Latest / Bad Decisions Podcast / #23 Adobe's AI Revolution: Firefly and Generative Tools with John Nack | Bad Decisions Podcast
Transcript
- 0:00This stuff's moving so fast nobody even knows what it can
- 0:02do. And then Dolly hit and it was
- 0:03like, Oh my God, you are not gonna pry bar me out of here,
- 0:06Firefly. It it sounds like a model, but
- 0:09of course, in reality it's many different models.
- 0:11Are we gonna have fireflies stand alone as a product or know
- 0:15if you're gonna have Firefly inbidity to every single product
- 0:18suit that I do be has or this is still a big question that you
- 0:20guys are figuring out? I don't know that there's a
- 0:23world where the exact same number of what we currently
- 0:25consider creative jobs stays the same.
- 0:28The value I would provide is trying.
- 0:30To be a good storyteller. How big is the A I team
- 0:33currently at Adobe? Adobe can or should try to do
- 0:36everything. I think that's an insane
- 0:38approach. He's been in Adobe since, I want
- 0:40to say, probably 1985, maybe even earlier.
- 0:44And we didn't solve a problem. We avoided a problem.
- 0:47If the way we get there best is a lot of people come at it and
- 0:50they compete with each other, they compete with us and vice
- 0:52versa. Great, Let's do it. 321 Let's go
- 0:57everybody. Welcome back to the Bad
- 0:59Decisions Podcast, where we give you guys a backstage pass into
- 1:02the minds and brains of creatives, entrepreneurs and
- 1:04visionaries from all domains and today.
- 1:07We are going to the world of artificial intelligence and it's
- 1:10special because we have the principal product manager of
- 1:13Adobe here with us today for the most.
- 1:15Part of the past 23 years he's been with Adobe working on
- 1:18critical components that you probably have experience with,
- 1:21like Adobe Camera, RAW, Smart Objects content.
- 1:24The Wear feel and he was also inducted into the Photoshop Hall
- 1:28of Fame. And today he's playing a major
- 1:31role in the Adobe AI team. We have John Nack in the house
- 1:34today. John Nack.
- 1:35What's up? Guys, thank you so much.
- 1:37That is that is too kind of intro and and I mean that
- 1:40literally because as as you may guess at a big company like
- 1:44Adobe there are a lot of people directing the vision and and so
- 1:48I say this with no false modesty.
- 1:50It's funny because my friend Deke.
- 1:52Did a interview with me a couple months ago and he put on the the
- 1:58poster screen John Nak, head of a I at Adobe and I was like Oh
- 2:01my God like I will never I will never live this down with all
- 2:05the people who are way above me in the org chart.
- 2:07So I am but I am very lucky I I am.
- 2:11I'm just you know one of the one of the product managers.
- 2:14It is one of those like it really takes a village kind of
- 2:17things and. I mean it.
- 2:20It's really been a fun ride. I mean, we get to see everything
- 2:23from like data gathering and annotation to training to all
- 2:27the folks who are doing infrastructure.
- 2:29There's people building, you know, user interfaces, there's
- 2:32people doing the pricing, getting good deals with cloud
- 2:36compute. So I I just say all that by way
- 2:39of like very genuinely humble intro and and say, yeah, where
- 2:45where I get to sit is trying to help.
- 2:48People make the world more beautiful and fun.
- 2:50Like that's kind of what it boils down to.
- 2:53And and you're right. You know when when you mention
- 2:55passion, it's interesting. One of the one of the guys who
- 2:59really inspires me is named Russell Brown.
- 3:01He's been Adobe since I want to say, probably 1985 maybe even
- 3:06earlier. And you know it.
- 3:08It's like if I had one Flex, if I could say, yeah, I there were
- 3:12four names on the Photoshop 10 splash screen and one of them is
- 3:15mine. I I you know that.
- 3:18That would be true if I were Russell.
- 3:20But but what's what's cool when when you see him and and other
- 3:25people like him is, you know, even though they've kind of done
- 3:28everything by this point, they're still so wildly
- 3:33enthusiastic. And I I feel like for a lot of
- 3:36us, I mean, my wife included, I'm, I'm sitting at her desk up
- 3:41in Colorado. She worked at Adobe for 12 years
- 3:44and before Dolly came out last year.
- 3:47I think we were both kind of in this, I don't know, lull.
- 3:51It was sort of like, yeah, I came back to Adobe to work on AI
- 3:54first creativity. I'd been at Google.
- 3:56I met guys like Bilavel, which was really inspiring, But it it
- 4:00felt like things had just kind of flattened out.
- 4:02Everything was getting like, really increbattle.
- 4:04And it's like, yeah, well, you know, here's another keyboard
- 4:07shortcut for you or something. And it's like, I don't know,
- 4:09couldn't I have done this 20 years ago and a copy of of my
- 4:12tools? And then Dolly hit and it was
- 4:14like, Oh my God, you are not going to pry bar me out of here.
- 4:19This is like, I hope I still have some runway left and in it,
- 4:23right, right and and what's so cool And just to circle back
- 4:27quickly to to Russell, obviously he was around.
- 4:30You know, when if you ever see those like original Photoshop
- 4:33demos, it's like, you know, black and white in in some
- 4:36cases, you know, everything takes 5 minutes to run a
- 4:39Gaussian blur or whatever. And so he remembers what a big
- 4:43deal it was when layers came out, you know, before layers.
- 4:46You guys I'm sure are too young. But it was like to do a drop
- 4:50shadow was like this pro operation, you know, it was like
- 4:54because it was like people don't remember this, but I mean, for
- 4:57the first eight years of Photoshop, there was one level
- 5:00of undue, right. So like, if you did something
- 5:02right and then you, like clicked away, I know, like, so you just
- 5:05like you did something and deselected.
- 5:07That's it. Because all you could do is like
- 5:09reselect but you couldn't even go back one more step and you
- 5:12know text wasn't editable. So if you wanted to make text
- 5:15with a drop shadow it was like it you were like a high wire
- 5:18Acrobat. You know it's like OK, right.
- 5:22Like OK, don't don't screw up. Like open this up try this OK
- 5:25you know Command J float the thing and and so for for Russell
- 5:29he's he's so so wildly excited. You can see it in these videos.
- 5:34I should I should share some with you guys so you can.
- 5:37Put him, put him out for folks who want him.
- 5:39But yeah he's like this is like when layers hit and that was 30
- 5:43years ago, you know and and I feel the same.
- 5:45It is, it is that kind of step change and and what's crazy is
- 5:49that we're just a year into it. I mean, not even a year into
- 5:52Adobe. We're like 3 months in, and I
- 5:54was tweeting the other day, like, I'm old enough to remember
- 5:56when we launched Generative Phil, right?
- 5:59That's March 2020. Exactly, exactly.
- 6:02And so and people just these last days have been posting
- 6:05comparisons with Mid Journey from like July of 2022 versus
- 6:10July 2023 and it's it's insane. Of course a year ago we thought
- 6:13that was insane and it was because you've never seen
- 6:15anything like it and yet now it's like.
- 6:18People are doing zooms in on like the corneas of people who
- 6:22don't exist. And so yeah, I I I brought my my
- 6:26soda here because of course I get so excited, you know talking
- 6:30I I like and it's also very thin air up here I should note.
- 6:34So I will certainly certainly run out of steam and but that's
- 6:37good because then I can let you guys get a word in.
- 6:39So that's my intro. So many so many beautiful things
- 6:44that I can just touch on based on what you just mentioned.
- 6:47The one thing I want to talk about is, and I think a lot of
- 6:51people will have questions based on based on this, before we talk
- 6:53about your passion and the concept of burnout, which is
- 6:55something I want to cover because clearly after working
- 6:57for so many years, you're more excited than ever before, which
- 7:01you don't tend to see in a lot of people.
- 7:03We'll get back to this. Can we just go back to the very
- 7:05beginning? You mentioned there's a lot of
- 7:07people working, you know, at Adobe right now, specifically on
- 7:09A I, what is the numbers like? Maybe not an exact number, but
- 7:13how big is the A I team currently at Adobe?
- 7:16Well, it's a really good question and I don't actually
- 7:20have an answer. Well, reasonable and I can make
- 7:23some guesses, but but part of where it's hard to answer and I
- 7:27think this is actually a good problem to have, is there's so
- 7:30many teams now who are kind of rethinking their Rd. maps and
- 7:35saying, well, of course, you know, we.
- 7:38Customer needs many times don't really change.
- 7:40It's like, you know, what's the process of telling a good story
- 7:43or like how would you structure a film shoot or you know, how
- 7:46could you convey your ideas in an illustrated way, right.
- 7:50A lot of the stuff goes back to Aristotle.
- 7:52You know, you can look at the arc of like what makes good
- 7:55tension and so. So those things don't
- 7:58necessarily change. Obviously what's changed is our
- 8:00ability to. Provide tools and structures
- 8:05that people could then use. And and even though some are
- 8:08still outside our grasp, you know, I'm sure you guys see
- 8:11where people are synthesizing video strictly from text.
- 8:14And there's all of these what we call multimodal inputs where you
- 8:18know, you can mix in sketches, you can mix in photos, depth
- 8:22video content, 3D is is extremely exciting.
- 8:27And so because all that is moving so quickly, I think
- 8:29you're seeing a lot of teams say, well, yeah, of course we're
- 8:32not going to forget. The basics because you know just
- 8:35to get excited and only do a I would be would be crazy.
- 8:38But but they're also spinning up a lot of little internal
- 8:41prototypes and saying like well gosh okay if Photoshop can do
- 8:44that, you know what can we do over here.
- 8:46By the way, one thing that was always amusing to me back in the
- 8:51day when I worked on on Photoshop full time was the
- 8:53After Effects team who I love these I mean if you want to talk
- 8:56about people who are like. An incredible ratio between
- 8:58like, accomplishment and humility.
- 9:00Like looked over. They're like incredibly down to
- 9:02earth. Very sweet, sweet folks.
- 9:06But I would go on stage internally in demos and I'm
- 9:08like, hey, check out this cool Photoshop thing.
- 9:10You know, people would applaud. It was very exciting.
- 9:12And then like 2 seconds later, the After Effects folks would
- 9:14come up and they're like, hey, that's really cool, but what if
- 9:16it moved? And I'm like, Oh my God.
- 9:20Like. Can we just have like a second
- 9:22day for video and then and then I could just.
- 9:24Not have it on the same day. Yeah, there you.
- 9:26Go It's so cold blooded and and they would come right up after
- 9:29me and anyway so so I I say that just because these folks and
- 9:34others are thinking like what is the next evolution?
- 9:37Like how do you make things temporarily coherent right.
- 9:40Like a lot of the cool video stuff we've seen today it is is
- 9:43neat but. You know it's a lot of like
- 9:46flickering and and it's a novel effect but of course that kind
- 9:48of burns out just like you know like paint box effects in the
- 9:5180s it was like Oh my God like a little squiggle went by but but
- 9:56of course then it's you can only do that so many times and so so
- 10:00anyway yes it's extremely longwinded way of saying a lot
- 10:03of product teams are are thinking about how they should
- 10:06intersect with a I then there's also just a lot of growth in
- 10:10teams building the infrastructure so.
- 10:13I mean when we say Firefly it, it sounds like a model but of
- 10:16course in reality it's many different models.
- 10:18You know there's there's like the core model and that's
- 10:22evolving and and so there's like various versions we're trying
- 10:25like well you know if we constrain it in this way you
- 10:29know if we can only access these data sources, how should we do
- 10:31that or is there data we can add that would make it better?
- 10:34How would you know that takes a while.
- 10:35Then there's things like multi diffusion which is what
- 10:38generative fill uses and that conditions on surrounding
- 10:41pixels. And then there's things where
- 10:44people are playing with technologies like Controlnet
- 10:46where you can use you know edges, you can use depth, you
- 10:49can use these other signals and then of course there's how do
- 10:52they combine and it's like Oh well could you do multi
- 10:55diffusion with depth and I don't know could we and and so let's
- 11:00try right and and as a result there are there are so many
- 11:04promising ideas that are all kind of competing for bandwidth.
- 11:08Literal, you know bandwidth just in terms of like how many GPUs
- 11:11can we get and how can we train efficiently and then sort of
- 11:15just engine testing bandwidth because you know when we we do
- 11:19these things one of the really important dimensions is making
- 11:22sure they're fair. And and by fair you know we mean
- 11:25sort of aligns with the values of of the company and with with
- 11:29you know most of the creative people we meet where you know we
- 11:33want to make sure that the quality of the edit is.
- 11:37Equally good whether you know you're young or old, light skin,
- 11:40dark skin, heavy, thin, whatever the the human diversity might
- 11:44be, we want to do a good job for everybody, right?
- 11:47And so that takes a ton of time. You know, back when I was at
- 11:50Google, we were working on meshes for faces.
- 11:53In fact, just a side note, I saw my my Google teammates put out
- 11:57some new stuff on that yesterday where where, like they're
- 11:59feeding the meshes and I'm like, Oh my God guys, I'm on vacation.
- 12:03Like am am I supposed to be reading these GitHub anyway?
- 12:06Leave that aside. So, so you know back in the day
- 12:09being like four years ago, right, ancient history we were
- 12:13working on mesh fitting for for faces and you know there again
- 12:18it seems like a simple thing but we really wanted to take the
- 12:21time to say okay, you know can we kind of get these big data
- 12:25sets and slice by geography which is in a perfect.
- 12:30Measure but at least it's a proxy for you know differing
- 12:33appearances among humans and and make sure that like the error
- 12:37rate we detect is within the error rate of of a normal human
- 12:42like hand annotating something and and all of that stuff goes
- 12:45on behind the scenes. It's easy to skip for sure and
- 12:49it's one of the reasons that it's it's not hard probably for
- 12:52other people to bring things to market quickly it's it's not a
- 12:55criticism it's just saying. Really spending the time on that
- 12:59long tail of of development and testing is really important and
- 13:03it is one of the key, the key things that has to happen before
- 13:06you like roll something out in Photoshop.
- 13:09But anyway, that's one of the many reasons that there's an
- 13:11awful lot of people who are now working really hard on all these
- 13:15pieces of the puzzle. But having all these A I tools
- 13:19coming in and with all the research that you guys are
- 13:21doing, are you guys reframing the product suits as well?
- 13:25Like how is going to look in the future?
- 13:26Are we going to have Firefly stand alone as a product or know
- 13:31if you're going to have Firefly embedded into every single
- 13:33product suit that Adobe has or this is still a big question
- 13:36that you guys are figuring? Out.
- 13:38It is a big question. It's a great question.
- 13:40You know The thing is obviously in order to.
- 13:44To get to keep playing this game right where that is, you know, I
- 13:48get to keep coming to work and working with these awesome
- 13:50people. We have to make some money on it
- 13:51and make sure it's like a real business.
- 13:54And therefore we have to align with what we believe people care
- 13:58about and are willing to pay for.
- 14:00And we have to do it in a way that's thoughtful, where, you
- 14:03know, it's not like we've traded dollars for pennies and it's
- 14:06like, hey, good job, like here's an amazing Photoshop feature,
- 14:08but it's also over here for free, you know, like I I think.
- 14:13It's a very solvable problem, right?
- 14:14There are really good reasons some people go to Photoshop and
- 14:18there's really good reasons some don't.
- 14:20There's tools like Adobe Express, you know.
- 14:23Again, this is a tangent but we were at a 4th of July thing
- 14:27yesterday, met a super interesting guy at the airport
- 14:30who had just flown in. He he does like remote undersea
- 14:34vehicle stuff like super random, but.
- 14:38Obviously a pretty tucky guy. And he was trying to show us
- 14:41some some photos he had from one of these adventures with the
- 14:43Discovery Channel. And I was sitting there quietly
- 14:46and he's using his phone and he's like scrolling and
- 14:49scrolling. And I felt like Doctor
- 14:52Strangelove where, like, my hand keeps like trying to fly up and
- 14:55like, grab the phone and like, oh, but I have to be polite.
- 14:59I don't know this guy and I'm like, oh like, have you
- 15:02considered searching? And he's like and he's like,
- 15:05yeah, anyway, he just keeps going and I'm like, I'm like,
- 15:08well is that Apple or Google? Because Google, you know Photos
- 15:10is like really good at searching and you can back up for free and
- 15:14and and you know an apple like for indexes text, right.
