobody at Amazon want to like seriously invest in it?
link |
It felt like you can make so much money
link |
if you took this effort seriously.
link |
And it feels like they have a committee
link |
of like two people just sitting back,
link |
like a meeting, they meet once a month,
link |
like what are we going to do with Mechanical Turk?
link |
It's like two websites making me feel like this,
link |
that and craiglist.org, whatever the hell it is.
link |
It feels like it's designed in the 90s.
link |
Well, Craigslist basically hasn't been updated
link |
pretty much since the guy originally built.
link |
Do you seriously think there's a team,
link |
like how big is the team working on Mechanical Turk?
link |
There's some team, right?
link |
I feel like there isn't.
link |
Well, if nothing else, they benefit from the other teams
link |
like moving things forward in a small way.
link |
But I know what you mean.
link |
We use Mechanical Turk for a couple of things as well.
link |
And yeah, it's painful UI.
link |
It's painful, but yeah, it works.
link |
I think most people, the thing is most people
link |
don't really use the UI, right?
link |
Like we, for example, we use it through the API, right?
link |
But even the API documentation and so on,
link |
like it's super outdated.
link |
Like I don't even know what to...
link |
I mean, the same criticism, as long as we're ranting,
link |
my same criticism goes to the APIs
link |
of most of these companies.
link |
Like Google, for example, the API for the different services
link |
is just the documentation is so shitty.
link |
Like it's not so shitty.
link |
I should actually be...
link |
I should exhibit some gratitude.
link |
Okay, let's practice some gratitude.
link |
The documentation is pretty good.
link |
Like most of the things that the API makes available
link |
It's just that in the sense that it's accurate,
link |
sometimes outdated, but like the degree of explanations
link |
with examples is only covering, I would say,
link |
like 50% of what's possible.
link |
And it just feels a little bit,
link |
like there's a lot of natural questions
link |
that people would wanna ask that doesn't get covered.
link |
And it feels like it's almost there.
link |
Like it's such a magical thing.
link |
Like the Maps API, YouTube API, there's a bunch of stuff.
link |
I gotta imagine it's like, there's probably some team
link |
at Google responsible for writing this documentation
link |
that's probably not the engineers, right?
link |
And probably this team is not where you wanna be.
link |
Well, it's a weird thing.
link |
I sometimes think about this for somebody
link |
who wants to also build a company.
link |
I think about this a lot.
link |
YouTube, the service is one of the most magical,
link |
like I'm so grateful that YouTube exists.
link |
And yet they seem to be quite clueless on so many things
link |
like that everybody's screaming them at.
link |
Like it feels like whatever the mechanism
link |
that you use to listen to your quote unquote customers,
link |
which is like the creators, is not very good.
link |
Like there's literally people that are like screaming why,
link |
like their new YouTube studio, for example.
link |
There's like features that were like begged for
link |
for a really long time.
link |
Like being able to upload multiple videos at the same time.
link |
That wasn't missing for a really, really long time.
link |
Now, like there's probably things that I don't know,
link |
which is maybe for that kind of huge infrastructure,
link |
it's actually very difficult to build
link |
some of these features.
link |
But the fact that that wasn't communicated
link |
and it felt like you're not being heard.
link |
Like I remember this experience for me
link |
and it's not a pleasant experience.
link |
And it feels like the company doesn't give a damn about you.
link |
And that's something to think about.
link |
I'm not sure what that is.
link |
That might have to do with just like small groups
link |
working on these small features and these specific features.
link |
And there's no overarching like dictator type of human
link |
that says like, why the hell are we neglecting
link |
like Steve Jobs type of characters?
link |
Like there's people that we need to speak
link |
to the people that like wanna love our product
link |
Let's fix this shit.
link |
Maybe at some point you just get so fixated
link |
on the numbers, right?
link |
And it's like, well, the numbers are pretty great, right?
link |
Like people are watching,
link |
doesn't seem to be a problem, right?
link |
And you're not like the person that like build this thing,
link |
So you really care about it.
link |
You're just there, you came in as a product manager, right?
link |
You got hired sometime later,
link |
your mandate is like increase this number,
link |
That's brilliantly put.
link |
Like if you, this is, okay, if there's a lesson in this
link |
is don't reduce your company into a metric
link |
of like how much, like you said,
link |
how much people watching the videos and so on
link |
and like convince yourself that everything is working
link |
just because the numbers are going up.
link |
There's something, you have to have a vision.
link |
You have to want people to love your stuff
link |
because love is ultimately the beginning
link |
of like a successful longterm company
link |
is that they always should love your product.
link |
You have to be like a creator
link |
and have that like creator's love for your own thing, right?
link |
Like, and you're pained by these comments, right?
link |
And probably like Apple, I think did this generally
link |
They're well known for kind of keeping teams small
link |
even when they were big, right?
link |
And, you know, he was an engineer,
link |
like there's a book, a creative selection.
link |
I don't know if you read it by a Apple engineer
link |
named Ken Koscienda.
link |
It's kind of a great book actually
link |
because unlike most of these business books where it's,
link |
you know, here's how Steve Jobs ran the company.
link |
It's more like here's how life was like for me, you know,
link |
an engineer here, the projects I worked on
link |
and here what it was like to pitch Steve Jobs, you know,
link |
on like, you know, I think it was in charge of like
link |
the keyboard and the auto correction, right?
link |
And at Apple, like Steve Jobs reviewed everything.
link |
And so he was like, this is what it was like
link |
to show my demos to Steve Jobs and, you know,
link |
to change them because like Steve Jobs didn't like how,
link |
you know, the shape of the little key was off
link |
because the rounding of the corner was like not quite right
link |
or something like this, right?
