Algorithms + Data Structures = Programs - Episode 250: Sean Parent on AI

Episode Date: September 5, 2025

In this episode, Conor and Bryce interview Sean Parent about his thoughts on AI, its impact on the software industry and society, and more!Link to Episode 250 on WebsiteDiscuss this episode, leave a comment, or ask a question (on GitHub)SocialsADSP: The Podcast: TwitterConor Hoekstra: Twitter | BlueSky | MastodonBryce Adelstein Lelbach: TwitterAbout the Guest:Sean Parent is a senior principal scientist and software architect managing Adobe's Software Technology Lab. Sean first joined Adobe in 1993 working on Photoshop and is one of the creators of Photoshop Mobile, Lightroom Mobile, and Lightroom Web. In 2009 Sean spent a year at Google working on Chrome OS before returning to Adobe. From 1988 through 1993 Sean worked at Apple, where he was part of the system software team that developed the technologies allowing Apple’s successful transition to PowerPC.Show NotesDate Recorded: 2025-08-21Date Released: 2025-09-05Snowcrash by Neal StephensonTech LayoffslumeWall-EAltered CarbonTerminatorIntro Song InfoMiss You by Sarah Jansen https://soundcloud.com/sarahjansenmusicCreative Commons — Attribution 3.0 Unported — CC BY 3.0Free Download / Stream: http://bit.ly/l-miss-youMusic promoted by Audio Library https://youtu.be/iYYxnasvfx8

We should have a device with the raw capabilities of a human brain in our pocket sometime between 20 to 40 years from now. Okay. And if you think about the ramifications of that for society, that's huge. But there is huge risk, and I think we are underestimating the risk, right? Right. When people talk about kind of, you know, oh, the existential threat of AI and think Terminator, I don't think it's Terminator. I think it's a massive wealth disparity and, you know, leading to war, right? Right. I think that's the existential threat. A civil war? Or, you know, overthrow the government type of, you know. Welcome to ADSP the podcast, episode 250, recorded on August 21st, 2025. My name is Connor, and today with my co-host, Bryce, we interview Sean Parent and get his thoughts on AI, its impact on our industry, society, and more.

So here's a question that it's been in the back of my head ever since you said that. that, you know, will your book, you know, maybe it's the last book that'll ever be read. What is your, like, you've had a very long, you know, experienced career? Like, what is your take on where the art industry is going? Because, you know, there's lots of folks that are like us. They're treating this stuff like rocket fuel. It's an amplifier of your productivity. Other folks are super against it.

And then, yeah, you made the comment as like, you know, are books going to be needed? Anyway, so I wonder if you have, like, larger thoughts on where the industry is headed and how people, I guess, maybe not how they should be feeling about it, just your thoughts and then people can feel how they want. Yeah, so I don't think it's an industry issue so much as a societal issue. And if I kind of do the back of the envelope calculations, we should have a device with the raw capabilities of a human brain in our pocket, sometime between 20 to, 40 years from now, okay? And if you think about the ramifications of that for society, that's huge. That changes everything and in a relatively short time frame.

And so there's all this like, like, you know, we're burning up the planet building data centers and building power plans to power these data centers. Well, by the time most of those come online, if you plot the curves, they won't be necessary. Great. When we think about a human brain, the advantages that a human brain has over our current silicon is it's a three-dimensional structure, which provides some amount of density. It's analog-based, which means you get more bang per neuron than you do if you're running your transistors as digital transistors. And the whole thing runs on about 20 watts. And we are not yet at the point where we're simulating the full number of neurons in a single human brain. even with our large LLM models. But we will get there relatively rapidly.

