Hello, welcome to the DevTools FM podcast.
It's a podcast about developer tools and the people who make them.
I'm Andrew and this is my co-host Justin.
Everybody, this is a little bit of a different scene for us.
So we would like to thank our panel for being here.
We are at the Epic Web Dev conference and really excited to be here.
So what we're going to do is just give our panel an opportunity for each of you to introduce yourself
and to kind of tell us a little bit about the work you do.
And then we'll dive in and do some questions and that'll be it.
Yeah, so...
My name is Scott Hanselman.
I'm a VP of Developer Community at Microsoft.
I've been programming professionally for 32 years.
I remember when the internet was there and I was there for eternal September.
If anyone remembers that, if you don't, go and Google with Bing for eternal September.
And I work on DevTools at Microsoft.
Hi, I'm Michael Chan, Chan or ChanTastic.
I answer to all of them.
I am in between jobs.
I'm fun employed right now.
I just left a position at Chromatic doing developer experience type stuff.
And I am headed towards WorkOS to do developer education.
So super excited about that.
I like making videos as like a final stage of my learning process.
It's been something I've done for like, I don't know, 15 years.
I've been in tech.
So yeah, it's a little bit about me.
Nice.
Well, hello.
My name is Yuna, like Unicorn.
And I currently work at Google on the Chrome team focusing on web platform,
specifically UI features.
I also do a lot of work with DevTools to make sure there's support of those features.
Before this developer relations job, I was working in front end for many years.
So doing UI engineering, front end development, kind of that intersection between design and code.
That's kind of where I find my happy place.
So that's me.
Yeah, sort of started doing like the community stuff before it became my job,
which was neat.
And now I get to do that for work.
So as DevRails, you're like the first line of defense before the developers.
You meet the community where they are.
So in that, how do you handle the community feedback?
I'm sure there's a lot.
And then like, once you take that feedback, how do you communicate it to the engineering teams?
And then have you championed a feature from that community feedback that you're proud of?
I mean, I can go first while you're deliberating.
Yes, I love community feedback.
It is the reason I like going to events like this,
because I want to hear about what you're working on, what you're struggling with,
what you're frustrated with, what you're excited about.
All those things are so critical to making sure that as we build for the web,
the web is supporting the users that build for it.
It's as simple as that.
So like a huge part of my job is making sure that we're working on the right things.
We're prioritizing the right things.
So the things that we're landing make sense to developers who are using them,
that they like the solution to the things and that they can use them.
So I really love feedback.
Please come talk to me if you're listening to this.
You can always message me, send a message on Twitter, DM me, send me an email.
I really appreciate the user feedback.
And then we bring that to our team, whether it's qualitative analysis,
like talking to developers, like bringing back quotes, bringing back experiences.
We're quantitative, like we sponsor surveys like the state of CSS and those sort of things,
so we can get more aggregated data.
And then we use that data to really make decisions on what to prioritize,
like how to kind of focus the work that we do in the future.
It makes a big difference.
I don't think the developers realize how big of an impact it has.
Like talking to the DevRel team, the engineering teams,
like that's how you make change, work for what you need in the ecosystem.
So definitely love it.
Yeah, I think that's a perfect answer.
And I'd like to add to that because I think that there's...
In terms of how you roll that feedback back, it's very team dependent.
Like everything's, you know, it depends.
But I feel like a lot of it is like context application
because the world is very unkind.
And I think that, you know, we develop these shells around like the things
that we've built, we're very protective about them, we like them,
we have our reasons.
And so when people come in, you know, from out of nowhere
and like really hate this thing that we're making,
it's natural to be defensive.
And I think a lot of times, like the engineering teams,
like, you know, we'll feel that faster.
And like having that kind of like layer of tissue
in between the like DevRel layer is really helpful
because we can have this like disconnect about it.
Like, well, we didn't build it.
Like we just want to make sure that it's best for everybody.
And at least, you know, that was my position.
And I think that a lot of it is applying the right context
and being able to communicate that in a way that's softer,
that communicates the things that you understand.
Like, I understand that we're trying to get here
and this is the future and this is what we're trying to build.
And some of these things have to get sorted out first.
But like, this is still something that people want, right?
And if we can meet them, like even a little bit,
like we'll be able to indicate, like to them,
that we care about what they care about as well
and like meet them somewhere in the middle.
