AA251 - AI Product Managers Are Just Product Managers

AI product manager jobs are everywhere - but are they really different from regular PM roles? 

In this episode of Arguing Agile, Product Manager Brian Orlando and Enterprise Business Agility Consultant Om Patel wade deep into the muck-filled pool of hype vs. reality around AI-specific product management roles.

Listen or watch to join us as we do the dirty work of discovering if most AI PM job descriptions are just copy-pasted PM responsibilities with 'AI' slapped on top, or if there's real insight to be found!

Stick around past the buzzword bingo to learn:
- the 'find and replace test' for job descriptions
- the meaning behind "probabilistic vs. deterministic"
- the real AI-specific skills that matter
- how a 20-50% salary premium (and beyond) are justified
- why we think continuous learning beats specialization

Whether you're a product manager considering rebranding or a hiring manager crafting job descriptions, this episode will help you cut through the noise. 

#ProductManagement #AIPM #CareerDevelopment

Marty Cagan: Inspired: How to Create Tech Products Customers Love, The Lean Startup, Sinan Aral: The Hype Machine, Teresa Torres: Continuous Discovery Habits

LINKS
YouTube: https://www.youtube.com/@arguingagile
Spotify: https://open.spotify.com/show/362QvYORmtZRKAeTAE57v3
Apple: https://podcasts.apple.com/us/podcast/agile-podcast/id1568557596

INTRO MUSIC
Toronto Is My Beat
By Whitewolf (Source: https://ccmixter.org/files/whitewolf225/60181)
CC BY 4.0 DEED (https://creativecommons.org/licenses/by/4.0/deed.en)

