Welcome to Tech Stuff by'm moz Veloschen here with Cara Price.
Hello. Hello, So I just read this piece about how AI has already started displacing jobs, but it's actually not the jobs you would think.
So do we have something to celebrate this upcoming Labor Day? AI not stealing our jobs?
Well, notebook LM, as you know, is trying to make podcast hosts obsolete. But no. According to MIT's State of AI and Business Report, AI is replacing jobs that are usually outsourced to other countries. That's the short term. Long term, around twenty seven percent of white collar jobs will be eliminated. Yeah.
I know a number of people who are worried about the security and longevity of their jobs. But I also know people who are thinking about the kind of glass half full perspective about new jobs that AI might create. And we're sharing a conversation with one of those people today.
Yeah. You and I recently spoke with journalist Robert Kapps, who has been reporting on what jobs could exist in an AI driven future, and he actually came up with a list of twenty two jobs that don't necessarily exist yet but are likely to exist when humans in AI work sort of in a hybrid capacity.
Yeah.
He lists a few jobs that sounded pretty interesting, and then wrote up this article for the New York Times magazine. I think my personal favorite was probably AI plumber. That's the person who will figure out why an AI system might not be working the way it should be, and, as Rob puts it, we'll snake the pipes.
Yeah.
And I was happy to hear him say that tastemaker jobs will last a while.
Is that because you consider yourself a test maker?
Yes, a nondescript creative professional. And he says that those will continue to exist, So I'm very excited. He mentioned one job called a world designer, where a person fabricates an entire universe, complete with fictional characters and locations, which could apply to everything from marketing campaigns to video games.
Yeah, I remember him talking about that. He also said that he wrote the first draft of the piece using AI to see what would happen and his Robert describing how that went down with his editor, I.
Thought it would be a funny joke to play on my editor, Bill Wassick that you know, he assigned me the article and then like two hours later, I'm like, here you go, here's my invoice. I love this new future. No, it was a little bit of a luck, but it was also you know, you're thinking about AI and jobs, and you're thinking about how what are the future jobs? So I just thought it was a logic place to start with. Okay, well what will my role be as writer?
And you know, freelance journalism is a hard place to be I and we're not extremely well paid people anymore. So it's if I can write more, I can make more money, you know, just as a purely capitalistic play, like that's the dream, right, But of course that wasn't possible. It wasn't nearly good enough. And I would suggest this like to anybody out there who's very worried about it.
It's like, just go have it, do your job. Just try it, because like it'll teach you a lot about how far it still has to go, which can be surprising in our current type cycle.
Right to absolutely no one's surprised. The AI version of Robert's story was not published because he couldn't risk his reputation on an article where AI may or may not be hallucinating, or may or may not be exaggerating things or understating things.
You know, We've talked about cognitive offloading a bunch on this show, basically atrophying your own skills by using AI too much, and that's something that Robert said he's very conscious of.
When you start to really work with the AI, at some point you start to hand over your sense of taste and your sense of uh, like is this good. You sort of hand over the authority to the AI, and the AI doesn't have any capability to accept that authority. It will tell you that like, oh, yeah, this is great, this is true, this is but like it doesn't It's just a machine, right, so it doesn't. It doesn't really
have the ability to accept that moral responsibility. Like at some level there are just these elements that like have to come from the human because the AI is just not capable of providing them.
You know what I love the most is that after using AI, his next step was to reach out to a bunch of people, which I know is basic reporting, but I think it really speaks to the core of his argument that AI will need the human touch for quite some time to come right.
Human work will continue, but the nature of the work may adapt to the AI Revolution and Robert have some interesting speculations about what the jobs of the future might look like and how they fit into three distinct buckets, trust, integration, and taste.
So we started out by asking how he came up with these three buckets, And here's the rest of our conversation.
The first person I called for the article was the Ethan Molloch, who wrote the book Cointelligence, who's Professor Wharton and who thinks and writes a ton about AI, and I just wanted somebody to talk with about like, Hey, how should I even be approaching this? And like right away, even from that first conversation, I knew that this was going to be way more philosophical than anticipated, because he wasn't. He was basically like, there's no way I can tell you specific jobs.
Why did he say that he couldn't tell you specific jobs? What was the kind of intellectual exercise or missing step from where we are today to knowing what those jobs will be part of?