- 15:17So even if there's like a word in in a sign in the background
- 15:20it'll often find it and and and and yet the guy wouldn't do it
- 15:25and and the same goes for editing.
- 15:26Right. And so.
- 15:28I I this is again, it's a very elliptical way of saying some
- 15:31people don't need or want to do that.
- 15:32And and I said, you know on a given day of all the people who
- 15:36use Google Photos, this is an old stat.
- 15:38But I I bet you it's still true. 97% do not ever hit the edit
- 15:41button, not once. And of the ones who hit it, they
- 15:45edit two to maybe three photos tops and those edits are
- 15:49basically the auto button and crop right.
- 15:52And so all the stuff that you know this is a.
- 15:55That's why it was weird for me to go to Google and it's weird.
- 15:57I I'm very stubborn which is why I lasted as long as I did.
- 15:59But but fundamentally, like if you're trying to make products
- 16:03for a billion people, it's OK. They they don't they don't need
- 16:05to want to do that. It's all right.
- 16:06Like if you can run an auto enhancement maybe sometimes you
- 16:09can make them a thing that's cool.
- 16:11It really doesn't matter compared to like the cold launch
- 16:14time of the Android app which I actually have the receipts on
- 16:17this. But I I think I got into all
- 16:20this just to say that. You know as we're thinking about
- 16:24the Adobe product line you want to make sure you're serving
- 16:29folks where they are right. You you want to give them an on
- 16:31ramp if if for the you know percentage of people who want to
- 16:35move from one product to another.
- 16:36You know maybe back in the day they would have started with
- 16:39Photoshop elements and then they move up to to full Photoshop but
- 16:43but I know I'm I'm way off topic to bring it back.
- 16:46I think what you'll see with the Firefly site is I'm I'm I'm I'm
- 16:51literally like. Playing hooky on my own vacation
- 16:53and writing up this, this vision for the team.
- 16:56Like, it's not to say it's the only mission.
- 16:58Maybe, maybe folks won't agree, but I think I think they will,
- 17:02insofar as we know we want to inspire, right?
- 17:05Like whether you're a pro, whether you're a newcomer you
- 17:09know, whether you're like a serious user or just you just
- 17:11want to have fun. This stuff's moving so fast,
- 17:13nobody even knows what it can do.
- 17:15So, like, let's let's welcome everybody in.
- 17:17Maybe they've never heard of. Firefly.
- 17:19They've never heard of Photoshop.
- 17:19They've never heard of Adobe. Or if they have, it's like, I
- 17:22don't know, I think my cousin uses that.
- 17:23She's a graphic designer. It sounds like really cool, but
- 17:25really hard, right? Like we we want to reframe and
- 17:28say no. Like you can use this tack to
- 17:30come in and do incredible things.
- 17:32If you want to do that as a creative pro, great, like let me
- 17:35show you some stuff you can use right away.
- 17:36And I'm not even talking like make a full image.
- 17:38I'm just talking like little bite sized pieces like recolor A
- 17:41vector or like maybe make a texture or just like change the
- 17:44aspect ratio on a photo. And then for for more, more
- 17:48newcomer folks like could this be almost more like a game?
- 17:52Like game's kind of a four letter word, you know, and you
- 17:54don't want to make it seem like, oh, this is we're doing Angry
- 17:56Birds or Snapchat. But but game in the sense of
- 17:59like things that are fun like Pictionary, you know, or
- 18:01tellustrations or charades or you know, fill in the blank and
- 18:05just give folks a really a really simple but but
- 18:09constructive way to kind of get their feet wet.
- 18:12And then. After all of that, to answer
- 18:14your question, I think there's still, I think there's room.
- 18:17There's plenty of room to put cool tech into the tools folks
- 18:21use today. There's also really a lot of
- 18:23white space, so to speak, around these new experiences.
- 18:27Like how does one get inspired? How do you manage the stuff you
- 18:30create? You know, you meet people who
- 18:31use Mid Journey, they're dropping 100 bucks a month in in
- 18:35Cloud Compute. They're generating thousands of
- 18:37images, which is delightful and amazing, but then they're like,
- 18:40Oh my God, I think I had like a. Was it a chicken made of armor
- 18:44and where did that go, you know, so, so all of these new, these
- 18:49new possibilities emerge and and I think we'll see a sort of a
- 18:53hybrid world where where we have new tech and new new product
- 18:56experiences and then hopefully if we do our job right does
- 18:59really dovetail with the the sort of creative ecosystem that
- 19:02lives in today's apps. First of all, can we just say
- 19:07congratulations to you and the team because you guys definitely
- 19:10knocked it out of the park. When Dolly and Mid Journey and
- 19:14you know Runway and all these came about, all of the stuff we
- 19:17were hearing from the community as people who are not part of
- 19:19Adobe was oh does the Photoshop killer this is going to like
- 19:23where is Adobe? What what what are they doing
- 19:25about this? Because clearly you guys are the
- 19:27ones that people would have wanted to see creatives would
- 19:29have wanted to be able to use these tools within Photoshop
- 19:32because exactly like you mentioned.
- 19:34The problem is if I want to use Mid Journey as great as great
- 19:37the tool is. You still have to go to discord
- 19:40and a lot of the people that I know don't have discord servers
- 19:43or discord accounts. You have to type in slash
- 19:46imagine which is something I believe you also mentioned in in
- 19:48somewhere where like there has to be easier ways than you know
- 19:52tips and tricks that you have to look online to figure out how to
- 19:55get to the image. It has to be a lot simpler than
- 19:57that. And then what do you do after
- 19:59you? Create the image.
- 20:01The problem is exactly like you mentioned.
- 20:02So many people are using these tools, but then all they do is
- 20:05they create the image and that's it.
- 20:06It stops right there. What are you going to do with
- 20:08that image? We're going to actually use it.
- 20:10And the applications is what's important.
- 20:12And clearly Adobe is in the most advantageous and privileged
- 20:17position to be using these tools because you guys have already
- 20:20every single possible, you know, application.
- 20:24You have audio, you have video, you have photo, you have 3D,
- 20:26which is the one that we're going to get into because that's
- 20:28what we're interested in. But I guess what I want to get
- 20:31back to is a similar question, but how much of the old.
- 20:35So clearly I would like to Give your opinion in in terms of
- 20:38Photoshop 1st and then outside of Photoshop since you have the
- 20:41most experience in that software.
- 20:42How much of the old Photoshop are we going to see in the
- 20:46future in the new Photoshop? Because clearly with all the
- 20:49text prompts and eventually maybe it's even voice.
- 20:51You're going to be simplifying things, right?
- 20:53You don't need to have all the depth of the tools, and maybe
- 20:56that's also the the the issue that is causing a lot of people
- 20:59not joining in Photoshop because it might be too confusing for
- 21:02them maybe. So do you think we're going to
- 21:05see a lot more simplified softwares with Adobe in the
- 21:09future? I I think so, but I think this
- 21:12is what what makes it really fun to develop these tools, you
- 21:15know? It's like, so I I had this joke.
- 21:20For years and and I, I made the same joke at Google because big
- 21:24companies are big companies are very frustrating.
- 21:26They're amazing. They they can do amazing things
- 21:28which is why, you know, we put up with them.
- 21:30It's like Churchill said about democracy.
- 21:31It's the worst system except for everything else we've tried.
- 21:34You know. So I'm like all right fine but
- 21:36you know I've got I've got this can of coke here and and I used
- 21:39to say if you asked Adobe like all of Adobe like 30,000 people
- 21:44like move this can of Coke. You'd be sitting there for six
- 21:46months and it would be like you know which VP is debating like
- 21:50should we be in the fluid transportation business and like
- 21:53what are the optics of high fructose corn syrup.
- 21:55I don't think that aligns with that brand values or or you
- 21:58could grab the coke and you just move the coke and then then it
- 22:00would be moved. And and sometimes I feel like
- 22:03that's true with tech where you know I was, I was looking at
- 22:07some some really neat stuff from another team.
- 22:09They're doing all this like semantic segmentation, you know
- 22:11and you've seen this in other tools like there's the the meta
- 22:13segment any. Anything model and a lot of this
- 22:16model. Yeah.
- 22:17Yeah, and a lot of it is in in Photoshop today in various
- 22:20forms, right? You have like select subject.
- 22:22In fact, it's funny, you can kind of like carbon date someone
- 22:26when they learned our tools by the techniques they use.
- 22:29And like somebody my age, it's like, Oh yeah, select subject.
- 22:33I forgot about that. But like you'll see somebody
- 22:35using the pen tool from like 1991, right?
- 22:38And you're like really bro. And they're like, hey man, it
- 22:40works. I know how to do it.
- 22:41Don't mess with me. I'm like, anyway.
- 22:44So all the semantic stuff is brilliant, but I think it it
- 22:46continues to coexist with the pen tool or with the lasso or
- 22:50with you know, channel based selections.
- 22:53Not because it's just because people get, you know, habituated
- 22:57and they're happy with what they had, but sometimes those tools
- 23:00are just the best way to do something.
- 23:01You know, they they're simple, they're straightforward, they do
- 23:05exactly what you said, they don't do anything smart so to
- 23:08speak and you know, selection. He's one of those things where
- 23:12it'll be like almost right. You know, it'll be like 95%
- 23:16right. But then there's like a chunk of
- 23:17like a, you know, if I was doing like the virtual background,
- 23:20like like the dresser would pop in and be stuck in out of my
- 23:23head, you know. And so with a lot of the, the
- 23:27classical tools, if you want to say, those are still valuable.
- 23:31In fact, many years ago we visited DreamWorks Animation and
- 23:35of course, you know, this was like early 2000s, but still they
- 23:37were doing like incredible production pipelines.
- 23:40And as I recall, they were using After Effects.
- 23:42And I thought, oh gosh, you know, the 3D stuff in After
- 23:45Effects, especially then, was really limited.
- 23:46And they're like, Oh yeah, that's not what we're using.
- 23:48We're just using it as basically Photoshop on wheels, you know,
- 23:51because it's sometimes the simplest and most effective
- 23:54thing. We don't want to, like, put this
- 23:56thing, like way up at the head of the stream and rerender all
- 23:58this pipeline and like, try to make it brighter.
- 24:00It's like, why not just put like a little track mat on that and
- 24:03then make it brighter? You know and it's just like
- 24:05maybe it's a quad of four points or it's a little thing and you
- 24:08just kind of shape it and then you're like putting in like a
- 24:10levels adjustment and and and sometimes even if those things
- 24:15seem low tech compared to like the the greatest new a I they're
- 24:20great. You know some of the some of the
- 24:23brilliant stuff you see today is people combining Mid Journey and
- 24:26Photoshop and you know they'll they'll use some of the new
- 24:29stuff like generative fill and they'll extend and and that's
- 24:31awesome but they'll also just use like.
- 24:34Old school selection layers, blending modes, traditional
- 24:38stuff. In fact when when Valley first
- 24:41emerged that was to me the starting gun.
- 24:43Like then OK, this has gone from like theoretical and papers and
- 24:45we were playing around with some like diffusion models.
- 24:47But if you guys go back that far again you know like 15 months it
- 24:52was stuff like you know disco diffusion and and I remember,
- 24:55you know I'll always remember it's not like watching my friend
- 24:57like we were on a a zoom call together and like taking 15
- 25:00minutes and it's like. Oh, I think I saw a face like
- 25:03in, you know, then it was right. But.
- 25:05But Dolly was where it was like, okay, boom.
- 25:07This is now practical, right? Or at least it's good enough to
- 25:10criticize in a lot of cases. And gosh, how do they get off on
- 25:14this? I guess the thing was, I wrote a
- 25:18doc called Yes And And And I wrote it.
- 25:20So that's like kind of the cardinal rule of improv comedy,
- 25:23right? It's like you, your friend.
- 25:27Walks on the stage and you have no idea what the plan is.
- 25:29And they're like, what's up, brother?
- 25:31It's great to be here on Mars and you're not like, well, we're
- 25:34not on Mars, actually, we're, you know, it's like, that just
- 25:36kills the vibe, right? So for me, the analogy is not
- 25:40that Adobe can or should try to do everything.
- 25:44I think that's an insane approach.
- 25:46What what we really want is like as long as we're making a living
- 25:50and get to what we're doing, awesome.
- 25:53I just want to provide. You know these these interfaces
- 25:56and hooks and platforms. Our friend Christian Cantrell,
- 25:59he worked at at Adobe for about 20 years.
- 26:01He left last summer and he's a screenwriter, so he's kind of
- 26:04going off and doing his own thing, which is great.
- 26:07And as he was leaving, he made a Photoshop panel which talks to
- 26:12Dolly and talks to stable diffusion and so well in advance
- 26:15of us being able to like do a deep native integration.
- 26:18It was possible for people to put those in.
- 26:20I've seen people do that with like Automatic 1111 and like tie
- 26:23into control that. And these are some incredible
- 26:25videos where people do a sketch and it like turns into a witch
- 26:28in the sky. And so wow.
- 26:30Yeah. And and I get excited because if
- 26:34we can just stay out of the way fundamentally, then I think good
- 26:37things will happen. Some of the stuff we've got to
- 26:39bring to the party for sure. And some of the doors like for
- 26:42better or worse, technically only we can open because they're
- 26:44like deeply wound into the products.
- 26:47But beyond that, I just want it where it's like, hey you, this
- 26:50is what I used to do as a designer.
- 26:51I'd string together, you know, 10 weird apps just for the
- 26:54pleasure of like A solving the problem and then, B, getting
- 26:58something nobody else could do. They're like, how did you do
- 27:00that, man? And so.
- 27:02So I think what'll happen is even as a lot of hard stuff
- 27:05becomes easy, it doesn't change who people are.
- 27:08They're still going to be ambitious, right?
- 27:09They're still going to want to set themselves apart and and
- 27:12that's where they'll just fill the time.
- 27:14Doing things they would have been completely out of the
- 27:17question. It's like, well let me see, let
- 27:18me see 10 variations of that. Anyway, I'm talking, I'm
- 27:22monologuing as I do. Let me stop cuz I'm sure you
- 27:25guys have more thoughts. No, it's very interesting, the
- 27:28points that you brought up actually.
- 27:30And it makes me think because again, Adobe has a lot of users
- 27:35from different ranges, like you have people who use their
- 27:38smartphone to take photos. And then edited in Photoshop,
- 27:42you have people who use After Effects, Premiere Pro, You have
- 27:45people who use 3D objects in Adobe, Substance Painter.
- 27:48Yeah, when you are thinking about AI, creative tools and
- 27:52when you want to design these products.
- 27:54Do you guys have a focus group or do you guys have a target?
- 27:57Design A target audience. I know to be a product.
- 28:00Manager before YouTube and before this podcasting and I
- 28:02used to be a product manager for a fintech company.
- 28:04We used to do digital onboarding and our goal was very simple.
- 28:08So we had to design products for people who visit the bank, they
- 28:11wanted loan and we know who our target audience is.
- 28:14So it was easy to, you know, design the product, design the
- 28:16features. But when I think of you guys,
- 28:18you guys have a wide range of the.
- 28:20Entire world per se. Like how?
- 28:22Do you guys manage that? Well, it it's a great question
- 28:25and especially when you work sort of at a core level like
- 28:29with the Firefly models, it ends up being this really interesting
- 28:34sort of hub and spoke exercise. And you know, part of part of
- 28:38the job of being a a product manager is largely knowing who
- 28:42to talk to or knowing who to connect with.
- 28:44It's not necessarily knowing the answer, it's like knowing how to
- 28:46get the answer. My old manager at Google said a
- 28:50PM is basically an expensive router.
- 28:53And it's like, yeah, there's some truth to that.
- 28:54It's not the whole story, but it's a lot of the story.
- 28:57And so like, as we think about, I don't know, let's pick January
- 29:02to Phil or something a lot. Like I don't.
- 29:06As as much as it this kind of kills me.
- 29:09I don't get to spend as much time directly talking to people
- 29:12as I wish. One thing we can do is like know
- 29:16the After Effects PM's, know the Adobe Express PM's know the
- 29:19folks on Photoshop and and they all.
- 29:23Really focus on their, their areas.