And I think that that's kind of,
being that buffer is a little bit challenging,
but like a really fun part of the role sometimes.
Yeah, I agree with everything that you both said.
I would add that it's a, it's an exercise in extreme
and ongoing empathy.
And it is a marathon.
It's not a sprint.
And it is helpful in my experience
that DevRel focus on the dev as much as the rel.
Because, you know, if you've just got it started,
maybe you came out of college and went directly into DevRel,
but you've never like carried the page,
you're run a run a live site,
your empathy is going to be slightly different.
But when, you know, the, I don't know,
chief architect, a little Debbie Snack cakes,
has a site down and you're like, oh man,
I remember what it was like to work for a company
that, you know, a thankless job and the site's down.
I really feel that in my chest.
So all these years later, you know, when someone has a problem,
yeah, the telemetry says it's only 3% of users
or 1% of users, but it's 100% of that
guy or gal who's having that.
And I really feel responsible.
It gets me in trouble though sometimes,
because Microsoft, the company I work for,
and of course you must feel this way about Google,
like I can't own all this stuff.
Like dude, my Minecraft account.
Ah, really, I feel it for you.
I really want to help you, but it's not my space.
You know what I mean?
Like I get emotionally attached
to solving these kind of problems.
Sometimes I'll do a PR, sometimes it's about raising a bug,
sometimes it's confirming with telemetry
that this is in fact a bug and not an outlier.
I'd like to transition a little bit and sort of just talk
about another part of public communication,
which is just education.
I mean, as a public figure and you're talking
about tech, either a product or representing a company,
it's like part of your role becomes education.
And it's interesting to think about the tech industry,
especially now.
So the tech industries went through a lot of different waves.
We had the dot-com boom and bust,
and then we had, well, bust.

Then we definitely had, you know,
there was a hiring boom during the pandemic,
and then as interest rates rose,
we definitely have had experienced a lot of layoffs.
And so it's an interesting time to be in the industry,
and especially for people who came up during the boom,
who were new to the industry.
So how do each of you feel about the state
of the industry as it stands today?
And then if you are talking to someone
who's considering joining the industry,
what do you say to them?
Respectfully, I think there's three different
complicated questions inside there.
There's the state of the industry,
there's advice for the emerging person,
but I wanna talk for my answer about education.
I think that the job is almost entirely education
and advocacy, which is itself being a teacher.
I've always long wondered, like, what's that tagline?
What's that Twitter bio?
What's that resume thing that you put?
And I've always just put teacher.
I was a teacher at Portland Community College,
and I was a teacher at Oregon Institute of Technology,
and I am professorial in my presentations and in my style.
I bring big dad energy to everything that I do.
And in doing that, I am trying to make you successful
because my teacher's the ones that I remember,
the teachers that I still talk to now,
40 plus years out of school,
the teachers that came to my wedding,
the ones that you see at the shopping mart,
this grocery store, and you're like, oh my God,
Mrs. Johnson, you said a thing in 1984
and you changed my life.
That's who I wanna be as a DevRel,
is I wanna be their favorite teacher.
And in doing that, I will advocate for their best interests,
and in doing that, the bug will get fixed
or the product will get shipped.
That's a very sweet sentiment, I like that a lot.
I say I'm a web developer like Tim Berners-Lieses,
so I aspire to be like him.
Yeah, I think a big part of the role is definitely education.
I think the advocacy part is also pretty critical,
which we sort of talked about in the first question.
Yeah, there's a lot of questions
that are sort of being asked there,
the state of the industry is really interesting.
It's in a place that we haven't seen in a couple years,
probably at least the last 20 at this point,
and it's hard to say.
I think the advice that I tend to give
if you're new to the industry is,
how do you help your potential future employers
understand who you are, what you know, how you learn,
and sort of on this thread of teaching, how you teach.
So I always give the advice of if you're new to learning,
if you're in a boot camp, or if you're learning in school,
write a blog, keep a blog.
For me, that was public notes that I was taking
when I was still in college and I was an intern.
And I wrote about when Kraft CMS was new,
how to set up Kraft CMS on your Mac,
and then I was able to talk to the developers of Kraft,
and they ended up linking to my little intern blog post
in their official documentation
because nobody else had written about it yet.
So I was like, wow, this is cool.
And that, for me, is also how I got really involved in community,
which is the second thing.