It is just too fast moving, you know. One of his big phrases they told me a couple of times, was like it depends on how good the AI gets
and how fast. Right. But then there are some things, and you know, he talks a lot about the jagged frontier of AI, which is sort of like the AI can be really good at some things and then just like really horrible at some things that you would expect even the base like estimating the word count of an article, right, like you expect any human to nail that pretty easily.
But when you think about all the different jobs, or even your job, or any job, what level are you bringing some sort of moral or technical or whatever responsibility to that job, Like you are signing off on something, you are the person saying like, yes, this is good and right and best and you know whatever it be. And that can exist in a lot of different things. In writing this article, it's I'm accepting responsibility for the truthfulness and accuracy of this article, and the AI can't.
It's just not part of our moral world. Right. You can't turn to the AI and blame it when something goes wrong. I mean you can, but it doesn't care. And so you know, you can think about the like extreme far end of that, in like autonomous warfare or something where like AI robots kill somebody, they can't accept moral responsibility. So the first category was trust and where are humans gonna still be very necessary? Whereas AI need them to authenticate and to provide that sort of trust.
I mean, to your point. This piece appeared in the New York Times magazine. And when we think about like the history of newspapers, or the history of publishing, or the history of publishing houses, like as this concept of impremoto, like I know if this article or this book or this movie comes from this studio or this producer or whatever, that I can trust it to a certain extent. The output of AI is often entirely divorced from its creator.
But you know, you mentioned in the piece some of the roles that may emerge in this trust bucket, like AI auditors, trust authenticators, AI ethicists, And I wanted to ask you, obviously, one of the things you have to think about is how much cultural demand will there be for these types of things. I mean, we're living in a world now of alternative facts and conspiracies, and I'm just wondering, do you think the demand signal will be there from the world, even if it should be morally?
I think that there. I think that there are a whole lot of tasks and a whole lot of things that we really would love AI to do, right like that that we would be just fine with AI doing and in fact, we probably already do a lot of them, right, Like I have AI transcribe my interviews, and I don't
think twice about the moral implications of doing that. And I think that like, as humans, we sometimes jump to these very extreme cases, right, Like I just used a warfare, but like you know, replacing human creativity and things like that, and yeah, we're gonna have to think very carefully about the lines there. But I think that there's a whole bunch of tasks that you're that you're just fine with.
And so you know, one example that I kind of reference in the piece in Trust is that, like, oh, I thought that like HVAC repairmen might be some of the last people to be affected by AI, But in fact, HVAC repairmen have to do a lot of paperwork, right, right, and a lot of things that's not really core to
what they enjoy about the job. But if they're using AI to do their contracts and to do their paperwork, at some point someone needs to validate that, like those contracts are accurate, that they're legally fair, because you can't
trust the AI, the AI is not worthy entity. So like even trust comes in there, like somebody needs to validate that, and it becomes a little bit different when you haven't created the thing, the contract or the whatever yourself, like the same problem I was having with the article.
It's like because you can't. You have to have a sort of slightly different set of skills to be able to be like I'm very familiar with where the jagged edge of AI what kind of mistakes it makes, and there really can be very weird mistakes, right, like unexpected not like it's not like backreading something a human rope, you know, and you can scale it up where it's more complicated than HVAC contracts, and it's something in an organization where they're where they have a whole chain of
systems and somebody has to really know the AI well enough to be able to go through that chain and give it the like human approval of like, yes, this
is this is trustworthy and accurate. And I think when I talk about ethicists, you know, a big organization might be integrating AI enough into the organization that they sort of have to have some level of explainability, right and some level of justification of like we let the AI make these decisions and not these decisions, and here's why we do it, and those things need to be explainable in a way that is satisfying to all kinds of constituents, right,
so investors, customers, you know. But it can be like if the organization ends up in court, right, they have to explain it to a judge and a jury in a way that they're like, okay, that's rational and ethically sustainable. So, you know, one of the things I like to say is that like the AI boon might actually be a big boost for philosophy majors who can think through the sort of philosophical implications of how the AI is used in an organization and create a rational chain of ethical
explainability for why it's done that way. Because so much of large corporations comes down to as they get bigger, they're just there's just more and more liability everywhere.