And then lots of people are working on, like, how do we build chips that are more 3D, right? How do we expand out? How do we build the cooling infrastructure to them? How do we combine the RAM with the computation within the chip instead of having it on separate buses? How do we do analog chips? All of these problems are being worked on,

and all of them are going to fall. So the advantages of the human brain are going to eventually lose out. What I don't think we know yet is how does intelligence scale? I think with even open AI, what you're seeing is a tapering. You're having to build significantly larger models to get marginal winds. And I suspect that intelligence scales by some quadratic function, right? which may mean that there's you know people talk about a super ai i think we'll eventually build ais that are smarter than us but i don't know how much smarter right there's going to be a

a law of diminishing returns on how that intelligence scales i also think there's a very fine line between you know intelligence and madness right right intelligence is being able to see patterns in data and being able to apply patterns you see from one set of data to another set of data. And, you know, madness has seen patterns that aren't there, right? Right? And there's a really fine line. And if you know super smart people, you see that line, right? So that might also be a limiting factor, right? We might build AI that start to see, you know, the pattern in pie, you know, and just reach a, reach a point where they're not useful because they construct patterns that don't actually exist to fit data, you know. So, so there's a lot of

unknowns. But I think everybody keeps focusing on like what happens to the computer science industry. Yeah, I was, I was just going to ask, John, this is a, I think, the toughest environment that I've seen for hiring in the CS industry in my time. And I'm just inundated by people asking for advice, looking for jobs. And I wonder whether you, as a far more experience than Connor and I, can maybe pontificate on what you think this means for the industry. And also, what should, what should like a new college grad do or somebody who's in school and thinking about a CS degree. But also, too, I think Sean was just about to say is that, like, there's a larger problem

that is a larger scope than CS. So feel free to answer both, but I'm curious to hear where you were headed before Bryce injected the CS. I think on this front, you know, we're the tip of the sphere. And the only reason why we're the tip of the sphere is because we're the ones building the systems. And so the problems we're solving first are the problems we have. But like I said, I think this is a societal issue, not a computer science issue, and it's going to become a larger societal issue rapidly, you know, and I don't know how society adjusts for that.

20 to 40 years put the kind of the late stages of my life, assuming I may get that far out. So this is more of a problem for you guys than it is for me. You know, that said, take it all with a bit of a grain of salt. When I switched my major in school from double E to computer science, my dad gave me this very long lecture about what a bad move that was because by the time I graduated college, computers would be programming themselves and I'd be out of a job. And that was just all around, you know, expert systems were the rage at the time. And you can look up what an expert system is, but it's basically just a decision tree and not particularly interesting these days. But that was the thinking at the time. But I do think that this is very much different, and the pace of change is incredibly rapid here.

And, you know, for a new computer science student, you know, what I would be saying is learn to use the tools, including AI. and use AI to learn and learn the fundamentals and learn when the AI is wrong. That will be probably the best bet you have to extend your career. But that would be kind of general advice to any college student, right? If you're coming out saying, I'm going to be a lawyer or an accountant or whatever, right? The legal industry is very much getting hit by the AI wave. So are medical doctors. So are everybody, right?

If you're a medical doctor, you have an incredible diagnostic tool sitting right next to you, you know, with a huge amount of knowledge and more than you could ever hold in your head. Plus, you know, with things like deep research, the ability to go do significant research on a given problem and come back with something well structured and well thought through. So I think focusing on AI or focusing on computer science is way too narrow. Like I get asked frequently, you know, do we hire more senior people so that they can use the AIs and be more productive or do we hire more junior people who are willing to learn the AIs and so they can grow into senior people who know how to use the AIs?

I'm like, it is way too soon to tell this, you know, the generation, rate that that's going on with the models right now is every six months the equation changes. That's nuts, right? So, and I think it will, I don't think it's, you know, a lot of people are like, everybody makes this mistake as people, you know, when you're at the start of a curve, you project things linearly, and you miss the fact that it's going to exponentially wrap up. And then when it's exponentially ramping up, you overshoot. almost everything works out to be an s curve right right exponential growth is not sustainable

right right almost nothing follows a linear path right right it's it's 10% better than it was the previous time not not a unit of 10 better than the previous time and so so you know and it will follow an exponential taper off at the top of that s curve and it's very hard to predict where you are in that curve right right now i think we are i don't think we are yet near the top we are some place my guess is bottom middle and so so the next five to ten years are going to be nuts and then it's going to taper off and and there's going to be a lot of bad decisions made along the way people are going to both underestimate the capabilities and then they're going to overestimate the capabilities and make decisions based off bad projections.