Keeping a blog is important because it helps you to kind
of remember what you learned, to kind of teach people what you know
and to learn it better because as you write or as you teach,
you remember a concept more.
So if you're learning, that's a great way to kind
of showcase your skills, but also learn skills.
And then also it helps you become a part of the web community
because when you start sharing your experiences,
your demos, your explorations,
that's when you have something to talk about with other people.
And I feel like the community is so important.
I've gotten all the jobs I've ever had
in my career somehow through the community.
And it was either from meetups,
where I remember being an intern going up to the speaker,
being like, I want to work for your company,
how do I do that?
You're too young next year,
here's some books for you in the meantime.
So I think that that's the thing that you can do
is really get to know people, embed yourself a community.
And it's a lot easier to find a job when there's a person there,
it's not just like a resume that's being floated into the ether.
So yeah, it's a tough time in the tech community,
but there's still things that you can do to make yourself stand out
and become part of the ecosystem around you.
And those cross-bivertual communities,
it doesn't have to be in person,
although I think meetups are great.
There's a lot of stuff you can do out there.
Yeah, I love that a lot.
I want to piggyback on that
because I have recently been affected
by the shrinking devraille space
and the fact that things cost more all of a sudden.
I think it's an interesting time
because with the zero interest rate phenomenon,
you get paid to learn,
and now you really have to bet on yourself.
You are the one who's paying and investing in those things.
And I agree with everything you said about creating
and showing people what you're about,
a friend of mine called Content Interviewing at Scale.
And it gives you an opportunity to put yourself out in front of people
who are going to be making decisions about you in the future,
whether that's to accept your conference talk
or to invite you on the team.
Or to...
There's so many opportunities
where people want to see you just communicating an idea.
And when you do that well, it's magnificent.
Could I impose upon you to expand on the definition
of what the zero interest rate phenomenon is?
Because I don't think everyone knows that meme.
Yeah, sure.
Je vais expliquer le meilleur que je peux
parce que je ne sais pas vraiment le meme.
Mais pour un long temps, le ménage est libre,
les investissements sont bas.
Il y avait beaucoup de ménage à faire,
et de l'investir en...
Les bêtes étaient peut-être de plus en plus risques.
Résearche et développement.
Résearche et développement.
En mon cas, ce bête était mon enjeu.
En emploiant en personne,
en faisant sens à la ligne basse de la compagnie.
Et...
Je pense que...
Je ne sais pas.
Je voudrais entendre ce que vous pensez.
Mais partie de ce qui est là, c'est qui nous sommes.
Nous avons ce qui nous veut être un part de la communauté
et engager de cette façon,
et se réunir de la façon dont nous nous sommes imprés
par les gens qui nous ont fait beaucoup.
Je pense que c'est un côté de la plage de Deverelle
que les gens peuvent aller dans le monde
et se faire parler en face des gens.
C'est vraiment cool.
Je pense que c'est...
C'est partie de ce qui est là.
Je suis excité pour les gens qui nous pressent maintenant.
Je pense que ce sont les gens qui...
...mènent l'amélioration de la façon dont nous aimons.
Mais je ne sais pas.
Peut-être que je vais faire des choses.
Je pense que c'est aussi intéressant de le dire.
Ce serait un cas de carrière.
Mais les gens qui font le plus de leur pensée
ne sont pas sales et ne sont pas marketing.
Je suis le programme de la people.
Je suis là pour aider.
Je n'aime pas que je fasse des choses
pour une grande company.
Je ne suis pas excité à faire des marketing.
Mais je me déçois comme un bébé à la nuit
parce que je suis éducé et j'ai l'amélioration
et si un sale marketing s'est annoncé
par virtue de mes actions,
c'est comme si je me déçais comme un coq
dans la machine.
Mais je suis vraiment excité à les coqs
que j'ai vus vendre.
Je l'ai utilisé dans l'exemple de la chambre
des gens qui mettent sur votre porte
pour convertir vous dans leur religion.
J'adore ça pour eux.
Vous avez entendu la news?
Je suis excité à mes choses.
Je suis excité à la plateforme.
Je suis excité à la web.
J'ai été éducé depuis le jump.
Je ne vais jamais arrêter.
Si vous me voyez me faire fêter
ou me laisser mon company,
c'est parce que je suis rétérent
ou ils ne m'ont pas voulu.
Ça fait du sens?