- 29:25I mean, even in those cases it's hard because you know, think of
- 29:27the diversity of people who use Photoshop.
- 29:29Even in something like After Effects, like Okay, that's just
- 29:31like it's it's nerds being nerds.
- 29:33It's like, but do you mean visual effects nerds?
- 29:35Do you mean special effects? Yeah.
- 29:36Like or or like motion craft people or sorry, you know,
- 29:40Anyway, so so a lot of it is kind of like building up this
- 29:44like Federated mental model in in your head.
- 29:48The other thing that I I always emphasize is.
- 29:52You can do like sell it and then build it, right?
- 29:55That's sort of like one of the the axioms of lean Startup where
- 29:59instead of creating the whole thing, come up with the concept.
- 30:02If you can't do that, don't. There's no point in that coding.
- 30:05Why would you know, like, like Amazon always says write the
- 30:08press release 1st. And I have an intern this
- 30:11summer. She's amazing.
- 30:12And this is one of the things we're working on.
- 30:14It's like hey, if if we're if we're helping people creatively
- 30:17ideate and think about like. All right.
- 30:20You know, today you type in a prompt, you hit go, you get 4
- 30:23pictures, then you, you change the settings, you hit go.
- 30:25If we're it works, right? But it only works to a degree.
- 30:29Like we said, you know, if it's majority, people get thousands
- 30:31of things. What do I do with those?
- 30:32Or, you know, how can I more efficiently visualize?
- 30:34So we have some ideas. We don't know if they're good
- 30:37ideas or not. I mean, there's ways I could
- 30:39think of them tipping over. So one thing we can do is like
- 30:43describe, describe the end state.
- 30:46You know, describe like who, who is your customer.
- 30:49By virtue of having used this, there's a great essay folks may
- 30:53have read called We Don't Sell Saddles Here by Stewart
- 30:56Butterfield, who was first, you know, created Flicker and then
- 30:59even more famously created Slack.
- 31:01And it's just it's such an incredible description of like
- 31:05what we're really trying to do. And and it a lot of it is like
- 31:08it make it. It's sort of obvious once you
- 31:10read it. But it's like we're not selling
- 31:12you a saddle. Like let's say we actually make
- 31:14saddles for horses. Like when they don't sell you
- 31:16the saddle, what we sell you is like.
- 31:18You guys on the beach, the dream, the best version of you.
- 31:22The one is like with friends looking at the ocean and like
- 31:25having a campfire. And then it's like, God, I got
- 31:28to be that guy. Like, what do I need?
- 31:30Well, you know, maybe get a horse and then can I recommend a
- 31:33good saddle. And you're like, give me the
- 31:34sail, right? And and so, so with the creative
- 31:37ideation thing, it's like, you know, who do we want folks to be
- 31:42because of what we delivered, right?
- 31:44It's like, I want you to be. More creatively satisfied.
- 31:48I want you to be more chilled out.
- 31:50I want you to have met a deadline that you you couldn't
- 31:52have hit otherwise. Maybe some.
- 31:55What are the things we have to guard against, You know is it
- 31:57like people became overwhelmed because we gave them too many
- 31:59versions too easily and and if you can kind of distill that
- 32:04then it it becomes possible to sort of write a press release.
- 32:07It becomes possible to work with design partners in playing with
- 32:12these ideas and and taking them in different different
- 32:15directions and then ultimately. If if we can do our job, you
- 32:19know, in its best form, it's like we didn't solve a problem.
- 32:22We avoided a problem, right? Because we we were able to make
- 32:26the vision so clear that we can actually test it.
- 32:29And if parts of it were bad, Oh my God, you'd so much rather
- 32:34find out up at the jump. You know it's 100 times cheaper,
- 32:37a million times cheaper than if you went and delivered this
- 32:40thing and then you got to unwind it.
- 32:42So, so to me. There is no substitute for
- 32:46interacting with people. I mean my blog is almost legally
- 32:49able to vote at this point. It's also in in blog years it's
- 32:53like an 18 year old dog. It's like you know it's it's
- 32:55moving service provider keeps falling over people like you
- 32:57know your certificate is bad. I'm like I know man I I just
- 32:59whatever it's it's an old dog like just try to be nice but but
- 33:04I do those things just as a some some way some proxy for like
- 33:09staying in the community. Twitter has been a great tool
- 33:11for all the complaining I can do about Twitter but.
- 33:15But yeah, I I think, I think for me that the enthusiasm and
- 33:18everything comes from the fact that this isn't, this isn't just
- 33:22like a side hustle. That's one of the things about
- 33:23Google. Like, it's it's an amazing place
- 33:25to go be a technologist. Right, people that, oh, I I
- 33:27learned Python And C++ and then I took an NBA class and like,
- 33:31but it'll be a PM. And then it's like what do you
- 33:33work on? Who knows?
- 33:34You just got hired at Google and like you'll work on payments and
- 33:36then you'll work on 3D and then you'll work on maps and it's
- 33:40like. Like, you're just like levelling
- 33:42up your skills and you're moving up in the career, but it's not
- 33:44really for most people anchored in a thing.
- 33:47And for me, it it's more like there's that song the Distance
- 33:51by Cake about this crazy guy who's a, you know, race car
- 33:53driver and it's like, everyone's gone, man.
- 33:56Like the race is over, but he's like, he's just still going and
- 33:59going and going because he's insane, you know?
- 34:01And and it's like, that's the that's the kind of weirdos I
- 34:05tend to find more at Adobe. Like Russell, right.
- 34:09Russell could have bounced. 30 years ago and be on an island
- 34:12and be a chill guy, but that's not what he wants.
- 34:16This stuff is too cool. How could you not be part of it
- 34:19so? Is that the reason you stayed
- 34:21for so long in Adobe and then you went back?
- 34:23Yeah, exactly. Exactly.
- 34:25That's the reason, right? Yeah.
- 34:27Well and the question you asked earlier kind of got it, part of
- 34:29it which was like, well, do we think?
- 34:32We'll just sort of do incremental extensions to
- 34:34existing tools, which I think again makes tons of sense, like
- 34:37generative fill. It's like great, I didn't want
- 34:38to relearn something. I didn't want to get into a new
- 34:40app. I just wanted the app I used to
- 34:42work better and now it does. I did this yesterday.
- 34:44There was like these donkeys running on the street in front
- 34:46of us, like it's a very interesting place.
- 34:48But there was also like this irritating car and I'm like,
- 34:51yeah, I can content or fill that, but what?
- 34:53Like I used generative fill, it was like way better.
- 34:55So. So the incremental stuff is
- 34:56cool. But the reason I came back was
- 34:58they were talking about a I first creativity.
- 35:01And what is that? I'm like, I don't know.
- 35:02They're like, well, we don't know either, but like want to
- 35:05fool around and find out and we can try to make something cool.
- 35:08And I'm like, all right. And so I think both, I think
- 35:14both will happen. You know, just this morning I I
- 35:16again, like literally sitting in the dog park reading one of
- 35:19these like daily newsletters, which are totally overwhelming.
- 35:22I I found somebody who's like doing an app for fashion
- 35:25designers and like, yeah, taking a sketch and get out all these
- 35:28like beautifully rendered. Copies and it's like it's
- 35:30probably like one guy or it's like it could be the three of
- 35:33us. I don't know if that's like a
- 35:35business or if it's really like a sustainable tool.
- 35:37But the fact that it's possible is incredible and and that to me
- 35:40is like you couldn't you couldn't just like.
- 35:45Kind of piecewise advance your your way there from existing
- 35:48tools like this is this is just a a a massively new thing and
- 35:52and that isn't even the start of like really what we'll see.
- 35:54You guys might say that something is like people are
- 35:56doing it with Gans now they're doing it with diffusion models
- 35:59as they get faster but like actually like putting down brush
- 36:02strokes and having art come out the other side and it's still
- 36:05very science fair but it feels again like the kind of thing
- 36:08that. It's just a different animal.
- 36:10It's not like, oh, it's a dog and it's like a different breed
- 36:13of dog. It's like, no, it's a bird.
- 36:15You know, like OK, cool, it's an animal.
- 36:17But it's like it's just very different in the in the
- 36:21evolutionary tree and and they're both really cool to
- 36:24experiment with. What I love about the the things
- 36:27you mentioned clearly about yourself and Russell is.
- 36:30And and I guess this goes back to the concept of competition
- 36:33because a lot of people fear competition.
- 36:35But you mentioned the moment Dolly came out.
- 36:37That's when it ignited sort of the passion you guys that Oh
- 36:40yeah, there's this new thing you can the block who can do crazy
- 36:44things. And now it gives you more
- 36:46passion to go back and and have a healthy competition in a in a
- 36:49way to who can create the best set of tools that can allow for
- 36:53more creativity amongst people. And that's I think beautiful,
- 36:56the beautiful thing that competition allows.
- 36:58Now I want to take it a little step back.
- 37:01And talk a little bit more about Adobe Firefly.
- 37:03Clearly Adobe Firefly has been one of the most revolutionary
- 37:06tools. If anybody watching or listening
- 37:08right now hasn't used it yet. It's free to test around I
- 37:11believe and and please go ahead and give it a try.
- 37:14It's mind boggling how crazy good some of the tools are.
- 37:18We were testing with some of them yesterday even and and it's
- 37:22just fascinating, right? You mentioned something about I
- 37:25believe in in the previous podcast about a wishing well,
- 37:27where right now a lot of the A i's and and they're getting
- 37:30better over time, but a lot of the A i's is something like a
- 37:32wishing well where you throw in a prompt and then you hope to
- 37:35get something good back. You never know what you're going
- 37:37to get back. How far is Adobe from a point of
- 37:41creating something that can work for creatives in in businesses?
- 37:44Because creatives in businesses, they need something that works,
- 37:46that something that's solid. And we're talking in terms of
- 37:49photo, in terms of video, in terms of audio and.
- 37:53And eventually 3D, because it's more complex to do 3D, but you
- 37:55guys are definitely doing that. You showed us something crazy.
- 37:58Can I just say what? Like the tool that you showed
- 38:00us, You can even talk about that a little bit.
- 38:04Yeah, yeah. OK.
- 38:05Gosh, there's a lot to unpack. OK.
- 38:07Yeah, yeah. Sorry that was just so much came
- 38:10out. Listen, don't apologize.
- 38:12That's all I'm doing. And like these like 9 minute
- 38:15answers if you even want to call them that, but.
- 38:19Gosh, where to start? Yeah, well let me just touch on
- 38:21the 3D thing because it also it it relates to an earlier
- 38:24question you asked about like simplification of software and
- 38:28it it's okay. I'll step back one level
- 38:31further. I have a a running discussion
- 38:34with a friend of mine. He's a a graphic design teacher
- 38:36really talented guy. He's teaching my son's design
- 38:40and and there is a lot of like sense of of kind of mourning
- 38:44actually like. It's just like sadness for
- 38:48people. And it's not just graphic arts.
- 38:49I I found somebody who's like a productivity expert and they
- 38:53were talking about, you know, large language models and
- 38:56they're like, yeah, it's great, I can do these things easily
- 38:59now. But like, I spent all this time
- 39:02refining my craft and learning how to work around the problems
- 39:05and then just goes away and it's it's a super valid, natural
- 39:11feeling. I mean, before our time, you
- 39:14know, I'm sure in the 80s when when people had spent 20 years,
- 39:18you know, cutting out letters physically from sheets, like if
- 39:21you wanted to put some type down, you had to like, get more
- 39:23sheets to come in. And I'm like, well, I can't
- 39:25really say that because I ran out of the letter C It's like
- 39:27really? That seems like an insane.
- 39:30Phenomenon. Right.
- 39:31And obviously computers solved that, but in the process, then
- 39:33people didn't have to sweat as much, you know, to do all that,
- 39:37if you can just put it in Page maker.
- 39:38And of course then people made, like, really weird kind of gross
- 39:41ransom note stuff. They're like, well, I paid for
- 39:4330 fonts. I better use all 30.
- 39:45It's like, no, no, no, no, don't do that.
- 39:48But like, that was just part of the evolution that the world had
- 39:50to go through. Like, just, is that a good idea?
- 39:52I don't know. Let's try it.
- 39:53It's like tail fins on cars. It's like, how big can we make
- 39:56them? Finally, people like, OK, that's
- 39:58enough. And it and so there there was a
- 40:02class of craft that that grew up in response to the needs and
- 40:05then went away. But people grew up new craft and
- 40:08and now we're seeing this too, where I'm I'm getting to this
- 40:12because of 3D. You know a lot of of 3D.
- 40:14It's it is. It's mostly unintentional I
- 40:18think. But there's a lot of gatekeeping
- 40:19that goes on just because it's a lot of stuff.
- 40:22I was talking to Gal Julie design.
- 40:26It's her. Handle and she's talking about
- 40:30picking up After Effects this week and I think I hope she's
- 40:33having a good time. But it's a lot.
- 40:35It is a lot because it's just it's not only all the Photoshop
- 40:38stuff but it all moves and then you have to think about like
- 40:40memory management what is a pre comp and composing those and
- 40:43it's just the grammar is more complex with with 3D.
- 40:48Well there's this this joking condition called Irish
- 40:51Alzheimer's which is where you forget everything except the
- 40:54grudges. And so for me, like, I don't
- 40:56look back like, Oh yeah, we shipped that.
- 40:57We shipped that. We can fix those bugs.
- 40:59It was great. Like what did I miss on and and
- 41:02back in in 2007 we put 3D into Photoshop and again it was we
- 41:08came in with like Blaze and Guns and it was just very exciting
- 41:12and it was like finally we're going to take this stuff, which
- 41:14is so hard and we're going to like regular graphic designers
- 41:17use it. I think we were just like
- 41:18really. Really naive like it was it was
- 41:21easy to maybe to go to the 1st 50% of these and the 1st 80% but
- 41:25it's the other percent that killed it.
- 41:27And it was like for most people they couldn't they just couldn't
- 41:30get around the corner to like make good looking 3D type or or
- 41:33just like put a texture on something and have it because
- 41:35it's like you have to understand materials and lighting and and
- 41:39normals and like what the hell does he mean?
- 41:41No, no, not that kind of normal, a different normal.
- 41:43And you know, it's not to say that.
- 41:47The fusion models will solve everything.
- 41:49But what I think is really interesting is, is how they're
- 41:54starting to solve a lot of the problems.
- 41:56You know, we we can take things. And on my Twitter feed I've,
- 41:59I've put up some of these very short demos where it's like Gray
- 42:01box art, right? Like I think it was like a Lego
- 42:03guy with no paint and and just by like putting in these
- 42:06primitives into space, you can then give the the system some
- 42:11guidance, right. And like we were talking about.
- 42:13You know, do we think traditional Uis will go away,
- 42:16like the lasso tool or the move tool?
- 42:17Well, no, because sometimes it's nice to just move the thing
- 42:19right. You don't want to be like a
- 42:20little to the left, a little to no, a little and actually a
- 42:23little bit more to the right, a little more to the right.
- 42:25Or you change one thing and three other things change.
- 42:27Sometimes you just want to, like, move the coke, change the
- 42:29deal, whatever. And with 3D, I think for a lot
- 42:33of the tasks, like just maybe making a background or an
- 42:36environment, using some simple primitives and describing what
- 42:39you want could be perfectly sufficient.
- 42:42You know, for the more like super, super demanding stuff,
- 42:45let's say you're Nike and you have like this, you know, CAD
- 42:47render of a sneaker. It can't just be like kind of a
- 42:51sneaker. It has to exactly be that
- 42:52sneaker, full stop. And so I'm I'm imagining we'll
- 42:56have a world of kind of hybridization where, you know,
- 43:00if you just want to make something like a cool scene or
- 43:04cool element in the scene, you know, for a guy like me, maybe
- 43:07using a simple 3D primitive plus some diffusion.
- 43:11Could be great. You know maybe if if that's like
- 43:12in Photoshop or it's in stager, that could be sufficient for
- 43:15other tasks maybe I still want to use you know traditional
- 43:18geometry materials renderer. I bet there'll be a lot of
- 43:21really cool ways to put those together And so I don't that's
- 43:24again it's it's the yes and idea it it's not to say, well God,
- 43:29now the world's even worse because you had all the
- 43:30complexity before plus the new complexity.