link |
He was famously a stickler for this kind of stuff.
link |
But because the teams were small,
link |
he really owned this stuff, right?
link |
So he really cared.
link |
Yeah, Elon Musk does that similar kind of thing with Tesla,
link |
which is really interesting.
link |
There's another lesson in leadership in that
link |
is to be obsessed with the details.
link |
And like, he talks to like the lowest level engineers.
link |
Okay, so we're talking about ASR
link |
and so this is basically where I was saying
link |
we're gonna take this like ultra seriously.
link |
And then what's the mission?
link |
To try to keep pushing towards the 3%.
link |
Yeah, and kind of try to build this platform
link |
where all of your, you know, all of your meetings,
link |
you know, they're as easily accessible as your notes, right?
link |
Like, so, like, imagine all the meetings
link |
a company might have, right?
link |
You know, now that I'm like no longer a programmer, right?
link |
Then I'm a quote unquote manager.
link |
That's less like my day as in meetings, right?
link |
And, you know, pretty often I wanna like see
link |
what was said, right?
link |
Who said it, you know?
link |
What's the context?
link |
But it's generally not really something
link |
that you can easily retrieve, right?
link |
Like imagine if all of those meetings
link |
were indexed, archived, you know, you could go back,
link |
you could share a clip like really easily, right?
link |
So that might change completely.
link |
It's like everything that's said,
link |
converted to text might change completely
link |
the dynamics of what we do in this world,
link |
especially now with remote work, right?
link |
With Zoom and so on.
link |
That's fascinating to think about.
link |
I mean, for me, I care about podcasts, right?
link |
And one of the things that was,
link |
you know, I'm torn.
link |
I know a lot of the engineers at Spotify.
link |
So I love them very much because they dream big
link |
in terms of like, they wanna empower creators.
link |
So one of my hopes was with Spotify
link |
that they would use a technology like Rev
link |
or something like that to start converting everything
link |
into text and make it indexable.
link |
Like one of the things that sucks with podcasts
link |
is like, it's hard to find stuff.
link |
Like the model is basically subscription.
link |
Like you find, it's similar to what YouTube used to be like,
link |
which is you basically find a creator that you enjoy
link |
and you subscribe to them.
link |
And like, you just kind of follow what they're doing,
link |
but the search and discovery wasn't a big part of YouTube
link |
like in the early days,
link |
but that's what currently with podcasts,
link |
like is the search and discovery is like non existent.
link |
You're basically searching for like
link |
the dumbest possible thing,
link |
which is like keywords in the titles of episodes.
link |
Even aside from a search and discovery, like all the time.
link |
So I listened to like a number of podcasts
link |
and there's something said,
link |
and I wanna like go back to that later
link |
because I was trying to, I'm trying to remember,
link |
Like maybe like recommended some cool product
link |
that I wanna try out.
link |
And like, it's basically impossible.
link |
Maybe like some people have pretty good show notes.
link |
So maybe you'll get lucky and you can find it, right?
link |
But I mean, if everyone had transcripts
link |
and it was all searchable, it would be so much better.
link |
I mean, that's one of the things that I wanted to,
link |
I mean, one of the reasons we're talking today
link |
is I wanted to take this quite seriously.
link |
The rough thing, I just been lazy.
link |
So, because I'm very fortunate
link |
that a lot of people support this podcast,
link |
that there's enough money now to do a transcription and so on.
link |
And it seemed clear to me, especially like CEOs
link |
and sort of like PhDs, like people write to me
link |
who are like graduate students in computer science
link |
or graduate students in whatever the heck field,
link |
it's clear that their mind,
link |
like they enjoy podcasts
link |
when they're doing laundry or whatever,
link |
but they wanna revisit the conversation
link |
in a much more rigorous way.
link |
And they really wanna transcript.
link |
Like it's clear that they want to analyze conversations.
link |
Like so many people wrote to me
link |
about a transcript for Yosha Bach conversation.
link |
I had just a bunch of conversations.
link |
And then on the Elon Musk side,
link |
like reporters, they wanna write a blog post
link |
about your conversation.
link |
So they wanna be able to pull stuff.
link |
And it's like, they're essentially doing
link |
on your conversation transcription privately.
link |
They're doing it for themselves and then starting to pick,
link |
but it's so much easier when you can actually do it
link |
as a reporter, just look at the transcript.
link |
Yeah, and you can like embed a little thing,
link |
like into your article, right?
link |
Here's what they said, you can go listen
link |
to like this clip from the section.
link |
I'm actually trying to figure out,
link |
I'll probably on the website create
link |
like a place where the transcript goes,
link |
like as a webpage so that people can reference it,
link |
like reporters can reference it and so on.
link |
I mean, most of the reporters probably want
link |
to write clickbait articles that are complete falsifying,
link |
which I'm fine with.
link |
It's the way of journalism, I don't care.
link |
Like I've had this conversation with a friend of mine,
link |
a mixed martial artist, the Ryan Hall.
link |
And we talked about, you know,
link |
as I've been reading The Rise and Fall of the Third Reich
link |
and a bunch of books on Hitler and we brought up Hitler
link |
and he made some kind of comment where like,
link |
we should be able to forgive Hitler
link |
and, you know, like we were talking about forgiveness
link |
and we're bringing that up as like the worst case
link |
possible things, like even, you know,
link |
for people who are Holocaust survivors,
link |
one of the ways to let go of the suffering
link |
they've been through is to forgive.
link |
And he brought up like Hitler as somebody
link |
that would potentially be the hardest thing
link |
to possibly forgive, but it might be a worthwhile pursuit
link |
psychologically, so on, blah, blah, blah, it doesn't matter.