Maybe one day there will be an AI hallucination interpreter, which would be very interesting. I think, I don't know. The second bucket is integration, and these jobs seem to be more technical in nature. Can you talk a little bit more about the integration category and what those jobs will be like?
Sure? So another expert I talked to was a fellow named Robert Siemens, who's a professor at New York University who studies a lot of this, and he's like, well, certainly there's going to be some technical rules, right, like people need to understand the AI, and not just like we know how to build models. We know how to
build AI. But these are sort of the most immediate jobs that you can see coming up that'll be very big for the next some amount of years, you know, and they may shrink as the AI changes, right, we may need less integrators over time, but like right from the jump, you need someone at your organization who really understands the models, can really dig down, can really understand what the I is doing and why, and can help map it to you know, the specific company's peculiarities and
needs and wants and desires, right, And so one of the first ones he talked about was just the AI auditor, someone who can go and just really create some sort of understanding of the AI for people in the business. So you know, you can almost think of these as like a sort of twist on or an addition to
your typical IT manager, but isn't quite so. Just technically focused, right, They're just like, Okay, our sales teams and models are not showing this, We're not hitting the optimization right, So just keeping up with the models and like which which one is actually best at algebra right now? Right? Like which one is actually best at writing right now? Right? It changes every few months. Just even keeping up with that if it changes at this pace, can be a
pretty substantial job. So you can see AI being a great optimizer and a great tool, but someone needs to understand the system enough to work with the ad to make sure that like, oh, we're having this problem, like people aren't washing their hands enough. How can AI help us in a hospital setting? Someone needs to be thinking about, like how to make the AI get to the outcomes
that they want to see. Because it's a very powerful tool for some of these things, but you know, it needs someone kind of prompting it and helping and integrating it to these very very complex, big organizations.
After the break, why taste making will be the industry of the future, stay with us.
So the first bucket was trust essentially human in the loop, a category of jobs around ultimately like taking responsibility right or being the final arbiter of like what is just what is legal, what is correct. The second bucket is integration, which is essentially like, how do we know enough about these tools to harness them, effectively.
Optimize them and harness.
The third bucket is taste, and this one I think relates most directly to what me, you and Kara do for a living. And I was curious, how did you choose that word and what does it mean in the context of your pend the jobs that may emerge.
Yeah, I mean, I think taste is going to be very core to a lot of things, and not just creative jobs. It's also the one that I feel like people kind of perk up because it's it's also sort of humanly appealing. But you know, as I was just thinking about it and talking to people, and I use this in the piece, I just had this viral e clip of Rick Rubin in my head right of him on sixty Minutes and Anderson Cooper talking to him and saying like, do you know how to work a soundboard?
And he's like, no, you know, want to play instruments?
No?
Do you know anything about music?
No?
And he's like, well, what what do you do? And his answer, I'll paraphrase, I don't have in front of me, is is you know, I am very confident in my taste, right, and so like I had this in my head because you know, just thinking about what do we still need humans for it? And at a base level, like the AI doesn't want anything on its own, right, Like, unprompted, it will just sit there idle. Right. So at some point, the basic human AI interaction is the human asking the
AI for something. And to me, one logical route you go down if you explore that enough, is that the human is providing the taste for what it wants created and by I mean created. It could be a creative thing like a piece of music, or it could be a non creative thing like a business process. Right, But at some point somebody is looking at it and saying like, hey, I have the vision for what I want. But what they are really doing is they're making creative choices. And
they're not wrong or right choices. They're choices of aesthetic or their choices of function, right, Like, so there's multiple ways to find a solution, but they're sort of using their taste and their judgment to make these creative decisions to get to their outcome.
Cara I'm curious for your take here because you know, wearing your other hat, you're a TV producer, successful TV producer. How do you think about this idea of taste being a key place of human irreplaceability.
Taste is certainly I think the final frontier to be messed with. I try to think of it. Does AI have taste? AI has taste insofar as what is fed to a large language model. If you feed a lot of Beethoven to a large language model and it's spinning out music that is supposed to be modeled after Beethoven or Bach or whoever, you still need a human being who's going to decide is this music good. I think things that are synthetically made by large language models can
be good. But I still think there's someone who's deciding ultimately if that thing is good or not. And that's a person.