- 43:32Hopefully that won't be true but but hopefully what'll happen is
- 43:36for a lot of the things that used to take a lot of this sort
- 43:39of like. Climbing of the stack.
- 43:40We used to say you must be at least 12 pixels high to ride
- 43:43Photoshop. You know you had to just learn
- 43:45all of these basic things you know some of that some of that
- 43:50either can go away or it could be deferred right.
- 43:53It can it can come in only when you need and want that level of
- 43:56precision and and for the rest of it if you just want to do
- 43:59something simple quickly it's kind of like photo software
- 44:01where a lot of the edits that used to be really demanding in
- 44:04Photoshop you can it can actually be smart and be like oh
- 44:07I I know that person. You know what what a normal skin
- 44:09tone. One is and therefore, oh,
- 44:11they're in a shadow. I can fix that for you, right?
- 44:13You didn't have to learn a lasso tool and you know, all all that
- 44:16stuff. But again, it also didn't mean
- 44:18you shouldn't or that those things wouldn't help you.
- 44:20It's just for the first N percent where you you that would
- 44:23have been a blocker that just would have cut the total number
- 44:26of people. Now that doesn't have to be
- 44:28true. And and I really, you know, I I
- 44:30joke with the 3D team and it's like, I don't know if it's a
- 44:33good or bad motivation, but it's like I I have my white whale.
- 44:37Like I want to. I want to get back and I want to
- 44:40fix all the stuff that, like, we started, We built a bridge like
- 44:43halfway over this ravine and it's been sitting there looking
- 44:45ridiculous, you know, since like the Bush administration.
- 44:49And it's like, well, can we just finish the damn job now?
- 44:51Like, maybe the tools will open those doors that that we can
- 44:54make it easy in a way that we just couldn't.
- 44:59It's definitely fascinating. I want to just.
- 45:03Consider the fact that what you guys have as a tool set for 3D
- 45:06artists, specifically Substance painters, the the software that
- 45:10we use which has allowed us to create such amazing textures for
- 45:14for all the 3D work that we do. And we know so many 3D artists
- 45:17who their go to software is Substance Painter.
- 45:20And we were just thinking how great would it be Because
- 45:23there's already a stupendous amount of textures that you can
- 45:28look for in the Adobe website that you can just download once
- 45:31you're subscribed which is already.
- 45:33More than enough, but what if you could type in a prom and
- 45:36create the texture that you exactly needed like?
- 45:40All these questions come to mind and we are just so excited for
- 45:44the. Future I saw fathers crying the
- 45:45other night talking about this while he was texturing the
- 45:48character, and he was like the same text prompt that I would
- 45:52use in the generative field if I could use it here.
- 45:55You could go to sleep earlier tonight, not not stay away until
- 45:58three I. Just need the score right?
- 46:00I was doing. I was doing up a meta human for
- 46:02for. On real Engine project and I was
- 46:04like, what if I could just add a scar on this face?
- 46:06I don't want to go and Google the image or even, you know,
- 46:09generate it and then come up, please.
- 46:10I just want to generate that texture just imagining all of
- 46:13these things being added to every piece of Adobe.
- 46:16It's gonna it's a bright future. It's a it's a future.
- 46:20I'm excited. About good, good.
- 46:21Me too. I think so.
- 46:23So a couple of thoughts. This will be digression, but
- 46:26hopefully entertaining. So, so way back in the day when
- 46:29Adobe bought Macromedia. You know, so, so like Adobe you
- 46:33know, was like these 40 something, you know Xerox
- 46:37scientists you know that's like its start, you know, and
- 46:39Macromedia was much more like kind of gunslinger.
- 46:42They were in San Francisco and it was like you know they made
- 46:44Flash and it was much more like Woohoo and and Adobe, you know
- 46:49like we we bought them and like Macrobeta people were like they
- 46:51visited the Adobe. San Jose office And they're
- 46:55like, I don't know. It looks like a bank.
- 46:56Like these guys are like no fun. But Adobe of course, like not so
- 47:00secret. We were like, Oh yeah, please
- 47:01make us fun. Like please make us cool.
- 47:03And we always do this by the way.
- 47:05And then some music researchers asked people like OK, if if if
- 47:10various apps were people at a party, who would they be?
- 47:12Like can you describe that person?
- 47:14And Photoshop was like this. Gray, bearded professor, right.
- 47:19Like, you could tell this guy's done a ton.
- 47:21It's very impressive and it seems friendly.
- 47:23You go talk to him, but, like, you're a little bit afraid.
- 47:26Illustrator was like this beautiful woman over on the
- 47:28corner and it was just like, OK, I got probably no shot here, but
- 47:31like, wow, I get it. That's, that's great.
- 47:34And Flash was like the young dude who was wearing, like
- 47:38those, you know, sunglasses like the louvers on him.
- 47:40And he's like. Women are like throwing drinks
- 47:43in his face but he's also going home with a bunch of numbers.
- 47:45Right. So, so and and so I'm trying to
- 47:49think like where would I put substance in there?
- 47:51And it's somewhere in in the realm of I I think like
- 47:57professor slash beautiful woman. Like it's like, Oh yeah, that's,
- 48:01that's, that's, that's cool. I I definitely should make time.
- 48:04Personally, I haven't. But I'm like, Oh yeah, like the
- 48:06stuff we we were starting to do in Photoshop with like, you
- 48:10know, projection painting and things.
- 48:13Like they actually really leaned into it and they really got it
- 48:15right. That's so cool with with the the
- 48:19question you're asking about generative texturing.
- 48:22So first of all I I have to just put a lot of caveats outside
- 48:24because I it's not a world I know very well.
- 48:26I I'm I'm intrigued by it. I'd love to to learn more.
- 48:30I mean I always kind of wonder like you say there are so many.
- 48:36Generated materials today. And the same goes for for
- 48:39shapes. You know, every so often, even
- 48:41this morning I found a couple research papers where people are
- 48:45using 3D to to to shape, or they'll take a 2D shape and
- 48:48they'll make a 3D shape and it is incredibly, incredibly cool.
- 48:53I my sort of devil's advocate wonders, I'm like, well, why
- 48:57wouldn't you just use one of the shapes that exist right?
- 49:00Like if you went on Sketch Fab or Turbo squid or you know,
- 49:03something like. Do we need a new, like, do I
- 49:05need to be able to, like make an attack helicopter just by
- 49:08describing an attack helicopter or would I be better off if it's
- 49:10like, Oh yeah, somebody went and they spent three months of their
- 49:12life making this thing and I can buy it for, you know, 20 bucks
- 49:16or 100 bucks or whatever. And.
- 49:19And so again, this is a naive question.
- 49:21I don't think there's a right answer, at least in my mind.
- 49:22But like with the textures, kind of what is the advantage of
- 49:26making something yourself versus using something that exists?
- 49:31You know, we talk about like the long tail and.
- 49:33And I think that's where. With like the the shape
- 49:36generation, it's like, okay, What are the things that I that
- 49:39don't, that don't already exist in the corpus of things people
- 49:42have made? Could I make those?
- 49:45And then would I be better served by describing it and
- 49:49generating geometry? Or now there's all these like
- 49:51cool photogrammetry techniques, you know, I could like walk
- 49:54around it, I could make a nerve, I could make, you know, some
- 49:56volumetric representation. I don't know I I again I think
- 50:01it's it's the not knowing that makes a lot of this fun.
- 50:03You know it's sort of like watching athletes like will that
- 50:05guy make it. I don't know I wouldn't try that
- 50:07but like let's watch and find out and so so it is kind of back
- 50:12to that high wire act of yeah can can we find a way where
- 50:17where these these technologies can can be useful.
- 50:20By the way I'll just put my cards on the table with with
- 50:22textures I mean one of the things that that people tell us
- 50:25they want with Illustrator which is the world I I know somewhat
- 50:27better. Is yeah, of course there's
- 50:30textures, but a lot of it is, it's, it's kind of, it's not as
- 50:33vibrant in ecosystem, right. I mean people did like, oh like
- 50:36here's some some postscript thing from 1992 like and it's
- 50:39still sitting around in the in the sample files.
- 50:42If you could do things more descriptively and more
- 50:45expressively but do it in context, that would be a huge
- 50:49boost. And again, I mean people are
- 50:50excited if we make a whole scene right, and do it in vector, do
- 50:53it in raster. But it's like, yeah, day-to-day,
- 50:55man. What I actually need is just
- 50:56like, fill a hole. So great gender to fill.
- 50:58Awesome. And it's like, yeah, then I'm
- 51:00also using some texture. So, like, help me make a
- 51:02texture, right. I don't need you to make the
- 51:03whole world for me. That's my job.
- 51:05Like to your point you was like, oh, I could have gone to bed
- 51:07earlier. You're not going to go to bed
- 51:08earlier if you're like me, you know, come on, you're going to
- 51:10stay up later. But you'll do something more
- 51:12interesting in the time you stayed out, you know.
- 51:14So at least you didn't you didn't waste your time And and
- 51:17not to be indelicate, but there's a great essay called
- 51:19shit work and it's basically says like, don't don't give you
- 51:22a user shit work. Right.
- 51:23It's like this was written in the time of like Google Plus
- 51:25making everybody like add friends to circles and you had
- 51:28to sit there and like, all right, fire out.
- 51:29Well, we we met and OK, now are we friends or are we coworkers
- 51:33And we'll kind of both, you know what I mean?
- 51:34And then and people like ask for this, I'm not going to do it.
- 51:37And it's like thankfully, you know, algorithms got good enough
- 51:41that that more or less like it could.
- 51:44It could just sort that out for you.
- 51:45And similarly if we can just kind of get that kind of dumb
- 51:50stuff out of your way. Again, like, yeah, if you can go
- 51:54to bed earlier, great. Please get sleep.
- 51:56That matters. But like, I also just know the
- 51:58nature of ambitious, creative people.
- 52:00They're just going to be like, OK, what the?
- 52:01With that out of the way, Oh my God.
- 52:03Now I could learn After Effects. I could try this other bananas
- 52:07thing that, like, was just been would have been prohibitive
- 52:09otherwise. That's exactly like the message
- 52:12that we are trying to send with AI is that it speed up your
- 52:15process. You can focus on the things that
- 52:17you love rather than doing things that are repetitive or
- 52:20taking a lot of your time, right?
- 52:22But I want to ask a question that you started the
- 52:24conversation with that you talked to a lot of your friends.
- 52:27And they are worried, rightfully so.
- 52:30But what I saw the other day, there was a live stream from
- 52:33Adobe actually and the there was Lisa Carney on it talking about
- 52:38how she designed movie poster using Adobe product suit and it
- 52:43was so interesting for me. She was an advocate for a I she
- 52:47was super excited to use it and the way she used it.
- 52:50It was crazy to me if you if I didn't watch that live stream, I
- 52:53would put her in the category of people who should be worried and
- 52:56who are worried about AI. But.
- 52:58Look at because she's working on a lot of the posters for like
- 53:01the films and stuff that we've seen exactly.
- 53:03And. How can someone for the people
- 53:05who are watching now, how can someone transition themselves
- 53:09from being a person who is worried?
- 53:11And now who's the person who's using AI to their advantage to
- 53:14save time, to create better products, to improve the
- 53:17quality? That.
- 53:20Yeah. I don't know how big a number
- 53:21you want to put on it. That's the $64,000 question, $64
- 53:25billion question. I don't know.
- 53:27Yeah. I I don't presume to have total
- 53:30answers. I I think when I see somebody
- 53:32like Lisa though I I've only recently met but she's been you
- 53:37know she's been doing this for 30 years I think you know and
- 53:40and it's like super, super inspiring much like Russell just
- 53:45to see somebody who's like innately curious.
- 53:48Right. Like I think that's that's one
- 53:50of the the interesting things. You guys mentioned Burnout
- 53:52earlier and I will tell you like as excited as I am now, I kind
- 53:56of, I kind of swing higher and lower, right.
- 53:58I look at a lot of, a lot of my peers, a lot of people who like.
- 54:02Way exceeded me organizationally.
- 54:04Like they could make a movie about me called the 47 year old.
- 54:07I see you know individual contributor and I have a new
- 54:10manager and she and I should have a talk and be like what do
- 54:13I actually want to do? I'm like I just kind of want to
- 54:15do this honestly. Like I just want to do this
- 54:17better, like help me be better at this, but like I don't need
- 54:20to be a director or VP or whatever, right?
- 54:22I I just, I want to do this. The thing I observed though is
- 54:25like what what tends to work in a career.
- 54:29Is a lot of well the Germans have this term sit flesh which
- 54:33it sounds like sit flesh because it kind of is a cognate and it's
- 54:37basically just means patience. Like it just means the power of
- 54:39like an iron ass, right. Like the ability like you guys
- 54:43know like sometimes a lot of the work it isn't fancy but you just
- 54:46got to pound it out right and you just got to grind it.
- 54:48I mean when I got out of college I didn't have a design degree.
- 54:50I somehow like BS my way into an agency and like I was willing to
- 54:54work for like no money basically.
- 54:57But they're like, yeah, we did like 200 animated GIFs made
- 55:00today because like we, we don't have a fancy thing.
- 55:02We just made a make. So fine.
- 55:03So anyway, the sense flesh thing, a lot of a lot of career
- 55:07advancement seems to come from just like be be reliable, be
- 55:11predictable, kind of like keep going and man, if I had more of
- 55:14that, I'd be better off my my wife's a program manager and
- 55:17she's so good, like. I I'd probably be, you know
- 55:20outside living in the van as opposed to just driving our van
- 55:24you know because but like she she's like oh did we actually
- 55:26had to pay the title. I'm like Oh yeah, right, cool.
- 55:30But but the thing with me is it this is ultimately back to like
- 55:33Lisa and Russell and curiosity and passion.
- 55:36It's like those are also rare currencies I find and I I think
- 55:43a lot of the art of of me and my own career is just trying to.
- 55:47To find like product market fit, right.
- 55:49It's like I'm not going to be that most reliable guy like in
- 55:53terms of just like did you do the like just you know like
- 55:58again I wish I was better. I'm not saying I won't try to be
- 56:00but like you can't fake curiosity and and when I see
- 56:04somebody who's been in the game for 30 years right obviously
- 56:07they've seen a lot of stuff come and go.
- 56:09I mean they probably weathered the storm the first time you
- 56:12know going from. God knows what it was, if if
- 56:14anything, you know, it's like maybe it was like rub down
- 56:17letters or quantile paint box, you know?
- 56:19And it's like, hey, guess what, I'm sorry to inform you your
- 56:21skills are now the value's going to 0 because we're just not
- 56:25going to do that anymore and it's like, shit that sucks, OK,
- 56:29have a drink, you know, wipe it off and like what are we going
- 56:33to do now? I mean, I can't.
- 56:36The part is I don't want to be too, too glib because it's like
- 56:39if if. If if somebody you know using
- 56:44Adobe Express or PowerPoint or Canva or whatever tool can make
- 56:48something, it's not going to be as good as what they would have
- 56:50gotten and they know it right? But they also can do it in 4
- 56:54minutes by typing some text prompts and like they got
- 56:57something approximating a solution.
- 57:00It's like kids cheating on homework.
- 57:01It's not as good as my essay, but it's like it's good now.
- 57:04And guess what I'm on. You know when I forgot to do the
- 57:05thing and it's this or fail, right?
- 57:07I'm going to do it. I I don't.
- 57:10I don't know that there's a world where.
- 57:12The exact same number of what we currently consider creative jobs
- 57:16stays the same. I hope what happens is even that
- 57:21they're they're not the same jobs and people don't pay for
- 57:23the same things. There is now such a a
- 57:26possibility of content that there's there's enough seats at
- 57:30the table. I started reading an essay this
- 57:32morning. I wish I'd I'd finished it, but
- 57:34it was already super interesting.
- 57:35Skype Benedict Evans. He was writing about automation
- 57:39and just like looking at the last, like 200 years of like,
- 57:42you know, the Industrial Revolution and and steam, and
- 57:44they're these interesting things which are paradoxes, right?