Do you have, I'm curious, I mean, this is, clearly you've spent, I mean, a lot of time thinking about this. Do you have, like, concrete examples or predictions of the kind of both mistakes and on the other side, like, not mistakes, but just like, you know, inevitabilities that are going to happen along this curve? So, like, like everything kind of in the, in the, in the political. structure, right, the mistakes that are frequently made in a war are fighting the last battle. And in political structures, the mistakes that are made are fighting the mistakes of the last

administration or the last round and not looking forward and planning for the future. So how we're approaching from kind of a political standpoint, you know, job growth and And profits for the wealthy and trickle-down effects, especially if you're in the U.S. these days, right? You know, my taxes are, I think, ridiculously low for how much money I make. And yet, our administration thinks that I need more money in my pocket and the people down the street from me need less. And that's ridiculous. And what they're trying to do is say, well, we're going to cut out infrastructure nets and try to provide reasons for these people to work where you're about to hit, a phase, and in some industries it's already hitting, where jobs, even high-level jobs,

are going to start to go away relatively fast. And so I think that's not planning for that and not getting out front of that from a societal nature is the underestimation that's going on right now. You know, what happens if you have 20% of your workforce, which is unemployable, okay, they'll burn the place down. Right. Picture Occupy Wall Street, but way worse. Way worse.

Yeah. So how do you navigate that and how do you avoid that? You know, I'm not saying that that's going to happen because people are very reactionary, right? Right. They will, they'll adjust as the problem starts to manifest, but it's kind of societal nature is to wait until the problem is critical before you address it. and then you probably overcorrect. So I think we're going to run into that. I think we need to give some serious thought.

Is the society that we want where there is a half dozen organizations that own the AI and basically everybody else is subservient to them? That's kind of the, I'm watching on Hulu, the alien earth. And that's kind of part of the background premise on that, which is that there's basically a small number of AI companies, six or seven, that run the entire planet. Basically, governments are gone. It's just just businesses in charge.

Or it reminds me of some of Neil Stevenson's books like Snow Crash. Yeah, yeah, yeah. It's, you know, or even if you just look, you know, you had kind of last century, you had robber barons and you had whoever controlled the transportation infrastructure ruled. So if you built the railroads, you were in charge. I think we're hitting a similar point in history, but greatly amplified. And I don't think our political structure is set up to try to address this, at least not now. In many ways, it's getting worse.

This is why all the tech people are, you know, big on the universal basic income, right? Yeah, yeah. Historically, I would have called myself conservative. Now, you know, I'm a registered Democrat, but I still consider myself kind of, you know, George Bush, the first conservative in nature. You know, I think the Republican Party is toxic as it stands now. But I think there is huge value in work.

I don't think people function well in a society where they don't have. to work at least the vast majority of people. There are social aspects to work. There's a sense of accomplishment. There's a sense of meaning and purpose that comes to your life through work. And so I don't think the answer is, is, hey, we'll just guarantee that you all have a home and food and a place to live. I think there should be a basic substrate that is guaranteed. But I don't think that's the answer. You will end up a huge number of societal problems from people who just, you know, have too much time on their hand, right? I think the pandemic kind of showed showed us the damage that that kind of isolation does

to mental health for society. So I think we need to figure out how do we make people productive, right? Right. How do we provide opportunities? How do we provide meaningful work? How do we provide a sense of accomplishment in a world where the machine can do it better. Uh-huh.