Transcript

0:00

um, for the prep for this podcast, I pulled up five AI product manager job postings that I saw on the old interwebs. And, you know what they said? Let me guess. They heard the letters AI in them at the very least. No, no, they said, gather requirements, prioritize backlogs, talk to customers, define roadmaps. Ooh, does that sound familiar? Yeah, that, well, that's just a product manager. That's why it sounds familiar. And it's just copy -pasted, with, AI, strapped to the hood, Okay. but Brian. I can't tell me there's nothing different about managing an AI product. I mean, if you want to have a nuanced discussion, yes, I can tell you there are nuances, like a product manager who can't learn a new business domain. That's not a product manager. we, we don't have database product managers or microservices product managers or, your typical segmentation here is just marketing segmentation plus resume keyword padding so okay uh so you're saying that the market is what's making this up i mean maybe uh but we guess we're gonna find out let's get into it to arguing Agile. If this is your first time, listen to the podcast, welcome. I am your host, product manager, Brian Orlando, and this is my host, the right, welcome back Sorcer of Sticky Notes, Mr. Ombatel. and business agility consultant to the stars. Sorry, I almost forgot. To the stars. To the stars. today we're talking about AI product managers. So by the end of today's episode, what we shoot for is we want the audience to be able to push aside the hype that comes from when companies slap AI onto the job of product managers. and able to justify, I don't know, budgeting, head count, both, all the above. and, by the end of this episode, you'll know exactly what AI specific skills. on the job title are real, and which ones are just repackaged PM fundamentals. That's what I'm saying, fundamentals. I think those in the audience that, are mature PMs would remember this whole thing happening around the digital product manager theme not too long ago. Mobile first. Mobile first, exactly. Right. So this is like, you know, we're here 2026. This is the new thing. So whether you are job hunting as a PM, whether you're in a role, looking, uh, to specialize or, or maybe push your boundaries. we'll have a framework here in this podcast. That's kind of our goal is to have a framework here in this podcast about the difference between a legitimate specialized role and a boatload of buzzwords. It was very difficult to say boatload. Boatload of buzzwords. A lot of hot air, basically. Yeah, right. here is a fun experiment. Oh, take any AI product manager job description that you find on the old internet, remove the word AI and then read the job description to see if the job description still makes sense. And spoiler alert, I already did this. The majority of them do not make any sense. So exactly what? our companies paying a premium for because i like we're going to skip over just saying on paper we all agree companies are paying a premium for the AI in the AI product manager part of a part of the job roles yeah and indisputably there's evidence that can you know pinpoint that as well right so what are they paying for that that's a good question this is what the majority of this podcast is all around is just say hovering around the difference between having AI as a prefix and not having air as a prefix right so I did a little bit of searching in between podcasts and the AI product managers as, as a role, again, a role with AI slapped onto the front of it, bolted onto the hood, right? They make, anywhere between 20 and 50 % more than a comparable product manager quote on the outside. So is there something here? something here for Well, there's the applicants. I mean, 20 to 50 % is a significant uptake right there. you know, if all it takes is two letters, prepended to a few places in my resume, hey, sign me up. However, we're going to figure out what really is going on under the hood today. so the main thing you're going to hear, in this category that I'll just throw out right now at the early, early, early part of the podcast. AI is fundamentally different than normal product management because AI products are probabilistic versus deterministic. That's the main thing you're going to hear probabilistic versus deterministic over and over and over again. You're going to hear that in this podcast. First of all, just want to throw it out now is the first point in the first category. Because it's meaningless to me. that is prevalent out there in the industry, right? However, when you look under the hood, it's always been probabilistic. even use the terminology though it's deterministic, we cannot determine the outcomes of many things. It is with good intent that we deliver features thinking that they will deliver the value we think they will. Right. But that's not really probabilistic because we often don't have really the evidence that says they will probably like this. But we do have the, the leading indicators, I guess, from our pilots, if we did any or any kind of market research, that says, we think they might like this. Let me, let's, let's back up a second and talk about probabilistic versus deterministic. Okay. If we solve your business problems, we, if we solve your business problems in a deterministic way, that means we have mapped out your business flow and can do XYZ and lead you to the solution. whereas the probabilistic solution is we take all of the people that are potentially solving your business flow and we take the median of all those activities and most likely the median of a lot of activities will solve your problem hopefully i think that's really the difference here is is i work a lot a lot of AI systems and developer type of working with AI systems and a little bit as a product manager in my career too um if you are working as a relying like if life or death is on the line and you're looking for a generative generative AI response to hook into for a life or death situation, I don't think the technology is there at that point. To say like every time I get a machine, every time I get a response back from the machine, it's rock solid and we can bet the patient's life on it. No, if that's what we mean by probabilistic versus deterministic is one is a roll of the dice and the other one is a well mapped out, path and flow to make sure that we. come to a predetermined path, then yeah, I agree with, I agree with this, which is, hey, they're