link |
It was very eloquent, very powerful words.
link |
I think people should go back and listen to it.
link |
And then all these journalists,
link |
all these articles written about like MMA fight, UFC fight.
link |
MMA fighter loves Hitler.
link |
No, like, well, no, they didn't.
link |
They were somewhat accurate.
link |
They didn't say like loves Hitler.
link |
They said, thinks that if Hitler came back to life,
link |
we should forgive him.
link |
Like they kind of, it's kind of accurate ish,
link |
but the headline made it sound a lot worse
link |
than it was, but I'm fine with it.
link |
That's the way the world, I wanna almost make it easier
link |
for those journalists and make it easier
link |
for people who actually care about the conversation
link |
to go and look and see.
link |
Right, they can see it for themselves.
link |
There's the headline, but now you can go.
link |
There's something about podcasts,
link |
like the audio that makes it difficult
link |
to jump to a spot and to look
link |
for that particular information.
link |
I think some of it, I'm interested in creating,
link |
like myself experimenting with stuff.
link |
So like taking rev and creating a transcript
link |
and then people can go to it.
link |
I do dream that like, I'm not in the loop anymore,
link |
that like, Spotify does it, right?
link |
Like automatically for everybody,
link |
because ultimately that one click purchase
link |
needs to be there, like, you know.
link |
Like you kind of want support from the entire ecosystem.
link |
Like from the tool makers and the podcast creators,
link |
even clients, right?
link |
I mean, imagine if like most podcast apps,
link |
you know, if it was a standard, right?
link |
Here's how you include a transcript into a podcast, right?
link |
Like it's just an RSS feed ultimately.
link |
And actually just yesterday I saw this company
link |
called Buzzsprout, I think they're called.
link |
So they're trying to do this.
link |
They proposed a spec, an extension to their RSS format
link |
to reference transcripts in a standard way.
link |
And they're talking about like,
link |
there's one client dimension that will support it,
link |
but imagine like more clients support it, right?
link |
So any podcast, you could go and see the transcripts
link |
right in your like normal podcast app.
link |
I mean, somebody, so I have somebody who works with me,
link |
works with helps with advertising, Matt, this awesome guy.
link |
He mentioned Buzzsprout to me, but he says,
link |
it's really annoying because they want exclusive,
link |
they want to host the podcast.
link |
This is the problem with Spotify too.
link |
This is where I'd like to say, like F Spotify,
link |
there's a magic to RSS with podcasts.
link |
It can be made available to everyone.
link |
And then there's all, there's this ecosystem
link |
of different podcast players that emerge
link |
and they compete freely.
link |
And that's a beautiful thing,
link |
that that's why I go on exclusive,
link |
like Joe Rogan went exclusive.
link |
I'm not sure if you're familiar with,
link |
he went to Spotify as a huge fan of Joe Rogan.
link |
I've been kind of nervous about the whole thing,
link |
but let's see, I hope that Spotify steps up.
link |
They've added video, which was very surprising
link |
that they were able to put on.
link |
Exclusive meaning you can't subscribe
link |
to the RSS feed anymore.
link |
It's only in Spotify.
link |
For now you can until December 1st.
link |
And December 1st, it's all, everything disappears
link |
and it's Spotify only.
link |
I, you know, and Spotify gave him a hundred million dollars
link |
So it's an interesting deal, but I, you know,
link |
I did some soul searching and I'm glad he's doing it.
link |
But if Spotify came to me with a hundred million dollars,
link |
I wouldn't do, well, I have a very different relationship
link |
I hate money, but I just think I believe
link |
in the pirate radio aspect of podcasting, the freedom.
link |
And that there's something about.
link |
The open source spirit.
link |
The open source spirit, it just doesn't seem right.
link |
It doesn't feel right.
link |
That said, you know, because so many people care
link |
about Joe Rogan's program,
link |
they're gonna hold Spotify's feet to the fire.
link |
Like one of the cool things with what Joe told me
link |
is the reason he likes working with Spotify
link |
is that they, they're like ride or die together, right?
link |
So they, they want him to succeed.
link |
So that's why they're not actually telling him what to do
link |
despite what people think.
link |
They, they don't tell them,
link |
they don't give them any notes on anything.
link |
They want him to succeed.
link |
And that's the cool thing about exclusivity with a platform
link |
is like, you're kind of wanting each other to succeed.
link |
And that process can actually be very fruitful.
link |
Like YouTube, it goes back to my criticism.
link |
YouTube generally, no matter how big the creator,
link |
maybe for PewDiePie, something like that,
link |
they want you to succeed.
link |
But for the most part, from all the big creators
link |
I've spoken with, Veritasium, all of those folks,
link |
you know, they get some basic assistance,
link |
but it's not like, YouTube doesn't care
link |
if you succeed or not.
link |
They have so many creators.
link |
Yeah, like a hundred other.
link |
So, and especially with, with somebody like Joe Rogan,
link |
who YouTube sees Joe Rogan,
link |
not as a person who might revolutionize the nature of news
link |
and idea space and nuanced conversations.
link |
They see him as a potential person
link |
who has racist guests on,
link |
or like, you know, they see him as like a headache,
link |
So, you know, a lot of people talk about this.
link |
It's a hard place to be for YouTube, actually,
link |
is figuring out with the search and discovery process
link |
of how do you filter out conspiracy theories
link |
and which conspiracy theories represent dangerous untruths
link |
and which conspiracy theories are like vanilla untruths.
link |
And then even when you start having meetings
link |
and discussions about what is true or not,
link |
it starts getting weird.
link |
Yeah, it's difficult these days, right?