You know, and I think that's really hard, right? Oh, man, this is podcast gold, Sean. I mean, it's not even podcast gold. This is just pure gold. I mean, so many thoughts. And it makes me think about there was a guy I used to work with 13, 14 years ago in 2011, in my first. co-op when I

came to Toronto worked for an insurance company this guy named Paul Paul was great I loved Paul I didn't work directly with him or maybe I did

he was on my team but I never worked with him but he was at a couple functions and I remember him telling me once that because you know actuaries they get paid quite well and he's like well if you're an actuary

you have to be you have to be you have to lean left I was like well you don't have to you can choose whatever you want and he's like no no no you have to be like a whatever is liberal Democrat wherever you are and I was like

like, okay, what's your reasoning for that? And he's like, well, at some point, you know, we're going to get to this point where we're going to evolve, you know, with, I don't sure he was talking about AI, but he just thinks there's going to be, you know, kind of people that are working and people that aren't working. And in that society, like, you want welfare systems and universal basic income and you need a way to support everyone in society. Because otherwise, the actuaries are going to have to go and get a shotgun, go take over their local no frills, which is like a grocery store. And we're going to have to like fend for ourselves because like society is going to flip

the board game table, you know, proverbially. Because, yeah, it's not going to work out if you don't. Like we need to care about all humans in society. You can't just this whole I pay less taxes so I can be more rich. It just doesn't work. Anyways. And that's basically what you were touched. And I never even thought of the fact that right now, unfortunately, America is undoing a ton

of these safety nets and programs. and obviously that sucks, but I didn't even think about it in the context of, you know, unemployment could be skyrocketing in the next, I don't know, if it's five years, ten years, however long, but, I mean, if you go to, it's only the tech layoffs that it tracks,

but there's a website called trueup.io slash layoffs. I can put it in the chat, and you guys can both click on it, but it doesn't look great. It's basically, if you take the average, it's either a linear line going up or a parabolic curve that's growing. It is kind of biased because Intel laid off 21,000 people in July.

They track it by month, and Microsoft has been doing a bunch of layoffs. And they're pretty high on the software's dead. I can't remember if it was the CEO or CTO that said something. But anyways, yeah, a lot of thoughts. I don't really, I have no solution. I'm not sure. Are you, it sounds like you read quite a bit of history because you're quoting the, I don't know much about George Bush number one.

Bruttle at Microsoft. Brutal at Microsoft here. Yeah, I'll put a link in the show notes, folks, if you want to go check. And I guess if you go layoffs by year, I guess... Can you look at it up by company? You know, I was actually a voting adult during, you know, George Bush is the first. Oh. So you don't need to be a scholar in history.

I've lived a lot of history. You just have to have been alive for it and paying attention. And I guess I was born, I think, like in the middle of his term. So I think that's the H-Bush, right? Yes. Yeah. He was 88 to 92. Yeah.

So I was working at Apple during the Iraq War. So I, yeah, I think Bryce and I were both born then. So we weren't. Yes. Yeah, I guess should we try and end on like, is there hope? I mean we didn't even talk about we didn't even talk about the robots

you know eventually there's going to be robots that all these things are going to be plugged into you know yeah would you have one of the robots in your in your home uh sure

yeah I would love you I mean it's like I'm always like you know we were promised by the you know from the Jetsons that we were going to have flying cars and

and you know Rosie the maid who would vacuum our house so we would have more time for our creative endeavors and instead what we've got is uh uh uh you know image generation and song generation and so so we've automated the creative endeavors so we have more time to fold the laundry and vacuum the house um have you have you heard about the loom the loom yeah we know what a loom is It's spelled L-U-M-E

You got to go to the website And watch They can play a little video for a second And then you've got to tell me You gotta tell me your thoughts, both of you Okay You might have to cut out the navigate to the loom here No, I'm already watching

This is Just wait for it Yeah, maybe not Maybe not attached to the bed I'm like, I've already seen the robots that are like a version of this that are on wheels that go around and pick things up and put them in a way

and we're definitely getting that because I don't want to talk poorly of my beloved but she is a little less organized than me in terms of... I'll talk about I'll talk about Ramona. There are clothes everywhere. She literally will not do

the laundry. She will not do the laundry. I'm the only person does laundry in the house. But okay, so describe for the listener, boom, the video starts off where there's this bed in front of you, there's some furniture next to it, and there's these two lamps, tall lamps next to the bed. And somebody comes and leaves some clothes on the bed and they walk away. And then suddenly the lamps come alive and little claw hands stick out of the end of the lamps. And then the lamps sort of bend over

and they start folding up the clothes. And when I first saw this, I really thought it was a joke. because it, I mean, it's an interesting form factor. It's just, it does seem like it's a joke, right? Yeah. I want it, though. You need a different version of this, though, because you don't want it just to fold it and put it on the end of the bed.