fundamentally different is you shouldn't use AI for these things that are really, really important. Yeah, I think, for clarity's sake, we should define these terms, right? So deterministic means given the same inputs, you're always going to get the same outputs. Correct. Yeah. Right. Probabilistic on the other hand means given the same inputs, the outputs may vary. and the variance will depend on the degree to which your models were trained, the types of models you use, the types of the input, the quality of the input, the guardrails that you provide, etc. So there are variables there, right? But the upside is, sometimes probabilistic will find paths forward that you didn't even anticipate, which is, which is really the hope, right, behind going with AI in the first place. With AI teams, a product manager would require specialist skills, one would think, evaluating models or training models, et cetera, or at least having some awareness of how these models are trained. So I'm not suggesting that they're the ones that actually sit there and train the models. That, by the way, is a very specialist skill for data engineers, data scientists, etc. PMs should be at least aware of the domain. right what are the what are the implications of training specific models in specific ways where do you get your input from how reliable are the inputs etc so with those definitions out of the way life and death situations i don't know about you guys out there on the interwebs i would prefer to stay on the side of deterministic because i know what the output's going to be if it involves a life or death situation. All right, well, knowing what side you're on, that helps me defend the other side, which is saying like you, you, you just defined probabilistic versus deterministic, okay, without, without making a statement on the business that we're in that we decide it's okay to have probabilistic outcomes rather than spending a good amount of time on deterministic outcomes. We've determined that as a business that it's okay. So I'm gonna put that in the against category. I'll put that aside to say okay, you can't really like knock of that because I'm not in the business of keeping people alive. I'm in the business of you know, whatever, filling seats on an airplane or whatever, you know, whatever like things that aren't really less drastic. Yeah, they're not really that important. And then, which is probably a terrible example because you definitely can fill seats on an airplane in a deterministic way. And then the other free seats when you're out of your deterministic seats. can be up to probabilistic outcomes. Yeah, it was terrible example. but the other one, that, that you outlined was the managing AI teams and managing AI products, when it comes to model evels and training and building specific models, that's a different skill set than your general run of the mill software development team would have. So that sort of lends itself in the against category is you kind of do need a product manager who understands, you know, the, the team of data scientists may not necessarily be able to pump out a new model in a two-week sprint or whatever. You're right? And, and maybe they can't even show progress at the end of two weeks in terms of like a finished working released product to the end user. They might need longer time boxes and that might be a point of contention that other product managers might not understand. So those are two points. The other point, here in the against is, hey, like the market demands this role. whatever we're talking about might be like academic. People might believe it's academic because the markets demanded this. So it must be true. So it is. So it must be true. So say we all. That's right. So those are, those are some, those are some solid points. Some, some solid points came out of this one. I agree. Yeah. In the against category, yeah, absolutely. I mean, the other side of this one, if I'm going to give any ground, the other side of this one, a good PM, good product manager, someone who's curious, they're, they're looking to put themselves out of a job. curious person. They want to know how things work, and they're looking to they're, they're a move on to the next bigger problem. So they're learning new domains and mastering like new areas all the time. So the idea is you get a good person who's curious and there's no, like there's no horizon here. even, even model e -vails and training new models, that stuff is not. it's in depth for certain i'm not putting down data scientists and whatnot that stuff is not working on rockets to launch and go into space it's it can be learned in a relatively fast amount of time and with tools with open source tools believe it or not you can do creating new models you can grab an open source model off the shelf and train it yourself with with the right you know if they have the right video card and something like that yeah i kind of concurred that that these pms that are worth their salt you know they're the ones that float above the rest right these people definitely lead with learning velocity they can learn these new technologies, these new ways of working, etc. So that's always been the case though. Just because we have AI now, it's convenient to slot the AI title in whatever it is that's new and vogue now, but that's always been the case. I mentioned digital earlier, there were other. fads, if you will, not that they fizzled out. They were just kind of transitory, right? blockchain, for example. So we've had these things before. And people that are really true PMs that know that they need to learn and will take steps to make sure that they're actively learning. They will have that in their, PDP or whatever is that that's used these days. personal development plans. I'm not worried about the corporate. side of it. They will always prevail. They will always be above those people that simply come in and say, I can specialize into this now because those people will be like going from pocket to pocket on a pool table. Uh, and then that's never a good thing. Yeah, blockchain. So that's, I know. That's an interesting one. demand in the moment for a title again only in my mind like maybe other people don't agree with this like it doesn't validate that that's the the market a real title the fact that like yo you threw it in your lincoln profile you got some traction for eight months Great. Like you, you're, you're riding the wave. maybe the selling