- 57:48You would think that like, oh, well, oh shoot, we're going to
- 57:51run out of coal. And people like, well, don't
- 57:52worry, we'll make steam engines more efficient and therefore
- 57:54we'll have enough coal. And they're like, well, actually
- 57:56that's. Not going to be what happens
- 57:58when you make them more efficient, use more coal,
- 58:01because now more things became possible, so the demand actually
- 58:05in aggregate goes up, even though the individual ones get
- 58:08more efficient. And he was also talking this
- 58:11notion of a lump of Labor, which is sort of a classic idea, which
- 58:14is like, well, there's there's this amount of Labor and there's
- 58:16this amount of stuff to do. And well, we made the stuff to
- 58:19do more efficient. So now you have this leftover
- 58:20labor and those people are screwed.
- 58:22And it's like, sometimes that happens, but.
- 58:26There's been this like literally now hundreds of years of humans
- 58:29reinventing themselves because like I say, you're probably not
- 58:31going to go to bed earlier, you're just going to hustle and
- 58:33do something better because that's your chance to set
- 58:36yourself apart. And I think again, if we can
- 58:39kind of get the crap out of people's way and that that will
- 58:44let people who are are curious and flexible, you know, keep
- 58:47keep being so. But man, I I'm honest, like I
- 58:50like. I say before Dali even I was
- 58:52like. Maybe we've chewed this gum long
- 58:55enough. Maybe there's just no flavor
- 58:56here. And and and I yeah, I thought
- 58:58first creativity was going to be a thing and I just couldn't make
- 59:01it happen, which I couldn't. I'd been at Adobe a year by that
- 59:04time and I'm like, I came back to shoot my shot and crap, I
- 59:07didn't get it done at Google either because nobody cared
- 59:09right now. All they were going to do was
- 59:11crop their photo, if we were lucky, you know?
- 59:13But then this stuff hit and they were like, ah, I'm glad the Sitz
- 59:16flesh kicked in. And I was willing and able to
- 59:19sit here long enough to till this next, you know, sort of log
- 59:22came along for me to jump onto. No, that's a beautiful way to
- 59:27put it. I think it was a Lex treatment
- 59:29podcast of you're watching very recently similar, similar sort
- 59:32of conversations where there's not a set number of Labor and
- 59:36work to do. The demand that there's always
- 59:39new demand that is created when you sort of intelligence or new
- 59:43efficiencies pop up and this time it's even greater than
- 59:46before. So we might see.
- 59:48A shift and it might seem negative for a lot of people
- 59:51now, but I'm sure eventually like you mentioned, there's
- 59:54going to be newer demands for newer things and we're going to
- 59:57be seeing a lot more of that and and you guys are clearly trying
- 1:00:00to highlight that by by creating the tools that.
- 1:00:04Allows people to put the crap aside and focus on the things
- 1:00:07that actually matter and focus on the things that they can
- 1:00:09improve on. Speaking of focusing on things
- 1:00:12that they can improve on. We touch a little bit on the 3D
- 1:00:14side of things and and and for for photo generative feel.
- 1:00:18What about video and and audio just want to talk on that do do
- 1:00:21you have any updates or any news that you can share with us about
- 1:00:24what's happening because we did see the Adobe Firefly for video
- 1:00:27and that was. Just incredible is, yeah.
- 1:00:31Do you know when that's coming out?
- 1:00:32Is it available? We just want to have some
- 1:00:34information on that, if that's possible.
- 1:00:36Yeah, yeah, really good questions.
- 1:00:39So what what you're referring to in case folks haven't seen it,
- 1:00:43it was like probably a 60 or 92nd little kind of vision and
- 1:00:47and teaser we put out at NAB and I I can find that I know what's
- 1:00:50in my my my Twitter feed somewhere.
- 1:00:53It's just it was kind of this rapid fire it was like.
- 1:00:56You know, how could we change storyboarding, How could we
- 1:00:58change, you know, B roll generation, how could we, you
- 1:01:03know, do track mats better and it's just kind of go go go if
- 1:01:06you want packet. Of course every one of those is
- 1:01:07like totally different technology and they're all like
- 1:01:09totally different stages. So I think a bunch of them, you
- 1:01:13know can come around relatively soon.
- 1:01:17I don't want to you know speak out of turn for for those teams.
- 1:01:20I will say I talk a lot to them. You know, the thing that drew me
- 1:01:24to Adobe 23 years ago was After Effects, basically.
- 1:01:27And you know, I was a Flash animator and I was like,
- 1:01:30thankfully again, I hope. I don't think you guys go far
- 1:01:32enough back to remember the pain, but it was like somebody
- 1:01:35literally took like a notes app for like an Apple Newton.
- 1:01:40In like 1994 and and like through this bizarre
- 1:01:43evolutionary process, you know, it became like, well, it's like
- 1:01:46a little bit of animation. It became this and it became
- 1:01:48like a motion graphics tool. But it was a nightmare.
- 1:01:50Absolutely. Look, I loved what I could do
- 1:01:51and I hated how it made me do it.
- 1:01:53And so Adobe was like, hey, what if we took like the After
- 1:01:55Effects approach brought that to the web And then we made and
- 1:01:59this was the part where where where the bridge was out and we
- 1:02:02didn't know it. But like, what if we did Web
- 1:02:04standards? We'll do SVG, We'll do CSS.
- 1:02:07We'll do JavaScript. And it was the right answer.
- 1:02:09It just was 10 years too soon. And so it's like, oh crap.
- 1:02:12But actually we can't do that. The only practical okay.
- 1:02:14So we're going to do, we're going to do flash.
- 1:02:16And then it was like, well, we can't win on somebody else's
- 1:02:20field, right. But anyway, so, so I love that
- 1:02:23world. And so I talked to those folks a
- 1:02:24lot and yeah, I think they're thinking about kind of the range
- 1:02:28of possibilities. I mean in in the limit that
- 1:02:30there is of course, like type words generate.
- 1:02:34You could generate a script, You could generate visuals, You
- 1:02:36could make it temporarily coherent.
- 1:02:37You know, there's folks like Runway who are doing awesome
- 1:02:39work now, you know, and and it seems like kind of, you know,
- 1:02:42month by month, maybe week by week, they're like, oh, we could
- 1:02:44do 2 seconds, it could do 3 seconds, How about 7 seconds?
- 1:02:47And it seems trivial, but it's like, yeah, but if it's
- 1:02:49compounding in that way, it gets somewhere pretty interesting
- 1:02:52pretty quick. I mean video and the nature of
- 1:02:56the beast is a lot around. I mean the compute, right.
- 1:03:01As as you well know with 3D and with video, I never had anybody
- 1:03:04by the way who worked in those tools tell me on a Photoshop
- 1:03:08customer visit like, and I wish Photoshop was easier.
- 1:03:11They're like that's never the issue, right.
- 1:03:12They would they would be like, dude, how come your 32 bit story
- 1:03:15sucks so much, you know and you don't know if you this targa
- 1:03:18file and you had the alpha channel in the wrong place and
- 1:03:20I'm like whoa, whoa, OK, OK, but like the nerdiest stuff.
- 1:03:22But it was never like this is complicated so so people get
- 1:03:26that it there's a lot of compute.
- 1:03:27There's a lot of mental horsepower.
- 1:03:30So I think it may take some time for for things like real, you
- 1:03:33know, generation of video to to be like production quality.
- 1:03:37I think there's tons of like surgical applications that could
- 1:03:39be interesting. You know, even there's a sort of
- 1:03:42interesting back and forth happening between my, my current
- 1:03:47Adobe teammates and my former Google teammates.
- 1:03:49So last August, I believe some of the Google folks came out
- 1:03:54with Style Booth. Sorry, Dream Booth.
- 1:03:58We'll get there. The names get except the dream
- 1:04:00booth, right. So you guys have seen this where
- 1:04:01it's basically custom model training and you can take like,
- 1:04:04hey, take 5 selfies or five pictures to your dog and we'll
- 1:04:06make a concept and then we'll be able to do things with that and
- 1:04:10then. A couple of months ago my Adobe
- 1:04:13teammates came out with this thing Instant Booth, and they're
- 1:04:15like, hey remember Style Booth? What if that was like 100 times
- 1:04:17faster? Like that sounds good because
- 1:04:20then instead of like I upload this stuff and I wait 15
- 1:04:22minutes, like maybe I could just do it in context.
- 1:04:24And then you know like 2-3 weeks ago the Google folks are like,
- 1:04:28hey, we got this thing called style drop.
- 1:04:31Like, hey, what if the instead of requiring 5 images, you had
- 1:04:35one image and and then you could just upload that.
- 1:04:37So now being again a crazy person, I'm going around.
- 1:04:41This Old Town we're in and finding like antique
- 1:04:43illustrations of it. Take a picture of that, take a
- 1:04:45picture of that, take a picture of that, like and I'm like DM
- 1:04:47ING these guys. Like hey, when can we have a
- 1:04:49build where I can like upload this and see if it's really
- 1:04:51true. Does it really work that well?
- 1:04:52Because that would just be mind blowing, right?
- 1:04:56Anyway, so I think there will be these these more focused
- 1:05:00applications, like the reason I'm going off about the the
- 1:05:03dream booth stuff is, you know if you want to tell a narrative
- 1:05:06with characters, right, you want to do storyboarding.
- 1:05:10One of the things we get asked a lot about is well, this is cool
- 1:05:14like I can say you know, a young woman in her 30s with blonde
- 1:05:17hair etcetera, etcetera. But then like every time you
- 1:05:20generate, you get a different young woman and blonde in her
- 1:05:2230s with with blonde hair. You you need to get like Sarah
- 1:05:24or you need to get Dave or whoever and putting more of like
- 1:05:30the the. The custom training stuff at
- 1:05:33people's disposal could could be like a very tractable way of
- 1:05:36saying, hey, it's not like you're not going to storyboard,
- 1:05:38you have to rethink everything you did.
- 1:05:40You can just do it better, you can do it faster, you can like
- 1:05:42convey your ideas with more precision, you know, as I
- 1:05:45mentioned our friend Christian, and he hasn't fully declocked on
- 1:05:48like what his startup at the concept dot art is.
- 1:05:50His is his domain what he's going to do, but given his.
- 1:05:57His rich history with with Adobe, he actually then went to
- 1:06:00stability for a brief period before it being like, no, I
- 1:06:02really do want to do a startup and like knowing what it takes
- 1:06:05to sell a script in Hollywood, I can, I can kind of speculate
- 1:06:10that you know, he'll be working on tools along those lines.
- 1:06:13You could also imagine, you know, very natural touch points
- 1:06:16with Adobe apps where you know, helping you kind of construct a
- 1:06:20script or a story and then turn that into.
- 1:06:24You know what are the ingredients you need?
- 1:06:25Do you need like a premier project?
- 1:06:27Do you need an After Effects mock up?
- 1:06:28Do you want to generate textures?
- 1:06:30How about logos? Like how how far into the
- 1:06:33process Do we want to get a running start based on the seeds
- 1:06:36of the ideas and the characters that you had.
- 1:06:38And so again lots, probably more open questions than there are
- 1:06:41answers there. But I can imagine rather than
- 1:06:44getting like a hey today we we, you know, we flipped the two
- 1:06:47switches and we turned the two keys and now there's a Firefly
- 1:06:49for video. What you'll end up getting is
- 1:06:52like. This sort of progressive roll
- 1:06:54out of like focused applications that will ultimately complement
- 1:06:58each other and it'll feel like forever.
- 1:06:59It'll, you know, let's say it's two years or something and then
- 1:07:02we'll look back and be like, well that was quick, but like
- 1:07:04while you're doing it like Oh my God, like I I can't believe we
- 1:07:07haven't done that yet. It's like people are going to
- 1:07:08die. It's like they're not going to
- 1:07:09die, but but good for you for for wanting to move quick.
- 1:07:13One, one thing you mentioned and I agree with you, I think it's a
- 1:07:16surgical applications of what Adobe Firefly for video will do
- 1:07:20that will inherently benefit a lot of people.
- 1:07:22I'm more so excited about those, the smaller tools that will come
- 1:07:25out rather than oh generate this video for me because more often
- 1:07:28than not we have the videos ready, we have the audio, we
- 1:07:31just want to make no small changes.
- 1:07:33Like for example for generative fill for photo, but imagining
- 1:07:36stuff like that for audio and like small sound effects and
- 1:07:38stuff like that. And exactly I'm sure a lot of
- 1:07:40people will be interested in in in that suit.
- 1:07:44Well, but just before, like they're going, I heard a siren
- 1:07:47in the background, right? And that's the kind of thing
- 1:07:50where it's like. Sometimes you maybe want to add
- 1:07:52a siren. Sometimes you just want to
- 1:07:53remove it, you know. And there have been cool tools
- 1:07:56like an Adobe Audition, like basically a healing brush.
- 1:07:58And you know, if you know what to look for, it's like the
- 1:08:00watching The Matrix. Like blonde, brunette, redhead.
- 1:08:02They're like, oh, there's the siren.
- 1:08:03And I can like draw over the siren and get rid of it.
- 1:08:05But but just kind of like you're saying these these very surgical
- 1:08:08applications. I think in that that concept
- 1:08:10video, they're talking about B roll or they're talking about
- 1:08:13sound effects and they're like, yeah, let me just maybe I don't
- 1:08:17want to search a stock library of audio and like, try to get.
- 1:08:20This one versus that one, I'm like, oh, I just want this kind
- 1:08:23of sound like and just make that on the fly.
- 1:08:26And then at least I know, you know, people aren't going to be
- 1:08:28like, hey man, I wait a minute, I know you bought that stock
- 1:08:30asset because I've seen that before.
- 1:08:32You know, I used that three years ago in a project.
- 1:08:33It's like, no, you couldn't have, it did not exist before,
- 1:08:37right. But but it is like you say, a
- 1:08:39very kind of pointed directed application as opposed to like
- 1:08:43trying to boil the whole ocean in one go.
- 1:08:47And I think that's exactly what you mentioned.
- 1:08:49That's exactly what people will be using the most.
- 1:08:52And one reason I believe Adobe's doing a great job at the moment
- 1:08:56is because Adobe's focused on creating all these smaller
- 1:09:00applications that you guys are clearly testing out.
- 1:09:03And that's why you're allowing people to test them out for free
- 1:09:05and just play around with them and understand how they work and
- 1:09:07you guys get to get more feedback from people. 1.
- 1:09:11One interesting point here, and I think we discussed this with
- 1:09:14Farred, is you have kids and clearly you are in love with
- 1:09:17them because most we we did stalk you a little bit before
- 1:09:20the podcast. I hope you're okay with that,
- 1:09:22but clearly. You are not that difficult, you.
- 1:09:25Are over too. Like.
- 1:09:27It's you have millions of videos and photos of your kids and and
- 1:09:34clearly you enjoy spending time with him.
- 1:09:35Clearly you won the best for them.
- 1:09:37Yeah, you you being in the field of design and in in in the field
- 1:09:41of creativity and you know, manning the helm in this field.
- 1:09:44What is your advice to them and potentially other, you know,
- 1:09:48kids who could be watching this and you know they're eventually
- 1:09:51going to get a career. They're looking to get into the
- 1:09:53creative field. What is your advice to them?
- 1:09:56What should they work on? What should they learn in order
- 1:09:58to be able to navigate and have success?
- 1:10:02Good, Good question. SO.
- 1:10:04So one son, the older one, he's 15, he's interested in fashion
- 1:10:10design. You know, in a month from now it
- 1:10:12could be something different. That's part of the fun of being
- 1:10:14that age, right? You get to keep discovering
- 1:10:16things, getting inspired. But that's really interesting to
- 1:10:19him now. And so he's and he's also
- 1:10:23interested in music. And so he's focusing on just
- 1:10:25some like basic stuff like learning the piano and like.
- 1:10:29You know, playing scales and it's like nobody loves like
- 1:10:32playing scales but like you you again, you have to be 12 pixels
- 1:10:35high to write Photoshop. Like learn some basics, you
- 1:10:37know, with with the fashion stuff, you know we just try to
- 1:10:42be supportive and like man, there's so many ways we could go
- 1:10:44like I don't want to force you to like, I don't know grind on
- 1:10:48things. I mean sometimes that has value
- 1:10:50but like if you're not interested in typography or
- 1:10:53sculpture, whatever, like there's there's plenty like come
- 1:10:55in here we found a great lady who.