link |
I worry more about the other side, right?
link |
Of too much, you know, too much censorship.
link |
Well, maybe censorship is the right word.
link |
I mean, censorship is usually government censorship,
link |
but still, yeah, putting yourself in the position
link |
of arbiter for these kinds of things.
link |
It's very difficult and people think it's so easy, right?
link |
Like, cause like, well, you know, like no Nazis, right?
link |
What a simple principle.
link |
But you know, yes, I mean, no one likes Nazis,
link |
but there's like many shades of gray,
link |
like very soon after that.
link |
Yeah, and then, you know, of course everybody, you know,
link |
there's some people that call our current president a Nazi
link |
and then there's like, so you start getting a Sam Harris.
link |
I don't know if you know that is wasted, in my opinion,
link |
his conversation with Jack Dorsey.
link |
Now I'll also, I spoke with Jack before in this podcast
link |
and we'll talk again, but Sam brought up,
link |
Sam Harris does not like Donald Trump.
link |
I do listen to his podcast.
link |
I'm familiar with his views on the matter.
link |
And he asked Jack Dorsey, he's like,
link |
how can you not ban Donald Trump from Twitter?
link |
And so, you know, there's a set, you have that conversation.
link |
You have a conversation where some number,
link |
some significant number of people think
link |
that the current president of the United States
link |
should not be on your platform.
link |
And it's like, okay.
link |
So if that's even on the table as a conversation,
link |
then everything's on the table for conversation.
link |
And yeah, it's tough.
link |
I'm not sure where I land on it.
link |
I'm with you, I think that censorship is bad,
link |
but I also think the show...
link |
Ultimately, I just also think, you know,
link |
if you're the kind of person that's gonna be convinced,
link |
you know, by some YouTube video, you know,
link |
that, I don't know, our government's been taken over
link |
by aliens, it's unlikely that like, you know,
link |
you'll be returned to sanity simply because, you know,
link |
that video is not available on YouTube, right?
link |
Yeah, I'm with you.
link |
I tend to believe in the intelligence of people
link |
and we should trust them.
link |
But I also do think it's the responsibility of platforms
link |
to encourage more love in the world,
link |
more kindness to each other.
link |
And I don't always think that they're great
link |
at doing that particular thing.
link |
So that, there's a nice balance there.
link |
And I think philosophically, I think about that a lot.
link |
Where's the balance between free speech
link |
and like encouraging people,
link |
even though they have the freedom of speech
link |
to not be an asshole.
link |
That's not a constitutional, like...
link |
So you have the right for free speech,
link |
but like, just don't be an asshole.
link |
Like you can't really put that in the constitution
link |
that the Supreme Court can't be like,
link |
eh, just don't be a dick.
link |
But I feel like platforms have a role to be like,
link |
Maybe do the carrot, like encourage people to be nicer
link |
as opposed to the stake of censorship.
link |
But I think it's an interesting machine learning problem.
link |
Machine, yeah, machine learning for niceness.
link |
It is, I mean, that's...
link |
Responsible, yeah, I mean, it is.
link |
It is a thing, for sure.
link |
Jack Dorsey kind of talks about it
link |
as a vision for Twitter is,
link |
how do we increase the health of conversations?
link |
I don't know how seriously
link |
they're actually trying to do that though.
link |
Which is one of the reasons that I'm in part considering
link |
entering that space a little bit.
link |
It's difficult for them, right?
link |
Because, you know, it's kind of like well known
link |
that people are kind of driven by rage
link |
and you know, outrage maybe is a better word, right?
link |
Outrage drives engagement.
link |
And well, these companies are judged by engagement, right?
link |
In the short term, but this goes to the metrics thing
link |
that we were talking about earlier.
link |
I do believe, I have a fundamental belief
link |
that if you have a metric of long term happiness
link |
of your users, like not short term engagement,
link |
but long term happiness and growth
link |
and both like intellectual, emotional health of your users,
link |
you're going to make a lot more money.
link |
You're going to have long...
link |
Like you should be able to optimize for that.
link |
You don't need to necessarily optimize for engagement.
link |
And that'll be good for society too.
link |
Yeah, no, I mean, I generally agree with you,
link |
but it requires a patient person with, you know,
link |
trust from Wall Street to be able
link |
to carry out such a strategy.
link |
This is what I believe the Steve Jobs character
link |
and Elon Musk character is like,
link |
you basically have to be so good at your job.
link |
Right, you got to pass for anything.
link |
That you can hold the board
link |
and all the investors hostage by saying like,
link |
either we do it my way or I leave.
link |
And everyone is too afraid of you to leave
link |
because they believe in your vision.
link |
But that requires being really good at what you do.
link |
It requires being Steve Jobs and Elon Musk.
link |
There's kind of a reason why like a third name doesn't
link |
come immediately to mind, right?
link |
Like there's maybe a handful of other people,
link |
but it's not that many.
link |
I mean, people say like, why are you...
link |
Like people say that I'm like a fan of Elon Musk.
link |
I'm not, I'm a fan of anybody
link |
who's like Steve Jobs and Elon Musk.
link |
And there's just not many of those folks.
link |
It's the guy that made us believe
link |
that like we can get to Mars, you know, in 10 years, right?
link |
I mean, that's kind of awesome.
link |
And it's kind of making it happen, which is like...
link |
And it's kind of gone like that kind of like spirit, right?
link |
Like from a lot of our society, right?
link |
You know, like we can get to the moon in 10 years
link |
and like we did it, right?
link |
Especially in this time of so much kind of existential dread
link |
that people are going through because of COVID,
link |
like having rockets that just keep going out there
link |
I don't know that it, just like you said,
link |
I mean, it gives you a reason to wake up in the morning
link |
and dream, for us engineers too.