point there is, Brian, all of tech is a wave up and down. And you just gotta know when the wave peaks are happening, grab your surfboard. paddle on out. You gotta know when to paddle out, oh, that's what we're saying in this category. When, when blockchain rises and the keyword searches on Google, you gotta paddle out into the waves and then make yourself the blockchain expert. And then when the next wave crashes and the next wave comes in, you gotta paddle out and become the AI expert. I got a little bit stitch happening on each shoulder now. So with every wave this happens, with every technological wave this happens. So if you're... in the picture and you've been there for X number of years, you would know this, right? And you should look ahead and say, hey, prime face, I'm a product manager. I know the fundamentals of product management. Beyond that, it's digital, blockchain, AI, whatever is next. I know I can prevail, but there is a nuance here, and that is you cannot stand still. You need to learn the domain, you need to learn the language, you need to learn the concepts. But that doesn't mean you need to be the de facto expert deep down or any of this. You're not a crypto expert, right? You understand about blockchain and digital wallets and all of that, but you're not a crypto expert. Same thing with AI. You're not necessarily an expert how to train a model. That should be a movie, but how to train, how to train your model. Yeah. but, you know, but you understand the fundamentals is what I'm saying. It's going to be a boring movie. I understand that. I understand that. It's going to be very boring. It's going to be a lot of math involved and most people are going to sign up. Okay, so there's a, there's a takeaway here. let's quickly dip into it. Sure. So the takeaway is the next time that you see AI product manager in the job title. Run away. Get a different one. No, no, no, no, just kidding. Just kidding. run a find and replace test. That's what I'm going to say. Find and replace test. So number one, I'm going to say, remove every instance of AI from the job description. And then, ask yourself, is this a coherent job? Was it written by a maniac? That's what I'm going to say right now. We're, Where's Mike Miller? Mike Miller should be on this podcast right now. Oh, absolutely. Dang. I, we always forget when we're down here is Mike Miller should definitely be here to say, was it written coherently by someone who understands what the, AI decorator means and it's like actual differentiation in AI not just slapping AI you this baby can fit so much AI inside of the job description that they just wrote a meme about it in job description format. Just just replace every instance of the letters AI with the word marble and then look at it and say, what do I know about marbles? AI, what do I know about AI? Can I still prevail as a PM knowing what I know? And if you don't think you can, guess what? You have a Delta, that's your homework. Same thing with the next fad in the next fad and the next fad and the next fad. I mean. So the number three, the last thing here, look for responsibilities that only exist in the context of AI or whatever technology fad that might happen when people watch this fill in the gap fill in the gap here of whatever technology you might be dealing with model evaluation criteria data labeling strategies bias auditing whatever is popular at the time You know, are, are these real things that you now have the call over for the whole business. Because if it is, then great, it's an AI product manager. otherwise. product manager with mean, it's just a some duct tape on it. Transparent duct tape. Transparent duct tape. I mean, what do you think? have you seen AI product manager, buckled onto the job description? Or have you seen the job roles that you looked at for product manager, AI job roles? Have you seen them actually be different? Let us know in the comments if we got this wrong. Let us know in the comments if I'm wrong here, where I, maybe I didn't look at enough AI product jobs. That is certainly true. I might not have looked at enough of them. Because once I look at like five or six, I just can't look anymore. That's a decent sample size, in my opinion. yeah, let's know if you've come across actual. nuances where you believe that to be true and we'll look at those separately and we'll see if you want to deep dive into another episode on those ones. Okay, so we've roasted the job descriptions. But maybe it's time to be fair. Is there anything legitimate in these job descriptions that would make them different for managing an AI product as opposed to like a regular product management product that a regular PM wouldn't already know? let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, let's, we love to steal man for a second. Let's think, if AI product managers really do behave different than, traditional. You I did air quotes. Air quotes, I did air quotes and completely evaporated a normal segment in the market that Marty Kagan wrote books about and Teresa Torres wrote books about it. They're, they're traditional now. Ah, you like that? AI. That's how segmentation works, kids. yeah. Anyway, models, if outputs are probabilistic. I'm going to keep, I'm going to keep, I'm going to keep using that word, then, and if data is also a big part of the product, because again, we have data engineering, bad data means, you know, bad probabilistic outcomes. I said it again, but I need a counter every time I say probabilistic in this. Should be a drinking game. Does that mean, that you need a whole different kind of product manager. That's, that's basically the question that we've asked so far. you just need a product manager who does their homework and and does all their extra credit assignments and turns them in on time that's what i'm asking if i was hiring i know which one i would hire again, the steel man arguments sound a lot like the arguments. We have already heard. Yeah, AI outputs are probabilistic. Once again, I'm gonna keep saying probabilistic. You know, if I say or doesn't mean that probabilistic enough times. Number one, I lose the ability to say it. Number two, it loses all meaning if I just keep saying it over and over and over again. But if I'm going to say something that's not probabilistic, I don't even think I'm saying that word anymore. Probable, probabilistic, probabilistic, paralytic. Paralytic? No, not not yet. you need to understand model drift and