- 1:10:58Who runs a just like an open studio and she's teaching him
- 1:11:01like some you know textile stuff and and just like basic sewing
- 1:11:06and and he's it's funny his younger brother is much more
- 1:11:10theatrical you'll see him all over the social stuff whereas
- 1:11:13the the older one is it's more like those fish that live in a
- 1:11:15cave you know and it's like if the sun hits them they like
- 1:11:18catch on fire so they're like if he's it's almost like Sasquatch
- 1:11:22like if there's very occasionally like you'll find a
- 1:11:24picture of him but but. But he prefers to work like
- 1:11:28behind the scenes in the theater department and just like learn,
- 1:11:31like how do you, how does all the stuff work in the costuming?
- 1:11:33You know the younger one is much happier being on stage and and
- 1:11:36like being being a very bombastic with the younger one.
- 1:11:42He and I did a photo workshop out in Ely, NV in in February
- 1:11:47and it was it was amazing. It's like.
- 1:11:50It's like, I'm, I'm pretty cheap by nature.
- 1:11:52Like I don't want to spend money, but I'm like, man, they
- 1:11:54they always say, invest in experiences and not in things,
- 1:11:56like have stories to tell instead of stuff to show.
- 1:11:59And it's a little cliche, but it's pretty true.
- 1:12:02And So what was amazing is, you know, part of it is like, OK,
- 1:12:07you don't have to try very hard. Like you're going to be next to
- 1:12:09a steam engine at dawn in the snow.
- 1:12:12Like, unless you have the lens cap on, you're probably going to
- 1:12:14get a pretty cool photo because it's a cool thing, right?
- 1:12:16But. But part of the, the part that
- 1:12:19was neat is, you know, we're in this machine shed.
- 1:12:21He's already interested because he loves the subject which is
- 1:12:24trains. And we're sitting there with
- 1:12:26these like accomplished pro photographers and they're like,
- 1:12:28well, let me show you like let's how to let's do a portrait
- 1:12:32photography on one of the engineers who's like you know,
- 1:12:35he's wearing the clothes, he's greasy, he's like cool looking
- 1:12:37guy and it's like. It's a.
- 1:12:40They provided the equipment. So it's like, all right, well,
- 1:12:42here's like the little hot shoe adapter and like, we'll put this
- 1:12:45light over here and, like, all right, take this photo.
- 1:12:47That's what's you know. Again, I feel like old man Mac,
- 1:12:51you know, tell them in my day you had to you had to wait for
- 1:12:54your photo to come back, you know, And like, I didn't know,
- 1:12:57it's like, is the 15th of a second good or is it a 30th of a
- 1:12:59second? Like, I don't know.
- 1:13:00I'll try a bunch. And then, like, I don't really
- 1:13:02remember what I did. I mean now he's sitting there as
- 1:13:04a you know train engineer takes a picture and it's like oh
- 1:13:07that's pretty cool. But like what if we tried like a
- 1:13:09little bit more open aperture and like what if I moved this
- 1:13:12light over here and then like 3 dimensionally like now you can
- 1:13:15understand the guy's face. It wasn't just like Picasso half
- 1:13:18cut off and and just that real time feedback is cool.
- 1:13:22So but but I also tell them you know the the basics are great to
- 1:13:28learn. You know I mentioned my friend
- 1:13:29Kevin who's the. The the design instructor and
- 1:13:33you know Kevin was in the Olympics a couple times totally
- 1:13:37previous life like doing track and field stuff.
- 1:13:39And so he, he was posting a picture of himself from a few
- 1:13:44years back where he was like throwing a hammer, right like
- 1:13:46going around in the circle and I was like oh God, my younger son
- 1:13:50can be a little bit impatient and I'm like, I hope.
- 1:13:53I hope this isn't him throwing a mouse in your class next
- 1:13:55semester. When he discovers like, that
- 1:13:58Illustrator has two arrow tools and they're different.
- 1:14:01And like, why is that? And it's like, I don't know,
- 1:14:03somebody in 1987 made a decision and that's just the world we
- 1:14:05have. So, so I know like A, the the
- 1:14:09kids will get to learn those basics, B they'll have to go
- 1:14:12through the frustration. But C like sometimes that's
- 1:14:16that's the best part in that it's it's only when you try to
- 1:14:19use something in practice. Right.
- 1:14:22Then you're like, Oh my God, this sucks.
- 1:14:24Like, that's why I came to Adobe.
- 1:14:26I'm like, do you guys not use these tools like Siri?
- 1:14:28Like, I mean, I love you, but like and I love the tools.
- 1:14:30But like, have you not been trying to like put Photoshop
- 1:14:33files into Flash? And they're like, yeah, not
- 1:14:35really. I'm like, bro, it is a 20 layer
- 1:14:37file. It took me 168 steps to get
- 1:14:40that. Not animated.
- 1:14:41Not at all. Literally just looking like this
- 1:14:44to looking like this in the other program, like and it's
- 1:14:48like. Part of the the value I would
- 1:14:50provide is trying to be a good storyteller and get get the
- 1:14:54engineers to understand. Like, like let's literally go to
- 1:14:57this lady's house Like Photoshop is taking 5 minutes to launch on
- 1:14:59her computer. Like well, her computer's screwy
- 1:15:01and it's like maybe, but like I bet she's not the only one.
- 1:15:05And if we can like like make that one person happy, I bet
- 1:15:10there'll be 10,000 other people. We never get to meet them.
- 1:15:13But you'll, you know, Helen and Helen in Wisconsin is like super
- 1:15:17appreciative, right. And and so it's like people,
- 1:15:20people don't, you know, this is a million cliches, right?
- 1:15:23But it's like even soldiers in war, it's like, well, you're
- 1:15:26fighting for the cause, sure. But really you're fighting for
- 1:15:28the guy next to you and vice versa, right?
- 1:15:30And and if you can make things like personal and meaningful,
- 1:15:33then then those things tend to get addressed.
- 1:15:36And and so ultimately with the, with the kids and the advice,
- 1:15:38it's like, you know, learn the basics, stay curious.
- 1:15:42Guess what It it none of it's going to stick around my job
- 1:15:45too. Our jobs, right.
- 1:15:46I mean a lot of this stuff that used to be hard and you, you
- 1:15:48sweat really hard. They don't Other people have to
- 1:15:50and it kind of sucks. It's like going to the gym and
- 1:15:53it's like I I struggle and I try to lift these weights and then
- 1:15:55it's like and then we come in and everybody gets an Ironman
- 1:15:58mech suit and it's like, hey, guess what?
- 1:16:00You can all lift £20,000 and it's like, oh, but by all means
- 1:16:03it's still, you should totally still keep coming to the gym
- 1:16:05because if you can add £50 and like really though, it's like it
- 1:16:09kind of cheapened the coin. And so there is that real sense
- 1:16:12of loss in mourning. But I also think then the dawn
- 1:16:17comes again. And then there's like, well,
- 1:16:20geez, dude, I had no idea that I could, like, take one letter
- 1:16:23form from like, an 1890 topographical map in a.
- 1:16:28Antique store and like instantly teach a robot what that is and
- 1:16:32then have it make me like, I don't know, a Penguin in the
- 1:16:36style of that letter from like, what does that even mean.
- 1:16:37That's doesn't make any sense. I'm like, I don't know.
- 1:16:39Let's find out that's cool, like and and then you could do like
- 1:16:4250 other things and it's just like so so it it's the learning
- 1:16:45to learn. It's like it it's it's even if
- 1:16:48fine except that maybe you have higher highs and lower lows but
- 1:16:51if you could just get through the lows, the highs tend to come
- 1:16:55back. You know there's there's
- 1:16:56something new for you to. Contribute to.
- 1:17:00There's one thing that just came out in the middle of all of this
- 1:17:04and I was thinking, I don't know if this is something Adobe has
- 1:17:06considered, but I do. I don't know.
- 1:17:08I just thought it would be a great day.
- 1:17:09I want to. I want to get your opinion on
- 1:17:10it. But an AI teacher per se?
- 1:17:12Because imagine using all these applications.
- 1:17:15They're there. They can be complex for someone
- 1:17:17who's just beginning and some might have not been tech savvy
- 1:17:21the entire life. Now they want to get into
- 1:17:22Photoshop. Oh my God, like you.
- 1:17:24How do I do this? Or?
- 1:17:25Maybe the AI could be watching you and then as you're doing
- 1:17:28something wrong is like, hey, why don't you go ahead and try
- 1:17:31it this way, Why don't you go ahead and try it this way.
- 1:17:32So it's literally like someone holding your hands as you're
- 1:17:35performing the tasks and as you mentioned all this while you're.
- 1:17:39Job and the engineer's job has been how can we simplify this?
- 1:17:43How can we make it easier for people to get to from point A to
- 1:17:45point B and use less amount of steps and also make sure they're
- 1:17:50doing it the right way? So what if there was an AI tool
- 1:17:53that can also help you and it's smart enough to know what you
- 1:17:56wanna do and also help you get there?
- 1:17:58That would be. I think that will make it much
- 1:18:00easier for a lot of people to get into Photoshop and
- 1:18:02Illustrator and After Effects because it's not like this giant
- 1:18:06complex thing and I don't need to go and watch 30,000 tutorials
- 1:18:09on YouTube. Maybe I just need to watch two
- 1:18:11and after that the AI will help me.
- 1:18:14Has. Has Adobe considered that
- 1:18:16thought? Oh, oh, absolutely.
- 1:18:18I mean, it's funny because back in the day when Microsoft came
- 1:18:22out with Clippy, you know, the little talking dude, right?
- 1:18:27It's too. Soon.
- 1:18:28Maybe it was too soon. I think it was too soon.
- 1:18:31I think you're right. I mean that's, that's Marc
- 1:18:33Andreessen says, you know, being early is the same as being
- 1:18:35wrong. And again, it's like, yeah, we
- 1:18:37were, we were right. I mean the world did want like a
- 1:18:40good motion graphics tool for Web standards.
- 1:18:42There just weren't any Web standards.
- 1:18:43So it's like there's stuff probably from 2033 that we can't
- 1:18:47even imagine. And it's like we made a tool for
- 1:18:49that now. Good luck, right?
- 1:18:53Yeah, I I think. The notion of an assistant has
- 1:18:56has always been with us. Back when Clippy was around, I
- 1:18:58used to joke about making brushy the talking airbrush for
- 1:19:02Photoshop. And he'd be like a little, like
- 1:19:04a friendly little convoy airbrush and like, hey, you're
- 1:19:06trying to retouch a photo, you know?
- 1:19:09And it was like, it was just even worse, super annoying.
- 1:19:12But I think it is actually kind of possible.
- 1:19:14In fact, you know, when you guys were were giving me this like,
- 1:19:17very kind intro, I was like, oh God.
- 1:19:20You know, there's these actual people in those roles.
- 1:19:22So there's a guy named Alexandra Kostin who's a VP at Adobe and
- 1:19:27he, he's really the guy, you know, organizing and driving our
- 1:19:30our division. But I mentioned him because part
- 1:19:34of my journey back to Adobe from Google was in 2020 and it was
- 1:19:39because that team Alexandra had been leading worked on two
- 1:19:44things. One was neural filters in
- 1:19:47Photoshop, So like. Kind of like the first desktop
- 1:19:50cloud connection where you could do like this hybrid calculation
- 1:19:54and deliver models and they learn a lot in that process.
- 1:19:57The other was this Discover panel in Photoshop where it's
- 1:20:01more sort of task oriented and again didn't have Llms and all
- 1:20:04this hotness now but. It was a weird time for me
- 1:20:10because my I was working on like 3D cloud rendering and
- 1:20:12streaming. Cool.
- 1:20:13It was amazing. It's like no one cares.
- 1:20:15Like let me just bottom line this like no one no one needs a
- 1:20:18three, three GB Volvo streaming into their their their lap while
- 1:20:23they're lying in bed you know but but good try anyway, so I
- 1:20:26was a little bored and there that was going to go away.
- 1:20:28So I reached out to Alexandra and I was like hey look at these
- 1:20:31things you got into Photoshop. At the time my my old Adobe blog
- 1:20:35was still running and. I sent him a link to this PDF
- 1:20:38where I'd asked for feedback in like 2009 and I'm like, but you
- 1:20:42actually did it and I don't mean that like what we were first
- 1:20:44doing. It's like do it again it was
- 1:20:46before it's time it it just the stars didn't align.
- 1:20:50But but like I say the needs were there.
- 1:20:51It's like you go back to Aristotle like what makes a good
- 1:20:53story, right. I mean a story is a story like
- 1:20:56we're still mammals like we used to be and so I think by.
- 1:21:03By virtue of laying down these foundations, a lot of
- 1:21:06possibilities exist. I think the folks who worked on
- 1:21:08that would say, yeah, it was a step forward.
- 1:21:10It was not the be all end, all right.
- 1:21:12But maybe, you know, it's the was it Archimedes, you know,
- 1:21:16give me a place to stand and I'll move the world, right.
- 1:21:19And so it's like if we can not try to do everything ourselves
- 1:21:22but give people a place to stand then.
- 1:21:26You know, smart enterprising could be you guys.
- 1:21:28Who knows? Like you'd say, oh, what if I
- 1:21:31use Bard or Chechi PT And then I like, how can I train on this
- 1:21:34corpus, you know? And then like, what if, what if
- 1:21:36it was like, And again this is like literally me guessing and
- 1:21:39making stuff up, but like if you looked at all the Adobe held X
- 1:21:42tutorial content, right. And it's 30 years of people
- 1:21:44teaching each other how to do stuff.
- 1:21:46There have to be some best practices we can learn you and
- 1:21:48things, right? And so.
- 1:21:52I think just making it contextual like again the same
- 1:21:54same thing with Chandra to Phil. It's not like you couldn't go to
- 1:21:57Dali and and do stuff and then have brought it over you know
- 1:22:00before I mean same same I I keep waiting for mid journey to
- 1:22:04frankly explode when they they stop dragging around this like
- 1:22:08piano size weight behind them which is a a freaking command
- 1:22:11line interface from 1982 you know that that runs through like
- 1:22:14a gaming chat like it it they get a lot of right they they get
- 1:22:18a lot of. Social uplift from it And so it
- 1:22:21was probably a really brilliant call in retrospect.
- 1:22:25But to this day I'm like at some point I'm going to wake up in
- 1:22:27the mid journey.com just going to have a gooey right.
- 1:22:29I mean obviously like how could you not do that like and they're
- 1:22:32sneaking bits they. Haven't done it until they're.
- 1:22:34Doing. They're working on it, Yeah.
- 1:22:36Yeah. I met some guys at a party a
- 1:22:37couple months ago. They're working on it.
- 1:22:39But yeah, no, I It does seem like they could do it.
- 1:22:43They must have. They must have good reasons.
- 1:22:45I hope it'll be. I hope and fear that it'll be
- 1:22:47amazing, I'll be honest with you, right.
- 1:22:48Because like if somebody just wants to make creative tools,
- 1:22:50I'm like, oh God, that's probably going to be really
- 1:22:52cool. And it's like, but that's okay
- 1:22:54because honestly and sorry, it's another digression.
- 1:22:57But like the thing that that made Adobe Lightroom a success
- 1:23:01despite Adobe was Apple coming out with Aperture and Aperture
- 1:23:07is like long since cancelled. I mean a lot of the tech made
- 1:23:09its way in Apple Photos, so it didn't die.
- 1:23:10I mean it's it's good stuff, but like, they decided.
- 1:23:14Yeah, we're going to go after Adobe.
- 1:23:15Like, we're not going to go after Photoshop, but we're going
- 1:23:17to take what we know with iPhoto and we're going to bring it out
- 1:23:19as like its own new thing. And man, for three years I'd
- 1:23:24watched my my new venue, you know, Photoshop teammates who'd
- 1:23:28gone over to this Lightroom project, they were trying all
- 1:23:30this stuff, they were struggling.
- 1:23:32I was P Ming Camera RAW, which became the heart of of
- 1:23:34Lightroom. But it was, it wasn't even like
- 1:23:38the. They didn't know.
- 1:23:39It was like photographers didn't know.