link |
It is inspiring as hell, man.
link |
Let me ask you this, the worst possible question,
link |
which is, so you're like at the core, you're a programmer,
link |
you're an engineer, but now you made the unfortunate choice
link |
or maybe that's the way life goes
link |
of basically moving away from the low level work
link |
and becoming a manager, becoming an executive,
link |
having meetings, what's that transition been like?
link |
It's been interesting.
link |
It's been a journey.
link |
Maybe a couple of things to say about that.
link |
I mean, I got into this, right?
link |
Because as a kid, I just remember this like incredible
link |
amazement at being able to write a program, right?
link |
And something comes to life that kind of didn't exist before.
link |
I don't think you have that in like many other fields,
link |
like you have that with some other kinds of engineering,
link |
but you're maybe a little bit more limited
link |
with what you can do, right?
link |
But with a computer,
link |
you can literally imagine any kind of program, right?
link |
So it's a little bit godlike what you do
link |
like when you create it.
link |
And so, I mean, that's why I got into it.
link |
Do you remember like first program you wrote
link |
or maybe the first program that like made you fall in love
link |
with computer science?
link |
I don't know what was the first program.
link |
It's probably like trying to write one of those games
link |
and basic, you know, like emulate the snake game
link |
I don't remember to be honest, but I enjoyed like,
link |
that's why I always loved about, you know,
link |
being a programmer, it's just the creation process.
link |
And it's a little bit different
link |
when you're not the one doing the creating.
link |
And, you know, another aspect to it I would say is,
link |
you know, when you're a programmer,
link |
when you're a individual contributor,
link |
it's kind of very easy to know when you're doing a good job,
link |
when you're not doing a good job,
link |
when you're being productive,
link |
when you're not being productive, right?
link |
You can kind of see like you trying to make something
link |
and it's like slowly coming together, right?
link |
And when you're a manager, you know, it's more diffuse,
link |
Like, well, you hope, you know, you're motivating your team
link |
and making them more productive and inspiring them, right?
link |
But it's not like you get some kind of like dopamine signal
link |
because you like completed X lines of code, you know, today.
link |
So kind of like you missed that dopamine rush
link |
a little bit when you first become,
link |
but then, you know, slowly you kind of see,
link |
yes, your teams are doing amazing work, right?
link |
And you can take pride in that.
link |
You can get like, what is it?
link |
Like a ripple effect of somebody else's dopamine rush.
link |
Yeah, yeah, you live off other people's dopamine.
link |
So is there pain points and challenges you had to overcome
link |
from becoming, from going to a programmer to becoming
link |
a programmer of humans?
link |
Programmer of humans.
link |
I don't know, humans are difficult to understand,
link |
you know, it's like one of those things,
link |
like trying to understand other people's motivations
link |
and what really drives them.
link |
It's difficult, maybe like never really know, right?
link |
Do you find that people are different?
link |
Like I, one of the things, like I had a group at MIT
link |
that, you know, I found that like some people
link |
I could like scream at and criticize like hard
link |
and that made them do like much better work
link |
and really push them to their limit.
link |
And there's some people that I had to nonstop compliment
link |
because like they're so already self critical,
link |
like about everything they do that I have to be constantly
link |
like, like I cannot criticize them at all
link |
because they're already criticizing themselves
link |
and you have to kind of encourage
link |
and like celebrate their little victories.
link |
And it's kind of fascinating that like how that,
link |
the complete difference in people.
link |
Definitely people respond to different motivations
link |
and different loads of feedback
link |
and you kind of have to figure it out.
link |
It was like a pretty good book,
link |
which for some reason now the name escapes me,
link |
about management, first break all the rules.
link |
First break all the rules?
link |
First break all the rules.
link |
It's a book that we generally like ask a lot of
link |
like first time managers to read it rough.
link |
And like one of the kind of philosophies
link |
is managed by exception, right?
link |
Which is, you know, don't like have some standard template
link |
like, you know, here's how I, you know,
link |
tell this person to do this or the other thing.
link |
Here's how I get feedback, like manage by exception, right?
link |
Every person is a little bit different.
link |
You have to try to understand what drives them.
link |
And tailor it to them.
link |
Since you mentioned books,
link |
I don't know if you can answer this question,
link |
but people love it when I ask it, which is,
link |
are there books, technical fiction or philosophical
link |
that you enjoyed or had an impact on your life
link |
that you would recommend?
link |
You already mentioned Dune, like all of the Dune.
link |
The second one was probably the weakest, but anyway,
link |
so yeah, all of the Dune is good.
link |
I mean, yeah, can you just slow little tangent on that?
link |
Is, how many Dune books are there?
link |
Like, do you recommend people start with the first one
link |
Yeah, you gotta have to read them all.
link |
I mean, it is a complete story, right?
link |
So you start with the first one,
link |
you gotta read all of them.
link |
So it's not like a tree, like a creation of like
link |
the universe that you should go in sequence?
link |
You should go in sequence, yeah.
link |
It's kind of a chronological storyline.
link |
There's six books in all.
link |
Then there's like many kind of books
link |
that were written by Frank Herbert's son,
link |
but those are not as good.
link |
So you don't have to bother with those.
link |
But the main sequence is good.
link |
So what are some other books?
link |
Maybe there's a few.
link |
So I don't know that like, I would say there's a book
link |
that kind of, I don't know, turned my life around
link |
or anything like that, but here's a couple
link |
that I really love.
link |
So one is Brave New World by Aldous Huxley.