you need to understand training cycles and retraining cycles. You need to understand fine tuning. You understand things like that. Data quality is the product in an AI tool or in an AI focus tool, like data quality. And then... traditional PMs, like they don't understand precision and recall and, like F scores and whatnot. They don't understand those things. They're not brought up on those things in your typical fintech SaaS software. wow, there's so much here. Probabilistic outputs. I mean, there's so much, but I, I just said probabilistic five hundred times. Yeah, that's right. So probably ,istic outputs are a new domain and not a job. I say that, but I'm also reflecting on it, as I say it and say, if you're a traditional, I'm using the words lightly, traditional product manager who was focused on precision, more than accuracy, or accuracy more than precision, either way. you're looking at your data and you're still concerned with the data quality. You're still concerned about that because you wouldn't want to put out ambiguous stuff. what changes here is the data quality changes based on the car you're driving, the model and make. So whatever model you're using, AI model, could sometimes significantly impact the data quality. having the wherewithal to understand that and then having the authority to say, hey, I have a weird feeling. I don't feel quite comfortable under the caller about this. Let me validate that by using a different model. it sets apart the product manager with AI chops. I didn't deliberately say AI product manager, because I still don't think that's the thing. But a product manager that is concerned about how AI impacts the quality of their products. So if you're, if you're looking at this sort of thing, you know, left and right and you're taking that into your stride. I don't, I think you, you're, you're, you're fine. I don't think you can go wrong as long as you, again, to double down and double click what we already said. You're learning these things as you go. So you're learning with your learning velocity, right? Some of the fields already managed non-deterministic behaviors anyway. Sure, healthcare definitely, yeah, right. Finance is another one. pharma, like there, there are quite a few that already handle, probabilistic outcomes like the old bell curve of a you know, X percentage of patients are going to have these outcomes, but there is, there's a field of statistics in this, which is, it's, it's wild to talk about it out loud to the stock market is something that nobody's ever seen before, before AI. Yeah. Yeah. People talk about spreads all the time. If you don't know about that, you really should know about that. You should know about that because the American economy is being held up by gambling right now. Absolutely. You should. Held up at gunpoint. I thought you were going to say, that was the G word I thought you're going to use. They are being held up by gunpoint, by gambling by gunpoint. Anyway, like, no, that can stay in the podcast. There's so much out there that's already non-deterministic. you know, if you don't know about that, two things, right? If you don't know about it, you should definitely go learn that. That's the first thing. second thing is, if you know about it, you're already well in your way. It's becoming a savvy product manager who can handle what AI throws at you. Oh, I thought you're going to say, if you already know about that, you just, keep your mouth shut and, uh, like, happily employable and continue being employable. That too. And you don't make a big fuss about like the AI people on, like understanding precision versus recall versus whatever, model training versus whatever else. those skills can be learned in a week or two. Very quick order, I think you're right. But again, that predicates the fact that you have appetite to learn, right? The curiosity, that you are open to learning new things. You're not simply sitting back on your laurels and say, I know, I'm a product manager, Brian, I can handle anything. Sure. I know, I've handled digital, I've handled Bitcoin, and now I can handle AI. Yeah, yeah, guess what? not so, right? This is a different beast. It has a different number of horns, whatever, however you want to say this, but you just need to have that hunger to say, let me learn about this, how it impacts my product, how it impacts my customer, and based on that, you're willing to switch things up, and that's really what it's about, I think. the real skill is in translating those... Oh, uh, technical constraints into user value, right? Which, if you think about at the core, isn't that what all PMs do anyway? it, it basically is, yes. translating the, domain to technical or otherwise into business value, into business language, right? Yeah, that's basically what PMs do. Thanks for coming the podcast. All right, that's, that's it. Good night. Yeah, yeah, right. Yeah. Don't go anywhere. Yeah, we're not done. I mean, that, that doesn't sell chat, GPD subscriptions or, or make the AI, domain more alluring. So you're not going to find a lot of people with that nuance to take that it was just a tool and it's just adding your tool. to the resume right yeah that's like you're not gonna win friends with that attitude home and i'll thank you to get out of my office oh like i think about any technology that's coming to the stack over the years of you know you want think about this one to be open to the new technologies you want to be open to not only do you want to be open to the new technology you want to be open to um learning about new technologies giving your folks time to adopt them and to figure out how to apply them to what you're doing today and like not a lot of businesses have time built in the schedule to test out new things AI probably because of the amount of just pure amount of marketing dollars spent on it probably has more But, boy, like I, I can't imagine that the, the majority of just regular, you know, insurance companies or fintech companies or whatever have oh, oh, and we're gonna break off 25 % of today's, you know, of this week's sprint to say that you know, Monday, Tuesday and Wednesday, your team is going to be free from working on all their work items to just figure out what the latest, hottest tech is. most companies barely have time to even work on tech debt or, you know, clean up or things that they didn't do two, three sprints ago. I think, I think most companies will probably, land a spot