You want it to go, like, organize your closet and hang everything and fold it there, right? Like, attached to the bed makes no sense. On the upbeat, you know, I think there's incredible stuff going on here, right? I think, you know, in every aspect of the AI, right? I use generative AI for a bunch of stuff, coding, you know, editing photos, being at Adobe, of course. Writing, it helps me out. I would love to have it driving my car. I don't have a self-driving car yet, but I'll get there. You know, I'm also

looking at, you know, probably 10 more years. I'm going to need a self-driving car. I'm hoping that that coincide. So, so I don't have just a fully automated car. Wait, you think in 10 years time, Sean, you will be in such a state where you would not be able to drive a car? I think I'm getting close enough now, right? I'm past 60 and feeling the age. So, so definitely feeling like I have slowed down a lot in the last five years. So yeah, you know, I think there's huge potential upside. I just think there is a challenge to navigate this. Societies may navigate it well.

I suspect it will be chaotic. Hopefully it will be chaotic and navigated well, but there is huge risk. And I think we are underestimating the risk, right? When people talk about kind of, you know, oh, the existential threat of AI and think Terminator, I don't think it's Terminator. I think it's a massive wealth disparity. and, you know, leading to war, right? Right. I think that's the existential threat.

A civil war? Or? No, overthrow the government type of, you know. Because, you know, if you have a huge amount of unemployment, you know. Yeah. Yeah. I think if you have a, you know, it looks more like, you know, the fall of the SARS in Russia. which unfortunately probably leads to you end up with an you know in a in a in a in a takeover like that you you end up with with also an undesirable political structure and that has to go through a period before it gets gets you know reformed or overthrowing and and you know overthrowing governments tends to be violent and messy I think we tend to think of society is progressing along a linear path but historically there's no evidence that that society always improves, right? Empires rise and fall.

You end up with dark ages. You end up with huge, huge periods of turmoil. I don't think there's any guarantees that we navigate this transition well. My hope is we do. My hope is it remains a great opportunity and maybe the AI has helped us navigate this better. When you said Terminator,

I was not Terminator, It's going to be more like, I thought you were going to say Wally, where everyone's floating around in their, you know, hover chairs, sipping on soda. But that's, that's naive of me, I guess, to think that everyone's going to accept their lack of purpose and meaning, and it's going to be closer to, on the way to that, people are like, no, this sucks. Like, why, yeah, why stand for this and they take matters into their own hands? I guess there's no movie for that yet. I've heard that Andrew Garfield has been cast to play Sam Altman in a movie called Artificial. Really? So maybe that's how that movie is going to end.

We don't know. Yeah. You know, I think Star Trek is kind of the canonical, you know, bright future society here, right? Of course, they're Star Trek fan. They had to go through a third world war to get there. But, you know, they have a vision of a society where people are, are meaningful and productive and then taking care of right so is there ubi in uh in star trek i've only see no money that's like a post money society right yeah it's like a post money society you have

uh uh they certainly have you know intelligent computers not nearly as intelligent as i think is what we're headed for it is funny how sci-fi always underestimates the computers and overestimates everything else yeah So wait, what are they, what are they, because I've seen, this is, I'm going to lose some nerd cred here, even though I've mentioned that, you know, we're reworking our way through, I'm reworking my way through, Hishima's seeing them for the first time, Star Wars. I've seen the latest Star Trek movies, which were starring Chris Pine and the very good looking Zachary Quinto, I think, was the guy that played Spock.