link |
And it's kind of incredible how prescient he was
link |
about like what a brave new world might be like, right?
link |
You know, you kind of see genetic sorting in this book,
link |
right, where there's like these alphas and epsilons
link |
and how from like the earliest time of society,
link |
like they're sort of like, you can kind of see it
link |
in a slightly similar way today where,
link |
well, one of the problems with society is people
link |
are kind of genetically sorting a little bit, right?
link |
Like there's much less, like most marriages, right,
link |
are between people of similar kind of intellectual level
link |
or socioeconomic status, more so these days than in the past.
link |
And you kind of see some effects of it
link |
in stratifying society and kind of he illustrated
link |
what that could be like in the extreme.
link |
There's different versions of it on social media as well.
link |
It's not just like marriages and so on.
link |
Like it's genetic sorting in terms of what Dawkins called
link |
memes as ideas being put into these bins
link |
of these little echo chambers and so on.
link |
Yeah, I know, so that's the book
link |
that's I think a worthwhile read for everyone.
link |
I mean, 1984 is good, of course, as well.
link |
Like if you're talking about, you know,
link |
dystopian novels of the future.
link |
Yeah, it's a slightly different view of the future, right?
link |
But I kind of like identify with Brave New World a bit more.
link |
Yeah, speaking of not a book,
link |
but my favorite kind of dystopian science fiction
link |
is a movie called Brazil,
link |
which I don't know if you've heard of.
link |
I've heard of and I know I need to watch it,
link |
but yeah, because it's in, is it in English or no?
link |
It's an English movie, yeah.
link |
And it's a sort of like dystopian movie
link |
of authoritarian incompetence, right?
link |
It's like nothing really works very well, you know,
link |
the system is creaky, you know,
link |
but no one is kind of like willing to challenge it,
link |
you know, just things kind of ample along
link |
and kind of strikes me as like a very plausible future
link |
of like, you know, what authoritarianism might look like.
link |
It's not like this, you know,
link |
super efficient evil dictatorship of 1984.
link |
It's just kind of like this badly functioning, you know,
link |
but it's status quo, so it just goes on.
link |
Yeah, that's one funny thing that stands out to me
link |
is in whether it's authoritarian, dystopian stuff,
link |
or just basic like, you know,
link |
if you look at the movie Contagion,
link |
it seems in the movies,
link |
government is almost always exceptionally competent.
link |
Like it's like used as a storytelling tool
link |
of like extreme competence.
link |
Like, you know, you use it whether it's good or evil,
link |
but it's competent.
link |
It's very interesting to think about
link |
where much more realistically is it's incompetence
link |
and that incompetence isn't itself has consequences
link |
that are difficult to predict.
link |
Like bureaucracy has a very boring way of being evil,
link |
of just, you know, if you look at the show,
link |
HBO show at Chernobyl,
link |
it's a really good story of how bureaucracy, you know,
link |
leads to catastrophic events,
link |
but not through any kind of evil
link |
in any one particular place,
link |
but more just like the...
link |
It's just the system kind of system.
link |
Distorting information as it travels up the chain
link |
that people unwilling to take responsibility for things
link |
and just kind of like this laziness resulting in evil.
link |
There's a comedic version of this,
link |
I don't know if you've seen this movie,
link |
it's called The Death of Stalin.
link |
Yeah, I liked that.
link |
I wish it wasn't so...
link |
There's a movie called Inglourious Bastards
link |
about, you know, Hitler and so on.
link |
For some reason, those movies pissed me off.
link |
I know a lot of people love them,
link |
but like, I just feel like there's not enough good movies,
link |
even about Hitler.
link |
There's good movies about the Holocaust,
link |
but even Hitler, there's a movie called Dawnfall
link |
that people should watch.
link |
I think it's the last few days of Hitler.
link |
That's a good movie, turned into a meme, but it's good.
link |
But on Stalin, I feel like I may be wrong on this,
link |
but at least in the English speaking world,
link |
there's not good movies about the evil of Stalin.
link |
Let's try to see that.
link |
Actually, so I agree with you on Inglourious Bastard.
link |
I didn't love the movie
link |
because I felt like kind of the stylizing of it, right?
link |
The whole Tarantino kind of Tarantinoism, if you will,
link |
kind of detracted from it
link |
and made it seem like unserious a little bit.
link |
But Death of Stalin, I felt differently.
link |
Maybe it's because it's a comedy to begin with.
link |
This is not like I'm expecting seriousness,
link |
but it kind of depicted the absurdity
link |
of the whole situation in a way, right?
link |
I mean, it was funny, so maybe it does make light of it,
link |
but something goes probably like this, right?
link |
Like a bunch of kind of people,
link |
they're like, oh shit, right?
link |
But like the thing is,
link |
it was so close to like what probably was reality.
link |
It was caricaturing reality
link |
to where I think an observer might think that this is not,
link |
like they might think it's a comedy.
link |
Well, in reality, that's the absurdity
link |
of how people act with dictators.
link |
I mean, that's, I guess it was too close to reality for me.
link |
The kind of banality of like what were eventually
link |
like fairly evil acts, right?
link |
But like, yeah, they're just a bunch of people
link |
trying to survive.
link |
Cause I think there's a good,
link |
I haven't watched it yet, the good movie on,
link |
the movie on Churchill with Gary Oldman,
link |
I think it's Gary Oldman.
link |
I may be making that up.
link |
But I think he won,
link |
like he was nominated for an Oscar or something.
link |
So I like, I love these movies about these humans
link |
and Stalin, like Chernobyl made me realize the HBO show
link |
that there's not enough movies about Russia
link |
that capture that spirit.