Et cette authentité
m'a permis de me détenir
que je suis un sale marketing.
C'est une fonction.
OKR, etc.
Je veux discuter un peu.
Je vous le dis.
Je pense que le part que je discute
est que
DevRel est plus étendue
à la chambre des sales.
Je faisais ce type de travail
avant que je m'aie officially
dévoilé DevRel pour plusieurs années.
Je parle de CSS & UI Web Conferences
et oui, c'est ça que je voulais.
C'est une éducation.
C'est un éducateur.
Je ne vois pas ça
comme tenter de vendre quelqu'un
sur un futur.
Je pense que c'est important.
Je veux que vous le connaissiez.
C'est une option.
Ça ne s'est pas fait.
Tu as entendu la news
sur la plateforme web?
Oui, mais je ne vois pas ça.
Tu es vendu en Chromium.
Non, pas vraiment.
Je ne parle pas de
c'est une feature de Chromium.
Je parle de la plateforme web.
Je ne parle pas de features Chromium.
Je ne peux pas attendre.
Je parle de base-line.
C'est une plateforme web.
Je vais te demander ce que se passe
quand la machine de la capitale
se dit que
on parle un peu plus de Chromium.
C'est un bon travail.
Je pense que c'est un bon travail.
C'est pourquoi je ne fais pas de DevRel.
Je pense que DevRel est très différent
selon où vous êtes.
Je pense que
ce que je vois,
le travail que nous faisons
c'est très bien
sur la web web
et sur la façon dont les développeurs
ont les tools qu'ils ont besoin
pour arriver.
Mais pas tous les jobs de DevRel
sont comme ça.
C'est très vrai.
Je pense que nous sommes à 95% d'accord.
Non, c'est pas vrai.
Je suis vraiment en train de compromise.
Je ne veux pas avoir une bifre avec vous.
Non, une bifre.
Je suis d'accord,
être en train d'être en train de cataloguer
une petite compagnie,
parce que vous êtes tous
des compagnies gigantesques
de la mega.
Comme vous l'avez dit,
c'est un rôle différent.
Vous pouvez se séparer
du travail
en train de...
Il y a des gens qui se voient
en Chromium.
Il y a des personnes qui se portent
pour des expériences
développeurs.
Je pense que,
mais en plus, comme un point de vue, dans les petites entreprises,
ça fait que cette ligne est très hausse,
et vous devez vraiment savoir ce que vous avez appris dans la porte
pour savoir ce que vous allez laisser sur le marché.
Je ne pense pas que c'est une petite ou une grande company,
mais une grande company, vous avez aussi de Vervell qui est
focussé sur le sale de votre produit,
surtout quand vous avez des taux de paiement de choses,
ce qui est totalement bon, mais je vous accueille,
je pense que ça dépend de votre rôle,
de savoir ce que vous avez appris,
et de l'expectation qui est alignée avec votre intérêt.
Il n'y a rien d'inquiétant avec un position de dévrel
qui est plus sales, mais c'est un peu différent,
que je pense que certains personnes regardent le dévrel.
Il dépend de... Il y a beaucoup de choses, il dépend.
Et aussi, basé sur le nombre de la company et la longevity de votre carrière,
et votre niveau, il y a des choses qu'on ne parle pas de,
qui sont le privilège de niveau.
Vous êtes privilège comme un ingénieur, un staff ingénieur,
ce que vous pouvez faire avec ne pas parler de ça.
Je vais donner une présentation aujourd'hui sur l'A.I.
Je travaille pour une grande company qui veut que tout le monde
utilise l'A.I., et ils veulent que l'A.I. mette,
pour que le ménage s'entend.
Mais il y a des choses que je désagris,
donc comment je parle de choses que j'aime,
et ne pas être fiers pour les choses que je n'aime pas.
Donc, vous devez tenir vos opinions.
Je n'ai pas été en dévrel à Google,
où je n'ai pas appris avec le produit,
et j'ai été en pensant,
je n'ai pas les idées pour les mieux.
C'est mon design doc,
je peux apporter,
et ce n'est pas ce que j'ai été fiers pour.
J'ai fini de laisser le rôle,
et de rejoindre la team Chrome,
et de faire attention à un team qui a voulu
ce feedback de la développe,
qui a voulu avoir l'input de l'appliquer
des développeurs.