- 1:23:41They just like people couldn't see it And and Apple, like we
- 1:23:44were saying earlier if you have that clarity of vision and you
- 1:23:46can tell the story that is worth gold.
- 1:23:50And and what they did is they wrapped the entire Javits Center
- 1:23:54in New York City like this Convention Center in banners
- 1:23:57that were must have been 40 feet high and they said Aperture,
- 1:24:00everything you need after the shoot, it's $500.
- 1:24:02It's not Photoshop. There's no layers.
- 1:24:04You're welcome. And you need a super sick Mac to
- 1:24:07run it. And and the world was like, Oh
- 1:24:09my God, OK that's cool. And then Adobe was like, oh
- 1:24:12shit, like, oh you know what we have one of those and and we
- 1:24:15went through this period of like, Oh my God, we take the L
- 1:24:18we blew it. We had this yet started we
- 1:24:20what's wrong with us? And it was like this like the
- 1:24:23the five stages of grief and and you know anger, denial,
- 1:24:26bargaining, all that stuff. And finally then in like stage
- 1:24:29six was like F you you're not doing this to us.
- 1:24:31And then like 6 weeks later, I was in Europe and in a different
- 1:24:35country every day being like this is what we're doing.
- 1:24:37We're going to come out with public beta.
- 1:24:38It's going to be free. You don't need a sick man, you
- 1:24:40know, I mean it was super limit, could even crop something, but
- 1:24:43it like lit such a huge fire. And so I don't know that you
- 1:24:46know mid Journey or whoever that people we haven't even heard of,
- 1:24:49you know, playground AI. I just saw their new stuff.
- 1:24:51I'm like, Oh my God, like you guys are like productizing these
- 1:24:53research papers at a phenomenal rate.
- 1:24:56If that lights a fire, great. Ultimately, guys.
- 1:24:59I mean, companies are imaginary, right?
- 1:25:00They do not exist. Nobody doesn't exist.
- 1:25:02Google, Apple, none of it is real, right?
- 1:25:05Countries don't exist. They're just people and and
- 1:25:07they're they're just ideas we come up with to try to get stuff
- 1:25:11done. And it it kind of works, even
- 1:25:13though we all know it's crazy and so.
- 1:25:16I am. I'm very fond of Adobe and I
- 1:25:18will try to use that idea to make good things.
- 1:25:20But but it is still a means to an end.
- 1:25:22It is still an idea in service of something else which is a a
- 1:25:26more beautiful fun world, right where people can can support
- 1:25:29themselves doing meaningful work, you know?
- 1:25:32And so if the way we get there best is a lot of people come at
- 1:25:36it and they compete with each other, they compete with us and
- 1:25:38vice versa, Great, let's do it. Let's look, I I've seen the
- 1:25:42company rally. Let's do it again.
- 1:25:46Like now we talked about how your vision is, but I want to
- 1:25:49talk about your mindset. So when you were at Photoshop
- 1:25:52and you were product manager at Photoshop, I would assume a lot
- 1:25:55of people were using Photoshop. You had probably feedback,
- 1:25:58constructive feedback, but there was no, I wouldn't call it hate,
- 1:26:01but they wouldn't. There was no skepticism.
- 1:26:04But now with a I the the environment is a bit different.
- 1:26:07The things that you are working on, the things that you are
- 1:26:09building has a lot of people who are worried about or are angry
- 1:26:13about at this stage. It does beat up competition,
- 1:26:16yeah. So much higher.
- 1:26:16Does that make a big difference in terms of how you ship product
- 1:26:20or design product? Because you know people, some
- 1:26:22people by default, they are not happy about what is coming out.
- 1:26:26So yeah. It wasn't like that with
- 1:26:27Photoshop, right? No, no.
- 1:26:29Again, like I say, you know, 3D people never, never came at us.
- 1:26:32Like, it's too hard. They were like, here's all these
- 1:26:34other complaints, right? And so with Photoshop, I mean,
- 1:26:38there was some discussion of, like, the ethics of morality.
- 1:26:41I mean, like. It's like you'd visit a customer
- 1:26:43who's Playboy and then, like, let me show you how I use
- 1:26:46Liquify Tool to, like, make women's waists slimmer and like,
- 1:26:50you'd look at, like a fashion retoucher.
- 1:26:52And I remember this example of a woman who was wearing these,
- 1:26:55like, really tight leather boots.
- 1:26:56And they were tight to the point, like the skin on her leg
- 1:26:58was sticking out the back. And it didn't look very nice.
- 1:27:00And they're like, look what I do, I just like, move, move the
- 1:27:03skin. I'm like, well, you've just
- 1:27:04helped mislead people. And say you're supposed to look
- 1:27:07like this. And when you wear the boots, you
- 1:27:08should look like this too. She looks like this, but she
- 1:27:11doesn't look like that. You know, she's probably super
- 1:27:13uncomfortable, just wants to wear some some flip flops, you
- 1:27:15know, But, like, so there were like ethical dimensions or like,
- 1:27:19you know, swapping faces or, you know, sort of fake news.
- 1:27:24I think you're right, though. Two things have happened.
- 1:27:26One, those applications have just become even more
- 1:27:29accessible. Right.
- 1:27:31So, so like, deep fakes and stuff.
- 1:27:33And then the other is like just a different class of concern,
- 1:27:36which is like, it is this all just fruit of the poison tree,
- 1:27:39You know, where it's like, yeah, I love that I can like remove
- 1:27:42the car next to the donkey. But the only way that was
- 1:27:45possible is like training on the photos of millions of people
- 1:27:49who, yes, they legally signed an agreement and we can do this.
- 1:27:53It's okay in, in that sense. But because the tech didn't
- 1:27:57exist when they shot the photo, you know, they didn't make a
- 1:27:59decision that like, I'm okay with using that.
- 1:28:02It's a it's a it's a real question.
- 1:28:04I mean Adobe is is trying to be super thoughtful with this.
- 1:28:07I mean ultimately like we are or we like to think of ourselves as
- 1:28:11creators right. A lot of us came to this because
- 1:28:13we we love doing photography or 3D or animation.
- 1:28:16Like the last thing we want to do is like screw the people who
- 1:28:20you know who are our friends and and the people we admire.
- 1:28:23I mean even if we were jerks like putting putting our the.
- 1:28:28Putting out of business the people who pay our salaries does
- 1:28:31not seem like a great idea, right.
- 1:28:33So even besides like trying to be good people, like there's a a
- 1:28:36selfish motivation right to to help creators still get paid.
- 1:28:41But having said it, it's hard, you know, I, I don't know, for
- 1:28:44example all the data sets that all these other models use.
- 1:28:47But you know, I think there's like data sets like Lyon which
- 1:28:50are basically scraping the Internet.
- 1:28:51I saw a thing yesterday where I guess Google has changed their
- 1:28:54terms of service. Where?
- 1:28:56Basically they're like, yeah, we kind of reserve the right to
- 1:29:00train on data at scale. And you know, the the the way
- 1:29:05that ChatGPT and other things work, they, you know, they suck
- 1:29:08up the New York Times. Maybe that's what was with this
- 1:29:10whole weird Twitter quasi outage over the weekend.
- 1:29:13It was like, I see people speculating about that.
- 1:29:15And so I, I don't know, I, I mean there is a spectrum, right?
- 1:29:23I can definitely, like everybody agree okay, we don't train on
- 1:29:25like. You know pornography or
- 1:29:26something. OK, obviously great.
- 1:29:28And at the other end it's like you could have a world where the
- 1:29:31only way we train is to get individuals to affirmatively
- 1:29:35like take their stuff and volunteer it from the jump like.
- 1:29:38So we have a training corpus of 0 and we start there and we just
- 1:29:41hope that out of goodwill or or money people and it's like OK,
- 1:29:45but you know Lion is out there stable diffusion Dolly mid
- 1:29:49journey like. Part of what they're why they're
- 1:29:51so good is they've seen it all right.
- 1:29:53They've seen things that Firefly hasn't seen.
- 1:29:55And as a result like you can type in Norman Rockwell or can
- 1:29:58of coke or you know the Pope wearing Balenciaga.
- 1:30:01It's like well we've never seen the Pope and we've never seen
- 1:30:03Balenciaga. So we maybe we'll like we we'll
- 1:30:06like give you an old man in a in a puffy coat.
- 1:30:09But it's not him and it's not Balenciaga.
- 1:30:11And so I, I I what I hope we can do is.
- 1:30:17I mean it'll be active conversation for, for years, I
- 1:30:20think. I I hope part of what we can do
- 1:30:23is give people the tools for them to make decisions.
- 1:30:26You know, I mentioned the custom training thing and let's say you
- 1:30:28really do want to have this specific Pope in that specific
- 1:30:31scenario, right? Or you are Balenciaga and you
- 1:30:33want that specific coat. You know, if if you have the I P
- 1:30:39right, like why shouldn't you be able to use your IP, you know,
- 1:30:43like Balenciaga? Great.
- 1:30:43Go train your own model, put that into scenarios.
- 1:30:47You don't need us to do it without consent, necessarily.
- 1:30:51If you can kind of like put the chocolate in the peanut butter,
- 1:30:54maybe that's more more difficult in some ways.
- 1:30:56It's also more powerful. And, you know, part of what
- 1:30:59really excites me is this, this notion of, like I said at the
- 1:31:02beginning, like inspiration, a creative ecosystem.
- 1:31:05People don't know what they can do because what they can do is
- 1:31:07different than what they could do on Monday.
- 1:31:09Right? So, like, how do we let people
- 1:31:10know that and then take action? You know, I don't want this
- 1:31:13world, this, this will seem so quaint.
- 1:31:15It'll be like, hey, remember the 90s?
- 1:31:16And like, we all watched friends, you know, at some point
- 1:31:19it'll be like, hey, do you remember alt text?
- 1:31:21And, like, you would see a cool image on Twitter and you'd have
- 1:31:23to, like, click the alt thing and hope somebody put text in
- 1:31:25there, and then you copy. And then you'd launch that
- 1:31:27gaming. Chat and then you go try to find
- 1:31:29that bot and you type in slash imagine and you paste that and
- 1:31:32then you cross your fingers like that seems insane, right?
- 1:31:35Like why is it? Here's the picture.
- 1:31:37I click on a link with the picture and then I have made the
- 1:31:39thing great. And now from there I can modify
- 1:31:43and I can mutate but like just that absurd number of of of
- 1:31:49steps. Sure, maybe in in you know doing
- 1:31:5110 things to get a swagged out Pope.
- 1:31:53OK and. In the limit, that's not that
- 1:31:55many steps, but it's it's about 9 1/2 more than you should have
- 1:31:58to do, right? So, So I think, I think we'll
- 1:32:01get to a place where as these tools mature and we have the
- 1:32:04multimodal stuff. You have sketching, you have 3D,
- 1:32:06you have depth, you have photo references.
- 1:32:09Even if model makers like Adobe don't take it upon ourselves to
- 1:32:13put in all of these entities, right?
- 1:32:16If we can have more of a, you know, connect the dots like
- 1:32:19combined tools foundation. That will open the door, you
- 1:32:24know, to people then using their own judgment and their own IP in
- 1:32:29in ways that are, I think, not good.
- 1:32:32Well, first of all, I just want to say thank you for sharing all
- 1:32:36of the thoughts with us. I I think after this
- 1:32:39conversation it's it's it's clear to me what Adobe's vision
- 1:32:43is in terms of the future. Of course we all understand and
- 1:32:47you mentioned this, it's an active.
- 1:32:49Questioning of the process because there is no.
- 1:32:52Right or wrong? Fast, Yeah.
- 1:32:53There is no right or wrong answer.
- 1:32:55We're all testing things out and as long as we try to do it in
- 1:32:58the most ethical way, and as long as the Adobe's goal is to
- 1:33:03ensure people can achieve their creative vision in a simplistic
- 1:33:08manner. And also while doing so they can
- 1:33:12try and test new cool methods to to get to the you know from A-Z
- 1:33:16and. Yeah, what's fascinating is how
- 1:33:19you guys are doing it. Again, a lot of people thought
- 1:33:21Adobe's just staying behind or not implementing all of these
- 1:33:25tools and you guys came out and knocked out of the park.
- 1:33:27So we're we're very glad to see the improvements that you guys
- 1:33:30are making to every single part of every single product.
- 1:33:33And we're going to look forward to all of these tools.
- 1:33:36And I mean, this is just a personal selfish thing, but I
- 1:33:39would love to see the idea of the AI assistant entertained
- 1:33:42within Adobe's softwares, like different softwares.
- 1:33:45I I would love to see that happen because.
- 1:33:47I know the pain I had to go through to learn some of the
- 1:33:50different things and just how much simpler it would be if I
- 1:33:53guess like a chat GB T like model trained specifically based
- 1:33:57on the 30 years of tutorials online and and and maybe even
- 1:34:01experts at Adobe themselves who have been using the tools.
- 1:34:05It just would be so. And and all of cool things come
- 1:34:07from the users, right? Like, yeah, auto complete of
- 1:34:10memes on Twitter. Like yeah, that that thing blue.
- 1:34:13That wasn't the Adobe thing, right.
- 1:34:15That was. Users came up with the idea
- 1:34:17totally, totally. That was that was.
- 1:34:19I think that that brought a lot of people to Firefly.
- 1:34:21Did you guys see a spike of users when that thing happened?
- 1:34:24I don't know like to me because I immediately saw everybody
- 1:34:28taking memes and doing that's. When people who were not.
- 1:34:31You know, like us, like AI tech nerds, people who have never
- 1:34:35even tried AI. That's when the first moment
- 1:34:37they were like, oh, this is cool, I I want to play right
- 1:34:40with this. Right, exactly.
- 1:34:41They. Don't even want to know what is
- 1:34:43AI and they don't want to know what.
- 1:34:44They don't care you and they should care.
- 1:34:46Yeah, but that's cool, yeah. Well, and and that.
- 1:34:48So what what you're getting at is, is it I do think a really
- 1:34:51interesting phenomenon, right? Like if you if you just present
- 1:34:56stuff from like the. You've got to boil the ocean,
- 1:34:59right. You've got to like do all of
- 1:35:01these and then then you can get somewhere.
- 1:35:02There's this famous saying from you know the world of business
- 1:35:05like a professor observed that you know if you're, if you're
- 1:35:09selling hardware, people don't want the quarter inch drill,
- 1:35:11they want the quarter inch hole, right.
- 1:35:13They they just want the result. They just, I mean I think I
- 1:35:16might have said this to you guys like I don't really know how my
- 1:35:18car works in fact. I'm sure the more I learn the
- 1:35:20more like, oh, I really don't know how my car works.
- 1:35:22You know, it would make sense of it like put in the fuel and
- 1:35:25there's some combustion and there's some air and there's
- 1:35:26some you know, torque and so forth.
- 1:35:29But that's about it. Like mainly I'm busy.
- 1:35:31I need to focus on like get my kids to school and so forth and
- 1:35:34so. But people could see the purpose
- 1:35:37of a car. Right.
- 1:35:38Again it's like we don't sell saddles like, oh, I can be free.
- 1:35:41I can get a job but you know no one has to debate that.
- 1:35:44I think with with a lot of a I what what we saw initially with
- 1:35:47Firefly especially is like. Oh my God, this exists.
- 1:35:50What do we do with it? I don't know.
- 1:35:52Let's. What's the first weird thing
- 1:35:53that popped in my Elon Musk made of pasta on Mars?
- 1:35:57Great. Go.
- 1:35:58Can it do it? I don't know.
- 1:35:59Was it good? I don't know.
- 1:36:01I I found from a year ago I said something about Tiger Woods in
- 1:36:05Dali and it made me a bunch of, well, there was like 1 Tiger
- 1:36:10Woods, the human, and then the rest were a bunch of tigers in
- 1:36:12the woods, you know, so, so I was like, OK, that's kind of
- 1:36:16interesting. But but more instead of the fool
- 1:36:20around and find out like you want to do something hopefully
- 1:36:22useful. And so I hope you know the meme
- 1:36:27thing is a little bit it's silly, it's fanciful, but people
- 1:36:30get it right. They they kind of know what
- 1:36:32these like visual references are and then it's like funny to
- 1:36:35combine them. And so when I say games again, I
- 1:36:39I have to duck because people like, wait a minute, we're not
- 1:36:41going to, we're not doing Tic tac toe.