link |
I'm sure it might be in Russian there is,
link |
but the fact that some British dude that like did comedy,
link |
I feel like he did like hangover or some shit like that.
link |
I don't know if you're familiar
link |
with the person who created Chernobyl,
link |
but he was just like some guy
link |
that doesn't know anything about Russia.
link |
And he just went in and just studied it,
link |
like did a good job of creating it
link |
and then got it so accurate, like poetically.
link |
And the facts that you need to get accurate,
link |
he got accurate, just the spirit of it
link |
down to like the bowls that pets use,
link |
just the whole feel of it.
link |
It was incredible.
link |
It was good, yeah, I saw the series.
link |
Yeah, it's incredible.
link |
It's made me wish that somebody did a good,
link |
like 1930s, like starvation that Stalin led to,
link |
like leading up to World War II
link |
and in World War II itself, like Stalingrad and so on.
link |
Like, I feel like that story needs to be told.
link |
Millions of people died.
link |
And to me, it's so much more fascinating than Hitler
link |
because Hitler is like a caricature of evil almost
link |
that it's so, especially with the Holocaust,
link |
it's so difficult to imagine that something like that
link |
is possible ever again.
link |
Stalin to me represents something that is possible.
link |
Like the so interesting, like the bureaucracy of it
link |
is so fascinating that it potentially might be happening
link |
in the world now, like that we're not aware of,
link |
like with North Korea, another one that,
link |
like there should be a good film on.
link |
And like the possible things that could be happening
link |
in China with overreach of government.
link |
I don't know, there's a lot of possibilities there.
link |
Yeah, I wonder how much, you know,
link |
I guess the archives should be maybe more open nowadays,
link |
right, I mean, for a long time, they just, we didn't know,
link |
right, or anyways, no one in the West knew for sure.
link |
Well, there's a, I don't know if you know him,
link |
there's a guy named Stephen Kotkin.
link |
He is a historian of Stalin that I spoke to on this podcast.
link |
I'll speak to him again.
link |
The guy knows his shit on Stalin.
link |
He like read everything and it's so fascinating
link |
to talk to somebody, like he knows Stalin better
link |
than Stalin himself, it's crazy.
link |
Like you have, so he's, I think he's a Princeton,
link |
he is basically, his whole life is Stalin.
link |
And in that context, he also talks about
link |
and writes about Putin a little bit.
link |
I've also read at this point,
link |
I think every biography of Putin, English biography of Putin,
link |
I need to read some Russians.
link |
Obviously, I'm mentally preparing
link |
for a possible conversation with Putin.
link |
So what is your first question to Putin
link |
when you have him on the podcast?
link |
I, it's interesting you bring that up.
link |
First of all, I wouldn't tell you, but.
link |
You can't give it away now.
link |
But I actually haven't even thought about that.
link |
So my current approach, and I do this with interviews often,
link |
obviously that's a special one,
link |
but I try not to think about questions until last minute.
link |
I'm trying to sort of get into the mindset.
link |
And so that's why I'm soaking in a lot of stuff,
link |
not thinking about questions, just learning about the man.
link |
But in terms of like human to human,
link |
it's like, I would say it's,
link |
I don't know if you're a fan of mob movies,
link |
but like the mafia, which I am, like Goodfellas and so on,
link |
he's much closer to like mob morality, which is like.
link |
Mob morality, maybe, I could see that.
link |
But I like your approach anyways of this,
link |
the extreme empathy, right?
link |
It's a little bit like Hannibal, right?
link |
Like if you ever watched the show Hannibal, right?
link |
They had that guy, well, you know Hannibal of course, like.
link |
Yeah, Silence of the Lambs.
link |
But there were those TV shows as well,
link |
and they focused on this guy, Will Durant,
link |
who's a character like extreme empath, right?
link |
So in the way he like catches all these killers,
link |
as he pretty much, he can empathize with them, right?
link |
Like he can understand why they're doing
link |
the things they're doing, right?
link |
It's a pretty excruciating thing, right?
link |
Like, because you're pretty much like spending
link |
half your time in the head of evil people, right?
link |
I mean, I definitely try to do that with others.
link |
So you should do that in moderation,
link |
but I think it's a pretty safe place, safe place to be.
link |
One of the cool things with this podcast,
link |
and I know you didn't sign up to hear me
link |
listen to this bullshit, but.
link |
That was interesting.
link |
I, and what's his name?
link |
Chris Latner, who's a Google,
link |
oh, he's not Google anymore, SciFi.
link |
He's legit, he's one of the most legit engineers
link |
I talk with, I talk with him again on this podcast.
link |
And one of the, he gives me private advice a lot.
link |
And he said for this podcast, I should like interview,
link |
like I should widen the range of people
link |
because that gives you much more freedom to do stuff.
link |
Like, so his idea, which I think I agree with Chris
link |
is that you go to the extremes.
link |
You just like cover every extreme base
link |
and then it gives you freedom to then go
link |
to the more nuanced conversations.
link |
And it's kind of, I think there's a safe place for that.
link |
There's certainly a hunger for that nuanced conversation,
link |
I think, amongst people where like on social media,
link |
you get canceled for anything slightly tense,
link |
that there's a hunger to go full.
link |
Right, you go so far to the opposite side.
link |
And that's like demystifies it a little bit, right?
link |
There is a person behind all of these things.
link |
And that's the cool thing about podcasting,
link |
like three, four hour conversations
link |
that it's very different than a clickbait journalism,
link |
it's like the opposite, that there's a hunger for that.
link |
There's a willingness for that.
link |
Yeah, especially now, I mean,
link |
how many people do you even see face to face anymore?
link |
Right, like this, you know?