J'ai trouvé une différence d'un peu,
dans le genre de rôle que je vois
d'un devrel comme une équipe de marché,
et un rôle que je vois de devrel
comme une partie de la processusée.
Je dois parler de ça,
parce que je pense que vous êtes bien,
et c'est important de savoir,
et de réclairer ça,
parce que je suis en train de rejoindre
la team Chrome,
et le chargateur que je suis donné
était de faire,
connecter le livre de la histoire avec les développeurs.
Parce que c'est vraiment populaire
avec des systèmes de design,
et des libraries de component,
on veut en faire un plus
développé en espace.
Je pense qu'il y a des tools très intéressants.
Ce que je n'ai pas réalisé,
c'est que l'intent était de prendre
ceci aux développeurs,
et de ne pas
mettre ce que les développeurs veulent
dans le produit.
Et ce,
de savoir où vous vous prenez,
c'est important pour vous,
et pour tout le monde.
Si vous pensez que votre travail
est de faire ce que les développeurs veulent,
on le va faire,
et ce n'est pas que tout le monde est malheureux.
C'est pourquoi je me suis dit
que whenever someone is marketing,
DevRel should be a part of product,
it can have a go to market aspect,
which product is involved with go to market,
but I don't know,
at least I don't think of it as more aligned with sales,
or it should be,
and maybe that's why DevRel is getting a worse rep today
than I think we have in the past.
Well, I think we smell it.
DevRel is known to smell marketing
and go, it doesn't smell good.
Totally, totally.
So I didn't mean to imply that it was a sales
marketing function.
I am saying though that one is fooling themselves
if they think they have no relation
to sales and marketing.
But it's a three-legged stool.
Yeah, totally, totally.
Yeah, definitely.
On the product side,
I really love working with
some of the folks that I work with at Microsoft
who are like, I had no idea.
You found a gold mine of feedback.
People who are really jazzed about good feedback,
because it's like, this is how we thought it would be used
in production, and this is how people actually use it
in production, and we had no idea.
So whenever I, there's all the data,
there's telemetry and that kind of stuff,
but when I discover what I've always called
dark matter developers, dark matter is this,
you know what I mean?
Good term.
I've been talking about this for 20 years.
Dark matter is the space between the stars.
It can't be seen, it can't be measured.
For every one person who's like WTF
Microsoft with a dollar sign on Twitter,
there's a thousand people
that are not on Twitter who are not having those bugs,
or they're not talking about those bugs.
They're anonymous redditors,
they're on Discord, they're in the dark web.
And Twitter is not product feedback.
And we've got all these young,
emerging deverelles
get a couple hundred thousand followers on Twitter
by asking open-ended questions
about what's your favorite JavaScript framework.
And then they go viral,
and they call that devrel, and it's like, no,
go and find somebody. Talk to the person on the BART
that you overheard using
whatever framework
and go and visit them.
Do call down, show up at their house.
Have you heard the news?
That's devrel.
Yes, snaps for that.
Thank you, I got snaps, now we're...
So we live in
kind of turbulent times for developers,
like there's this new thing called AI
that's doing a lot of things,
and it's changing how we learn how to code
and how we learn new products.
So how has the way
developers learn, change throughout your careers,
and how do you think it's going to change in the next five?
That's you, that's you.
Stretch it out before you.
Get ready.
When I was a boy,
when I was a boy,
when I was sitting
on my BBS
on my bulletin board system
with my ex modem and my
my US robotics modem and my parents basement,
I would have a question,
and there were only two books, right?
There was like the KNRC book,
and then there was the assembly language book,
and that was the sum of all human knowledge,
and there was no other available information.
So I could log into a BBS and leave a note,
and then in like six to eight weeks,
someone would find that question.
There was no search, there was no metacrawler,
there was no altavista,
like Lycos was not a thing,
there was no, you were alone.
There was just this profound sense of grind it out.
And then community happened
and compuserve and forums and use net,
and the idea that you,
if you choose to be, you are never alone.
You know, you will always go
and Google with Bing to find the question
and then discover that it was you who asked it years ago
and answered your own question.
That sense of never being alone is really amazing.
Now in the AI world,
you have an opportunity
to have an infinitely patient friend
to ask questions of
at two in the morning on a Tuesday
and interview them.
In my opinion, too many people are trying to get the AI
to write the perfect for loop,
and oh, you didn't write it right,
and then ask again and then ask again.