And that's what I mean, at some level of someone's making those decisions. And I think, like, again, there's these far out examples of like, oh, I use the AI to do everything, and that's going to have a certain quality. But there's like, oh, I use the AI to do something, and I do some things myself, right, I do something's analog and that's going to have a different quality, right, Like those can both exist?
Do you find that quality that qceing has become a harder job? Like I do notice that just even in LinkedIn posts or in Instagram posts, people are relying more on chatch ept to create the language that they're using. And there is a sort of en shitification of things because we have become used to just accepting that things aren't as good anymore because people are using chat ept to produce content. I know, is there a job that exists to push back on in shitification?
I guess. I mean, like I've done enough writing experiments now that I can tell, especially like LinkedIn is just rife with it. Like there's just certain constructions M dash shit. The m dash is one but like I love m dasher is the one that gets me is the like it's not X, it's y. So like when you and you'll see these once I tell you this, you'll you'll see this on LinkedIn like every single post. It's like the future is not you know, apples and oranges, it's bananas.
It's like that that construction just standing on its own is like when the AI is trying to be like a muscular copywriter, especially Claude. It just it just it just loves that construction, right, And yeah, I think that it's funny because when I think about AI in creative fields, and I try to think about it sort of more optimistic in writing, Like there's this extreme example of like I just use it to just write the whole thing,
or I use it to do whatever. Like there's a lot of room between where we are and that being the outcome, and there's a lot of like sort of positive room between there. And so like one of the things that I also do is I work on documentary films and I'm making a documentary film right now about AI weapons. And it's going to cost a little over a million dollars to make that film, right, And that's
that's relatively inexpensive for a documentary film. And would I love it if AI could help me make that film for five hundred thousand dollars. Like there's a lot it can't do. It's not gonna like DP my shoots and stuff. But like, you know, I have an editor, and you know it's gonna take months and months to get the edit together, and I wouldn't want to replace that editor because that editor is a valuable story collaborator who's taste
I love. But what I love the tools to help him be able to do in ten weeks what takes him twenty because it can like do sort of first cuts quickly for him, and it can sort of like let him try things much faster, like so therefore we can make that documentary faster and we can move on to another one. I would absolutely love that, right, like, and I think that he would too, but like to go right to like, well, just let's just have the
AI edit the whole thing and it'll be super cheap. Yeah, but it it'll suck.
Now, But what about what about with this piece though? Because V one you knocked up in two hours using AI a little bit maybe came in a little bit of a strung man in the piece, right, But on V two that you wrote yourself, like, were there places where you helpfully leveraged AI to make it better? And if not, like, when do you imagine starting to do that and in what way?
So I didn't use AI in this piece for that, Like I didn't, And I would say one of the reasons is I have editors at the New York Times magazine who are very good.
And even better than GPT five.
But what I will say is I've done a lot of experiments with writing in both fiction and nonfiction, of trying to figure out where to use the AI, and what I find is that for me it becomes unglued very quickly in that even Male talks about this in the piece which I very much agree with. He says that he never lets the II create a first draft, and that's something that people have talked about, like, oh, use it to get over your writer's block, how it
created a first draft. But like his point, which I find for myself too, is that the AI starts to dominate your thinking. It puts you in the AI box and then you're like, well is this where I would have gone without the II? And it's sort of like you end up in your own exist crisis. But then I find even as I use as an editor, I'm sort of saying like, hey, is this good? Is this?
And it'll start to give some suggestions like oh, that's great, like you could probably lay on this point a little harder, or like maybe this is a good opportunity to introduce some moral ambiguity or whatever. But as you as you start using it to do that. You're handing that taste to the AI, right in the same way that like when I write a piece of New York Times, I give it to my editor because I'm so close to it that I'm like, is this is this even any good?
Am I any good? Have I ever been any good at anything? In my life? That's sort of the writers just sent into madness. But you're trusting your editor now to tell you, like, yes, this is very good, right that this part's working. This part isn't working. I find it really fraught to hand that to the AI because it really doesn't know, but it will tell you that it knows.