link |
It's like not that many people like in my day today,
link |
aside from my own family that like I sit across.
link |
It's sad, but it's also beautiful.
link |
Like I've gotten the chance to like,
link |
like our conversation now, there's somebody,
link |
I guarantee you there's somebody in Russia
link |
listening to this now, like jogging.
link |
There's somebody who is just smoked some weed,
link |
sit back on a couch and just like enjoying.
link |
I guarantee you that we'll write in the comments right now
link |
that yes, I'm in St. Petersburg, I'm in Moscow, whatever.
link |
And we're in their head and they have a friendship with us.
link |
I'm the same way, I'm a huge fan of podcasting.
link |
It's a beautiful thing.
link |
I mean, it's a weird one way human connection.
link |
Like before I went on Joe Rogan and still,
link |
I'm just a huge fan of his.
link |
So it was like surreal.
link |
I've been friends with Joe Rogan for 10 years, but one way.
link |
Yeah, from this way, from the St. Petersburg way.
link |
Yeah, the St. Petersburg way and it's a real friendship.
link |
I mean, now it's like two way, but it's still surreal.
link |
And that's the magic of podcasting.
link |
I'm not sure what to make of it.
link |
That voice, it's not even the video part.
link |
It's the audio that's magical.
link |
I don't know what to do with it,
link |
but it's people listen to three, four hours.
link |
Yeah, we evolved over millions of years, right?
link |
To be very fine tuned to things like that, right?
link |
Oh, expressions as well, of course, right?
link |
But back in the day on the Savannah,
link |
you had to be very attuned to whether
link |
you had a good relationship with the rest of your tribe
link |
or a very bad relationship, right?
link |
Because if you had a very bad relationship,
link |
you were probably gonna be left behind
link |
and eaten by the lions.
link |
Yeah, but it's weird that the tribe is different now.
link |
Like you could have a one way connection with Joe Rogan
link |
as opposed to the tribe of your physical vicinity.
link |
But that's why it works with the podcasting,
link |
but it's the opposite of what happens on Twitter, right?
link |
Because all those nuances are removed, right?
link |
You're not connecting with the person
link |
because you don't hear the voice.
link |
You're connecting with like an abstraction, right?
link |
It's like some stream of tweets, right?
link |
And it's very easy to assign to them
link |
any kind of evil intent or dehumanize them,
link |
which it's much harder to do when it's a real voice, right?
link |
Because you realize it's a real person behind the voice.
link |
Let me try this out on you.
link |
I sometimes ask about the meaning of life.
link |
Do you, your father now, an engineer,
link |
you're building up a company.
link |
Do you ever zoom out and think like,
link |
what the hell is this whole thing for?
link |
Like why are we descended to vapes even on this planet?
link |
What's the meaning of it all?
link |
That's a pretty big question.
link |
I think I don't allow myself to think about it too often,
link |
or maybe like life doesn't allow me
link |
to think about it too often.
link |
But in some ways, I guess the meaning of life
link |
is kind of contributing to this kind of weird thing
link |
we call humanity, right?
link |
Like it's in a way, you can think of humanity
link |
as like a living and evolving organism, right?
link |
That like we all contributing in a sway way,
link |
but just by existing, by having our own unique set
link |
of desires and drives, right?
link |
And maybe that means like creating something great.
link |
And it's bringing up kids who are unique and different
link |
and seeing like, they can join what they do.
link |
But I mean, to me, that's pretty much it.
link |
I mean, if you're not a religious person, right?
link |
Which I guess I'm not, that's the meaning of life.
link |
It's in the living and in the creation.
link |
Yeah, there's something magical
link |
about that engine of creation.
link |
Like you said, programming, I would say,
link |
I mean, it's even just actually what you said
link |
with even just programs.
link |
I don't care if it's like some JavaScript thing
link |
on a button on the website.
link |
It's like magical that you brought that to life.
link |
I don't know what that is in there, but that seems,
link |
that's probably some version of like reproduction
link |
and sex, whatever that's in evolution.
link |
But like creating that HTML button has echoes
link |
of that feeling and it's magical.
link |
Right, well, I mean, if you're a religious person,
link |
maybe you could even say, all right,
link |
like we were created in God's image, right?
link |
Well, I mean, I guess part of that is the drive
link |
to create something ourselves, right?
link |
I mean, that's part of it.
link |
Yeah, that HTML button is the creation in God's image.
link |
Maybe hopefully it'll be something a little more.
link |
So dynamic, maybe some JavaScript.
link |
Yeah, maybe some JavaScript, some React and so on.
link |
But no, I mean, I think that's what differentiates us
link |
from the apes, so to speak.
link |
Yeah, we did a pretty good job.
link |
Dan, it was an honor to talk to you.
link |
Thank you so much for being part of creating
link |
one of my favorite services and products.
link |
This is actually a little bit of an experiment.
link |
Allow me to sort of fanboy over some of the things I love.
link |
So thanks for wasting your time with me today.
link |
It was really fun.
link |
Well, it was awesome.
link |
Thanks for having me on and giving me a chance
link |
Thanks for listening to this conversation
link |
with Dan Kokotov and thank you to our sponsors,
link |
Athletic Greens, Only One Nutrition Drink,
link |
Blinkist app that summarizes books,
link |
Business Wars podcast and Cash App.
link |
So the choice is health, wisdom or money.
link |
Choose wisely, my friends.
link |
And if you wish, click the sponsor links below
link |
to get a discount and to support this podcast.
link |
And now let me leave you with some words
link |
from Ludwig Wittgenstein.
link |
The limits of my language means the limits of my world.
link |
Thank you for listening and hope to see you next time.