I interview, I want to show this in my talk,
I interview co-pilot
and I do it with my voice, I do text to speech,
and I just have conversations,
and at the end, it's written a book,
and it's 85% correct
and confidently wrong in other places,
but the idea that I could have a tutor
with me available,
and whether it be co-pilot,
which costs money, or continue.dev,
which is free and local
and can be run in airplane mode,
you can have whole conversations about your code with it,
means that rubber ducking,
you familiar with rubber ducking, right,
is now a real thing,
and you can vibe with a pair programmer.
That's a thing that we have yet to explore,
is an infinite book
of an infinitely patient tutor
that I can ask questions about my code to.
I love the framing that you give
for AI as a rubber duck,
because I think that there's a lot of potential there,
especially when you think of it as a co-pilot,
as a assistant, as a friend.
The thing that I worry about with AI a little bit,
is especially newer developers
who haven't fully, you know,
learned a concept yet,
just taking AI, you know,
what it says verbatim.
And then when you get to that point
where you're talking about, you have it 85% written,
you don't know how to debug what's wrong.
Like how do you find the thing that is missing
when it's valid code,
it just doesn't really work for what you're trying to do,
and that's kind of like, you know,
the next phase, I think,
and also just as a side note,
because I work in the CSS space,
AI is particularly not good at CSS.
So we have a ways to go,
which I think, you know, there's things
that we can do to improve code quality across the platform.
Ok, so I have a question for you then.

So I've used this analogy for years, I'll run it by you.
What do you do when Uber and Lyft exist,
but I want people to learn
to drive a stick shift
because it will make them better drivers?
Change your oil, change your tire, drive stick once.
I'm not trying to gate keep cars.
I want everyone to have transportation.
But when two people are sitting on the outside,
you know, like sitting on the side of the road
with their broken self-driving vehicle,
and they have no idea what a wheel is.
I, as old man who shakes fist at cloud,
keep finding myself saying,
let's go just find the layer that you operate at
and go one layer below.
Just go slightly out of your comfort zone.
I'm not saying, you know,
you don't get a garbage collector until you write assembler.
I'm not saying you don't get to use Lyft
until you learn how to drive a stick.
But I am saying, get a little uncomfortable
and go a layer below.
Oh, I hear you. I'm definitely not, you know,
not saying you need to learn the basics
before you can use AI.
But I'm never debugging my Uber.
I'm never back there trying to fix the engine.
You know, it's not my job to drive someone
from point A to point B if I'm taking an Uber.
That's someone else's job.
They better know how to fix their engine if they break down.
Yeah, exactly. As the driver, as the developer.
I just think that you can get yourself in a place
where you can get stuck a little bit more easily.
Which is a challenge to solve.
Oh, man, you both looked at me.
We've been talking a lot.
I like the...
I've always liked that idea of going one level deeper.
I just watched Kung Fu Panda 4 with my kids.
And there's this...
I wonder where this is.
One level deeper.
No, there's a scene where...
Oh, man.
Okay, there's another movie.
There's another movie that...
I'm going to spoil it anyway because it's the same scene.
But there's another movie.
Shoot, Big Hero 6.
And someone's caught in a cage
et ils sont tentant de se faire sortir.
Mais c'est trop difficile pour eux.
Ils ont réalisé qu'ils ont suffisamment de force
pour aller sous la cage,
aller au sol, à la place de la place de la place.
Et je pense que ça a toujours été...
C'est génial, parce que tu ne peux pas
se faire sortir dans la trappe que tu es dans.
Mais si tu peux se faire sortir,
c'est le moyen de la sortie.
Et...
Je ne sais pas.
Je pense que je dois avoir des pensées de la AI.
Je ne sais pas si c'est le bon format pour ça.
Mais je ne...
On va se faire des trinques.
Je ne suis pas affrayé de l'idée
d'une intelligence séparée de l'humain
et d'une émission de notre vie.
Ce n'est pas quelque chose qui me terrifie.
C'est là que la AI n'est pas la AI.
La AI est juste...
Si vous et je connaissons les autres longs,
nous allons finir les autres...
Sandwiches.
Ah, je l'ai mis.
Je l'ai halluciné.
Vous savez ce que je dis?
Je vais juste vous rappeler
que c'est un couteau de rouleau
parce que vous parlez de vous-même.
Il n'y a personne d'autre là-bas.