When we look at the three buckets in this article which I'll trust, integration, and taste, one of the things that strikes means is that these have a somewhat pyramid like structure, right, So like, ultimately, when it comes to trust, somebody has to take responsibility legally, morally or otherwise for outputs. When it comes to integration, like somebody has to diligence the tools and decide which ones are helpful and how to integrate them. And when it comes to taste, somebody
has to be a tastemaker. Somebody has to be a talented creative or an experience creative, whether it's in music
or organizational design or writing or whatever it is. So I guess my question is how much of how many of the jobs, because you also came up with a concept of the AI plumber, right, how many of the jobs that will emerge in the AI revolution are kind of very much for the top of the pyramid types of jobs, And was that a consideration the piece, Like, even if there are these new jobs created, will they make up for the jobs lost?
You know, I would say that, like the piece is going to be kind of and I'm going to be kind of woeful at answering that because I think while I did it under the structure of like listing some new jobs, like really my hope was to sort of like help people philosophically think about like where new jobs will be and maybe even what their own job will be, and like what interacting with AI will be like. As I say, I think that part of it is like
thinking about where the AI needs humans. It's not so much like I try to resist the idea of its humans serving the AI, because again, the AI doesn't want anything if you follow the chain far enough, eventually you get to a human that wants the thing. Right, So it's like, where does the AI need the human? It needs some trust, and it doesn't really understand your business. It needs some integration, and it doesn't really understand what
you want, so it needs your taste. So there, it's hard to say, like how much will be sort of AI plumber versus like sort of setting the taste. You know, I think that probably ultimately, like the taste maker part of it is so fundamental to human creation of any kind, that like that will be the longest lasting, right, Like I think integration might be the shortest lasting, but like
could still be decades. It's interesting because every single person I talk to express serious trepidation about where we are headed in terms of jobs. All of them would say AI is going to result in more prosperity, is going to increase wealth. But will that prosperity and wealth accrue to capital or will that or will it accrue to labor? Right? And that's the big unknown that I think people really kind of wring their hands about. And I think that like we are definitely at the place where it could
go either way. I am a little bit of a firm believer in you know, sort of the phrase the cars go where the eyes go, we will go where we sort of collectively point ourselves to go. Right, None of this is foreseeing the technology will not It does not make it guaranteed to go one way or the other. I think that in theory, right you have a bunch of companies working on this, A bunch of them are
playing with open source. In theory, the tools will be commodities, which is to say, unless open AI and a less anthropic and everybody decides that they get to a model that's so good that they just close it off for themselves. Like we'll all have access to these tools. That should in theory, and this might not be very comforting, but in theory, that should be democratizing, right, like the Internet was.
That should enable more of us to be able to do more things, and that should be actually threatening to bigger organizations. So like, my hope is that we should be entering a very entrepreneurial age where like small teams of people can take on big encumbered people. But I don't think that like that. We're in this weird quirk area right now where doesn't look like that, right, Like, what it looks like is we have these massive monolists that are just getting stronger and more impenetrable, which I
think is part of why it seems so scary. Right, They're just accumulate cash till the cows come home, and the rest of us are going to be poor. And I do think that, like in this sort of one sense of like being a corporate cog in one of those massive enterprises, Like there's never been a worse future for that. But like if you can start to think entrepreneurial, if you can start to think about, like, hey, how
can I use these tools to do something that excites me? Right, that I'm really thrilled about, Like I want to make documentary films, right, Like, and it's so expensive, but like, hey, maybe they can come down and I can start to tell the stories I want to tell in the ways I want to tell them, right, Like that's just for me. Like, you know, my hope is that like there's all sorts of stuff that we don't see coming that is going
to be really empowering to people. I don't know that that's true, right, there's certainly enough worry about just it's just capital building on itself right, which will require some other kind of intervention that is fairly bleak about. But again I think the tools will be commodities, not the people.
Robert Capps, thank you for joining us on tech Stuff.
Thank you, thanks for having me.
That's it for this week for tech Stuff.
I'm Kara Price and I'm mos Valasha And this episode was produced by Eliza Dennis, Tyler Hill and Melissa Slaughter. It was executive produced by me, Kara Price and Kate Osborne for Kaleidoscope and Katria Norvelle for iHeart Podcasts. The engineer is Beheid Fraser and Jack Insley mixed this episode. Kyle Murdoch wrote our theme song. Please do rate, review and reach out to us at tech Stuff podcast at gmail dot com. We love hearing from you.