Anyways, I think there was one or two or three of those. I saw those. I don't think there was much talk of the infrastructure about society. monetarily or fiscally operated. So what are the basics of Star Trek lore? You know, they don't go into details as far as how society is structured or governed, but it's a worldwide government, a single worldwide government. It's a non-monetary, somewhat socialist structure of some form.

They ended up there after the third world. World War, which was a eugenics war, you know, a faction of society was modifying genes to try to basically become superhuman and divide the structure of society between the superhumans that ruled and the rest of the people. And that ended up in a big war. And was that where Con came from? Yeah, that's where Con came from. Okay, okay.

That was in the movie. I remember that. you know and and in many of the movies they have you know shots of their cities and san francisco is you know commonly comes up in the movies and it's like oh these beautiful cities with everybody nicely dressed and clean and walking around and they have lovely apartments and everything's automated but they still go to work every day and they all have jobs and they all have tasks to do so you know it's a it's a very uh uh idealistic view of society be so aim for that minus the the third world war yeah if we can if we can get there without the third world war that would be nice if we could stop threatening you know taking over greenland that would be nice sentences i wouldn't have thought we would say uh some number of years ago given canada's in in giving connor's in canada maybe we can start with threat not threatening to do things to Canada.

How is things in our 51st state up there? It's all right. We elected a very, very, you know, unknown right until he was nominated and then everyone fell in love with him, Prime Minister. He's great. I mean, I was always a fan of Trudeau. I might have lost a couple listeners there, but I thought he was a great PR guy, always did the right thing, shakes the hands and Tim Hortons. He was a great, you know. Is that the Canadian thing, it shakes the hands? You don't kiss the babies. You just go to the Tim Horton.

I think people misunderstand what politics is a lot of the time, and a lot of it is just PR. And I think, for the most part, he was great at PR. And his downfall was just that he was the incumbent, you know, and people were upset globally. And so it was his time to end. But he always did these things where he would go to Tim Hortons and shake hands with someone. I was like, that's just brilliant.

It's just brilliant. you know like everybody knows us for tim hortons nobody hates tim hortons like i'm not a huge fan of their donuts but nobody hates it it's a good place for a cheap cup of coffee and a donut and uh anyways we we uh we elected the new guy and he's been great just watching him and how he deals with your leader is amazing it's amazing his first interaction he comes out he was the first he was the first nation state leader to come out and start talking about how trump was doing great things you know You didn't actually give any specifics. He just said, I think you're, you've made some really bold, you've done some bold things,

and you could see Trump being happy about it. And then look at this, just in the last week, you've got all the leaders of the European countries copying basically what our guy did and saying a bunch of, oh, you're great, you're great. We're going to get our podcast shut down, though, you know. I guess South Park is, they're still being aired. So as long as South Park is up, we're probably safe. But you don't have to worry because you're not the one who's going to get thrown in the gulock you're you're up and they'd have to they'd have to you know come for you right uh

i mean i mean we shouldn't start making predictions about what what your president will and will not do because uh he's already done a bunch of stuff that i never thought i'd see in my lifetime from a you know a u.s sitting u. that is true that is true yeah yeah anyways i feel like we've kind of flown off the rail here but that was uh somewhat hopeful aim for aim for star track minus the minus the World War and hopefully definitely not Terminator or the other show that comes to mind is Project Carbon when you mention the wealth disparity is because that's kind of the wealth disparity gone to 100% or 200% and the rich people live above the sky and they're basically immortal because they can pop in and out their consciousness and they treat bodies as you know just like replaceable things and so the wealthy

people, they always have spare bodies and their consciousness backed up, whereas poor folks have none of that and death is real for them. Anyways, let's not do that. Let's not do that. Fun show to watch. Probably a terrible reality to live in. Be sure to check these show notes, either in your podcast app or at ADSP thepodcast.com for links to anything we mentioned in today's episode, as well as a link to a get-up discussion

where you can leave thoughts, comments, and questions. Thanks for listening. we hope you enjoyed and have a great day. Low quality, high quantity. That is the tagline of our podcast. That's not the tagline. Our tagline is chaos with sprinkles of information.