- 1:36:42And I'm like, I don't want us to.
- 1:36:44What I want is to give people. Some end point that seems like
- 1:36:49relatable and attainable and interesting, you know and and
- 1:36:52I've always been fascinated by the phenomenon of Snapchat
- 1:36:55lenses where you know in a in in a world where 97% of people
- 1:37:00won't edit a photo and if they do they're just like going to
- 1:37:02crop out their arm because it looks fat or whatever.
- 1:37:06These companies have gotten hundreds of millions of people
- 1:37:08doing something creative every day.
- 1:37:10You know and and and they're not thinking like, well it's time
- 1:37:13for me to go and be creative and I have to you know practice the
- 1:37:15scales on the piano. And it it's like, no, I just, I
- 1:37:18I I wanna you know make my wife smile that that's like my main
- 1:37:23thing. Right.
- 1:37:24And so I just, I hire the app for that purpose.
- 1:37:26It launches. It's in camera mode and then
- 1:37:28there's like, oh, there's something Otter.
- 1:37:29So like oh she loves Otter. So like, oh, clip oh great, now
- 1:37:32I'm an Otter, right. And it's like that's awesome.
- 1:37:34Like, oh, you know Maria would also like that and probably my
- 1:37:36kids and like send it out. And I'm not sitting there trying
- 1:37:39to think like, I will now make fine art today.
- 1:37:41You know, it's just like, oh, here's a creative visual
- 1:37:43experience. Now, I don't think Firefly, you
- 1:37:46know, should be doing Snapchat filters.
- 1:37:47But I do think that notion of, like, purpose and focus and
- 1:37:51frictionlessness is really cool. And I would love to see a world
- 1:37:55where it's like, come for the X, stay for the Y, like, come for
- 1:37:59the memes, stay for the tools, you know, And if we can kind of
- 1:38:02get you in there, and then you're like.
- 1:38:04Like, you guys already know 3D. Most people don't.
- 1:38:06But let's say we have these literal building blocks in 3D
- 1:38:09and it's like there's a creative challenge and one I I should
- 1:38:13tweet this out. One of our designers or actually
- 1:38:15engineers, he was playing with that old how to draw an owl
- 1:38:19meme. You guys might say it's like
- 1:38:20first step, draw two circles, step in and then like, it's this
- 1:38:23beautiful owl. It's like now draw the rest of
- 1:38:25the ethic owl. Right.
- 1:38:26But just as a joke, he's like, OK, I'm going to take a cylinder
- 1:38:29and a sphere and put them together and say make an owl and
- 1:38:31it makes owls. And you're like, oh, my God, I
- 1:38:34didn't know. But the cost of entry could be
- 1:38:36that low, right. But it.
- 1:38:38But imagine you gave somebody a few building blocks.
- 1:38:40You know, there's some I did the other day where it's like a
- 1:38:42Taurus, right. That's like donut shape.
- 1:38:45And I piled a few of them around and I'm like, what does that
- 1:38:47look like? Oh, it looks like a donut.
- 1:38:49OK, cool. Make a few different Donuts.
- 1:38:51And then like, oh, it also kind of looks like.
- 1:38:53An inner tube, you know, on on the water.
- 1:38:55And I'm like, what if it was inner tubes?
- 1:38:57And then I was like, oh, what if it was actually Donuts, but the
- 1:38:59Donuts were the inner tubes and they were on the water.
- 1:39:01So I click and go into generative fill and and so that
- 1:39:04was just me screwing around. I mean I can't do 3D to save my
- 1:39:07life. But if you gave ordinary people
- 1:39:10a way to like, here's a little like a tiny micro creative
- 1:39:13challenge. Like what?
- 1:39:14What's the funniest thing you could do with just these three
- 1:39:17things? You see this on Twitter, It's
- 1:39:18like, you know, change one letter in a band name and ruin
- 1:39:20the band. You know, and people people have
- 1:39:23a lot of fun with that. And if there was something
- 1:39:26similar where like change one dimension of this creative
- 1:39:29composition and then you get X. What is XI?
- 1:39:32Don't know the answers. I know I'm talking to a game
- 1:39:34developer who is friends with the Cards Against Humanity folks
- 1:39:37and it's like just help us understand like what are those,
- 1:39:40what are those patterns? Because the purpose of Firefly
- 1:39:42is not to make memes, but it is, it is to draw people in and make
- 1:39:46them. More more creatively expressive
- 1:39:50and capable. And if hey, if memes got you
- 1:39:53started, great. You know, if like going in the
- 1:39:56the Yamaha, you know, keyboard store and pressing a button and
- 1:39:59hearing the demo got you thinking, oh, now I know what
- 1:40:01this thing sounds like. Cool, great.
- 1:40:05Like I I just, I think we can make this virtuous ecosystem of
- 1:40:08inspiration, you know, super attainability and then a really
- 1:40:12gentle on ramp into as much or as little creative tooling as
- 1:40:15you want, right? There's no right answer.
- 1:40:18What I might need from After Effects today and and five years
- 1:40:21from now could be different, could be higher today, could be
- 1:40:23lower than vice versa. If we can pair you with the
- 1:40:26experience that feels like, hey I got I got somewhere cool and
- 1:40:31it wasn't so trivial that like just push button get result and
- 1:40:33there's the thing you see with stuff like.
- 1:40:36What was it called? The, like, the AI selfie stuff?
- 1:40:38It is really cool, like people amazing.
- 1:40:39But it also felt kind of cheap. It's like, I take 5 pictures now
- 1:40:42I have 200 pieces of art. They're basically the same
- 1:40:45pieces everybody else has. And you saw a spike.
- 1:40:47I mean, this guy's made like $30 million apparently in a week.
- 1:40:50And that was that, right? And and I.
- 1:40:52Remember, the entire social media was filled out, everyone's
- 1:40:55profile photo changed for like a week.
- 1:40:56But literally down. Yeah, no, exactly.
- 1:40:59And you guys asked about like, did we see a spike with like,
- 1:41:01the mean thing? I don't.
- 1:41:02I haven't seen that with Photoshop.
- 1:41:03I did see. I was I I typed in Mid Journey
- 1:41:06on Google Trends just to see kind of like what's mindshare
- 1:41:08like Dolly has gone down you know like opening eyes focused
- 1:41:11more on on GB T but but there was a spike I want to say in
- 1:41:16like March with mid Journey and I'm going to go back and I'm
- 1:41:19like I bet that's when the swagged out poke thing happened
- 1:41:22because like they had hit with like 5.1 or 5.2 whatever, like a
- 1:41:26certain level of visual quality that people believed it was real
- 1:41:29and it was like a culturally funny like a resonant thing.
- 1:41:33Right. And it was it was the the union
- 1:41:35of those that made it go and it's come down some but it has
- 1:41:38still definitely it's way above where it was.
- 1:41:40So just as a tool for like making people think okay, I get
- 1:41:43what it's for I could do that. I want to try it that that does
- 1:41:48wonders And again where they're held back currently and I'm sure
- 1:41:51very temporarily is by. You know the the price of entry
- 1:41:55of this Discord UI. I mean, you don't have to
- 1:41:57download anything fancy, but you still have to type in a bunch of
- 1:42:00arcane jabber jabber, right? Certainly it will become
- 1:42:02possible. I'm I'm, I would bet in my next
- 1:42:05mortgage payment on this. You know when you see the shared
- 1:42:07thing you click a link and then like you're in the creation
- 1:42:10session, it's just it's just way too obvious not for everybody to
- 1:42:14do that. It's just a question of getting
- 1:42:15there when. You know, it's it's, it's
- 1:42:19definitely the way you. Plan it out and and and sort of
- 1:42:24wrote it out for us. That is definitely the way
- 1:42:26people should be getting into these things.
- 1:42:28And it's it's always about finding something cool, finding
- 1:42:31something useful, finding something funny.
- 1:42:33And then you go and test it out and then realize, Oh my God,
- 1:42:35there's this. Whole depth of other things I
- 1:42:38can. Relate to like now.
- 1:42:39I bet a lot of business owners, I'll make it formal.
- 1:42:42I'll make it like, not casual. A lot of business owners saw the
- 1:42:45meme and they were like, what if I can do that in my poster?
- 1:42:48What if I can do that in my banner ads?
- 1:42:51Yes, the meme looks fine, but do you believe that this can only
- 1:42:55happen by creators trying things out?
- 1:42:57Like big companies could never come out with all these things.
- 1:42:59It's just one random creator create one thing goes viral and
- 1:43:03it brings a lot of. Non casual application and use
- 1:43:07cases for this product. Totally, totally.
- 1:43:10It is. Then I remember like at the
- 1:43:13beginning of the pandemic, Sundar from Google was talking
- 1:43:15about telemedicine and he's like, I forget what the exact
- 1:43:18numbers were, but it was like, you know, prior to the pandemic,
- 1:43:20it was like 2% telemedicine. And then it went up to be like
- 1:43:2270 or 80%. And he's like now, I don't know
- 1:43:26what it'll come down. It won't be 70 or 80%, but it
- 1:43:29also won't be 2%, right? Like it'll it'll just kind of
- 1:43:31expanded people's comprehension of what's possible.
- 1:43:35And so like you say, I think people like, well that's
- 1:43:38interesting, I wouldn't do that, but it does make me think of X
- 1:43:41and I would not have thought of X.
- 1:43:42And in one other example, you know, just with generative fill
- 1:43:46in Photoshop, I'll, I'll be honest for a minute like.
- 1:43:50I've been admiring, like all these control net capabilities
- 1:43:53for a long time. We haven't been able to ship a
- 1:43:54lot of that just because we've been so busy scaling and like
- 1:43:57making sure it's like, oh, you know, on Tuesday we're going to
- 1:43:59run out of GPUs. Oh, crap, OK.
- 1:44:01The same as Instagram when they first launched, like so many
- 1:44:04more people used it. I think then we kind of
- 1:44:06bargained for. Is this the biggest challenge
- 1:44:08now? The GPU and computation?
- 1:44:10I I so, so there's probably some somebody who's you know running
- 1:44:15a fire drill that I don't even know about like making us look
- 1:44:17at. I I don't think so though I
- 1:44:19think. I think we're in a much better
- 1:44:20state but because of of the need to scale, we've kind of had to
- 1:44:24put a bunch of things on hold which would otherwise be really
- 1:44:26cool. So.
- 1:44:28So there's all these creative possibilities.
- 1:44:31I mean I've been showing examples just like me playing
- 1:44:33with like some hugging face demo and just like oh did you know I
- 1:44:37could like take this you know, vector clip art and like style
- 1:44:40it. In fact when I saw the guy this
- 1:44:41morning who was like. You know hey put a sketch in and
- 1:44:44get a a fashion rendering out of like I know what you're doing.
- 1:44:47Like you're charging 9 bucks a month for this.
- 1:44:49Like I could just go over to hugging face and I could just do
- 1:44:51it for free. But but hey man, shoot your shot
- 1:44:53like you let them cook. But somehow I got on to this I I
- 1:44:59guess oh it was in in Photoshop with generative film.
- 1:45:04People have found this like really funky.
- 1:45:06Technique where they're using Quick Mask, which again like
- 1:45:09predates layers. I mean this is like in the, you
- 1:45:11know, like in the human brain there's like the the mammalian
- 1:45:14brain and the lizard brain and there's like some, you know,
- 1:45:17like little watch battery in the back there that's like down
- 1:45:20there. But people are figuring out
- 1:45:21like, oh, I could do this and I could put like make a 40% opaque
- 1:45:25selection and then effectively do image to image.
- 1:45:28Conversions. And I was like, wait, wait, oh,
- 1:45:31I thought we were. We we haven't deployed any of
- 1:45:33that stuff. And they're like, yeah, don't
- 1:45:34worry about it. We would just like people like,
- 1:45:36hacked it. Hacked it in there, right.
- 1:45:38Or like I say, the automatic 1111 thing, It's like you don't
- 1:45:40have control of that. Well, whatever I do, I'm just
- 1:45:43going to stick it in there, right.
- 1:45:44And yeah, it's not going to be like super elegant, but it'll
- 1:45:47work. And so I I get super inspired.
- 1:45:51I mean, I'm sure it's like if you were Les Paul, right?
- 1:45:53And you were like making guitars and it's like.
- 1:45:55Jeez, I didn't think that like if you you know fed it through
- 1:45:57some like weird you know vacuum tube calculator thing like you'd
- 1:46:01get The Rolling Stones but you did.
- 1:46:03And as a you know as a as a creator that to me is one of the
- 1:46:09the greatest joys. Like it means you made something
- 1:46:12that isn't A1 Trick pony. It isn't.
- 1:46:13And again this is not a knock on Snapchat filters like, but like
- 1:46:16that's that's so far on the spectrum of like being an
- 1:46:18exactly blade. It can do exactly one thing and
- 1:46:21it's like, no, we want to do. We don't want to do like the
- 1:46:23full Swiss Army night for the toolbox, but we we do want
- 1:46:26something with more flexibility and then when people surprise
- 1:46:29you that's that's always a delight.
- 1:46:33Well, thank you so much, John, for everything that you
- 1:46:36mentioned. I want to encourage every single
- 1:46:39person who is watching and listening.
- 1:46:41If you haven't tried any of the Adobe, you know, Firefly tools
- 1:46:45that are out there right now, go ahead and try them.
- 1:46:48It's it's honestly going to be one of the most fun times for
- 1:46:51you to just get into it if it's the first time you're getting
- 1:46:54into it. If you've never tried, I would
- 1:46:55say it's the best place to begin with.
- 1:46:57And John, we're also going to put down your blog in the links.
- 1:47:01Below Twitter because he's Twitter, he's always posting all
- 1:47:05the latest stuff that they are doing with fireflies and a lot
- 1:47:08of other cool stuff that's happening in the A I board.
- 1:47:11I think that's a great source of knowledge for people if they
- 1:47:14want to get stay up to dated. Yeah, 100%.
- 1:47:16If you guys are looking for because this there's like.
- 1:47:18Thousands of newsletters you can look at but and and you you
- 1:47:21would know this out of everybody else it can get pretty
- 1:47:24overwhelming pretty quickly. But if you want to have a
- 1:47:27starting point John's blog would be an amazing place to start
- 1:47:30with. We we we found out about it
- 1:47:32recently ourselves and we we were just like you know reading
- 1:47:35everything. So if you guys are looking for
- 1:47:37something we definitely recommend that.
- 1:47:40Again John thank you so much. We appreciate all of your
- 1:47:43insights and personally we're looking forward to see all of
- 1:47:46the cool things that you and Adobe and the entire team.
- 1:47:48Was clearly working extremely hard to bring all of these cool
- 1:47:52creative tools to us. So we thank you guys so much all
- 1:47:56of you guys, and we're looking forward to see more stuff coming
- 1:47:59from you. Cool.
- 1:48:00And thank you guys. Yeah, I I think it's, you know
- 1:48:03the the world wouldn't be nearly as fun and interesting without
- 1:48:06you. Just yeah, bring in your own
- 1:48:09curiosity and helping spread the word.
- 1:48:10I think we all have have roles to play and so.
- 1:48:15Yeah, it's it's great to talk to you.
- 1:48:16Obviously we want this to be bidirectional.
- 1:48:18So yeah, check those things out, shoot me a tweet, DM, whatever.
- 1:48:22And again, that that's how I can help the team do their best
- 1:48:26work. It's just by understanding what
- 1:48:28what the heck is going on in the world.
- 1:48:30Like maybe I only stand you know understand like 18% of it, but
- 1:48:34at least if it's like 18 more than somebody would have seen,
- 1:48:37then it's a net win. So like please, please shoot me.
- 1:48:40Good stuff as you see it. Yeah, 100%.
- 1:48:43And again, just to touch on both of the things that you guys
- 1:48:46mentioned tweeting, John would probably do the best way.
- 1:48:49If you guys have some feedback, there's a very good chance he's
- 1:48:51going to see it. So make sure you go ahead and
- 1:48:53check it out. We'll leave all the links down
- 1:48:54in the comments section below. And for everybody watching and
- 1:48:57listening, thank you so much for staying through.
- 1:48:59We'll see you guys again next week.
- 1:49:02Until then, Ciao.