Vous êtes en train de parler de vous-même.
Un des gens m'a dit que je suis un preuve
et que je suis un preuve pour l'apocalypse.
Et je l'ai preuve par downloader
le mystère, un model de langage
de 8 gig.
Et je l'ai downloadé
tout le monde internet.
Et maintenant je peux juste demander des questions en ligne.
Je suis comme... Ah, c'est le pire.
Si vous êtes preuve,
vous pouvez downloader le wikipedia.
Mais 8 gigs ne vous donnera pas de rien.
Imaginez un monde
qui a été rébuildé après l'apocalypse
sur un de ces
petits models de langage.
C'est sur votre carte d'invidia.
Parlez-vous de toutes les informations
que vous avez confié que ce soit spuit.
Et personne ne va savoir ce que c'est.
Je pense que c'est tout un halluciné.
Et parfois c'est correct.
Comment nous avons été là?
Je pense que
mon préféré pour AI
est de faire quelque chose que je sais
de la manière de la faire
mais c'est...
Vous devez faire la code.
Je pense que le talk
était bien, parce que c'est pas
prototype en interface.
AI c'est juste de la logique.
Je ne suis pas showcase.
Je suis showcase dans le UI.
C'est Yadiyadiyad.
C'est de Schinefeld.
Quand vous faites quelque chose comme ça.
C'est ce que je pense que ça s'applique.
C'est comme la lait de la lait de la lait.
Vous savez comment faire ça.
Ça vous prend un moment.
Qu'est-ce que vous faites maintenant
où on a des essayants
en échecant et en pleine de travail.
Et ils vont dire qu'on va
garder la connaissance à ce moment.
Vous devez changer le prompt.
Pourquoi ils font des essayants?
Ils doivent faire quelque chose qui
intervient plus de pensées.
Vous ne pouvez pas juste
faire quelque chose de AI.
C'est intéressant.
Je pense que, en regardant le
basis de la tool,
je vous accorde que
il y a une séparation
que nous commençons à voir
qui est vraiment fun
où les compétences sont des compétences
et les humains sont des humains.
Nous avons essayé de
être des compétences

et
en parlant de les compétences
et en parlant de la langue des compétences
et je pense que
l'AI nous permet
de participer
dans l'expérience humaine
d'offloader plus de cela.
Et nous avons mieux de communiquer
aux compétences
à ce niveau.
Je suis excité.
Il y a beaucoup de choses qui sont
très difficiles pour nous
et qui nous ont appris de faire
des choses que j'aime.
J'aime les vidéos et tout.
Mais je veux
faire des expériences
comme possible.
J'ai spent huit heures
en faisant des transcripts.
Et maintenant, je peux le whisper.
Et dans 5 minutes,
j'ai un transcript
très bon.
Et ça me permet de être
un humain.
Je l'aime.
Mais on n'est pas là.
Tout est super, c'est aspirant,
optimiste et avec vous.
C'est le but.
Mais quelqu'un a dit que je m'ai
brûlé un forestier
pour faire un vidéo de Hamster.
Il est en espace.
C'est génial.
Comment ça a été utilisé ?
Il a utilisé assez de pouvoir pour
l'infos pour sa vie.
C'est beaucoup de pouvoir.
Les textes de l'A.I. sont cool,
mais je ne veux pas qu'ils soient artistes.
Je vais dire, j'ai un ami
qui est un designur. Il est un grand designur
dans l'espace. Il a eu un tweet
qui m'a pris le temps.
J'étais très excité pour l'A.I.
pour prendre les tasks menuels.
Je vais faire des tasks créatifs
et faire des arts.
Et faire des choses fun.
Et ça me fait que l'A.I.
est en train de faire des arts.
C'est faire des vidéos de musique.
C'est en train de faire des tasks créatifs.
Il y a un
intéressant...
La façon dont les gens parlent de l'art
assisté à l'A.I.
est intéressant, car il est parlé
comme un bruit.
C'est un aspect créatif.
Mais il est très bas.
Qu'est-ce que le copyright ?
Comment vous le faites ?
C'est un tout autre sujet.
On est presque au temps.
Vous savez comme la Chine.
Maintenant, tous pour que vous
parliez à un film,
loyalité à ce qui spoke...
2D page,
c'est le peanutons en mai !
Pour treatments,
on commence l'ombre.
Merci.