where they will say, look, you know, we hired these people to, to give us training. Oh, by the way. sorry it wasn't in your time zone but we recorded that so go watch the recording i in your own time right on top of your eight hour 10 hour a day yeah so you can imagine the effectiveness that that yields in the end but let's come back to individuals as p ms yes look at this and say what are some of the takeaways out of this particular segment of our podcast i mean takeaway right away for me is going to be if you're a product manager and you you see the company hiring for an aipm and they're hiring 50 % over what your salary is that you can you know what I mean because they put the job description up and you're looking at it you're saying oh my goodness like what is happening here like look at it look at the five AI specific skills that they list like the things that are listed by name in the job description that you don't have in your job description. There's probably at least five of them. They have specific skills. And then look at those things. I'm what, what are they? Model evaluation, data pipelines, data cleaning, bias detection, model evaluation. Data labeling. Look at what they are. Yeah. And then ask yourself, I have, I have some things on the screen here, then ask yourself, can I learn this? you know, in a week of studied focus, like of like dedicated studied focus, you know, two hours every day at the end of your day, at the end of your day, at beginning your day, whatever, how are you working? Is this something that I could learn? I say a week, whatever, two weeks, you know, don't cut you something. I think reasonably two weeks, right? And then compare to compare that learning curve to if you ever had to switch industries to get your product management job, you know, or, if you had to switch teams, you know, or if you were a product manager and you switch teams and you, inherited another product or sometimes product manager split, products, right, take, take up the slack for other people and stuff like that when people leave. is it really any different learning these skills than the last time that you learned a different product or a different discipline in your product? Because if it really, I mean, if it really requires you to go out and run out and get a PhD as opposed to whatever degree you have now. then okay, sure, as a product manager, I bet it's not. I, yeah, I agree. I'm just going to add a little bit of nuance in there as well. So maybe you are just finding yourself in a slightly different domain after perhaps, you know, an illustrious successful stint with a yet different domain, right? So that happens sometimes. Sure. You just say, you know, we've reached this point where you've been great at this. stuff that you do now you want we want you to come in do your magic over here so you learn the new domain so to your point is there any different two weeks i think is a reasonable time you know it's just it's contextual folks it doesn't have to be a week or two weeks three weeks whatever can you reasonably quickly equip yourself with the skill sets what is different about this right just tagging on to the back of your last statement there. What's different from the last time you switched industries or product types, I've added business domains or whatever. The difference is now you have all these different avenues available to you to learn that didn't exist back then, right? So you have online training, you have all these things. You can even just hit up a, hit up AI to give you a nice. manageable for you, syllabus, that you can follow, and also the content that you need to follow in order to reach the outcomes that you need. You need to specify the outcomes and say, how do I get there in a reasonable time frame, two to three weeks, and it's going to spit out everything you need. You didn't have all that before. So you, you're much, much further along and luckier for the technology that we are faced with today. You should have no excuse for not trying this if you're a... product manager that says, hey, I'm looking at this 20, 50 % more that people are offering for this label, you know, what gives, right? You can figure that out for yourself. I'm gonna say you probably can figure it out because, that's probably what everyone else who's applying for those jobs do is that they figure it out and then just throw it on their resume is, I did it at my last company because that's where I figured it out. Not saying people lie on their resume, but, hey. Some people are being economical with the truth, Brian. That's what I'm saying. Economical with their gas mileage. Hey, if, if you have a strong take on this category. Let us know in the comments. Let us know what the hardest part of learning AI products is that regular PMs, they just can't cut. They just can't cut it. Let us know what it is. Let us know if we're wrong. I really am interested to know if I'm way off base here. and also stay tuned because our next debate is going to be on why companies keep inventing these roles to throw people on their, job descriptions on their LinkedIn fast apply or what was it called quick quick apply instant apply what's it called in LinkedIn I can't remember the the don't bother applying because we don't read these anyway anyway so if the skills are learnable and the job descriptions are all made up and fake I was gonna say all the same but then all creating these roles why are they why are companies doing that So manager is basically just a product manager. with a couple things bolted on. Then why are companies posting thousands of these roles on is it ignorance? Is it a strategy? or is it just signaling to investors in the If an AI product board that you are quote, doing AI, you know, like quote, doing agile. I'm gonna keep doing this. Yeah, the rabbit ears. Listen, I, I, I'm a fan of that last one. I'll tell you why. I've seen it before, and I've seen it before that, but way back when the internet bubble was a thing. -oh. A lot of people won't remember that, but especially Jen Zetters. Welcome to school, ladies and gentlemen. Yeah, yeah, Jen Zeters won't remember that. Have a seat. It was a thing when I was driving my imaginary Porsche home every night from the dealership. Different color. different models. when everything was said and done and the dust settled, I had a bunch of papers to put on my fireplace, okay, wasn't worth the thing. Those were my stock options. So the hype machine was a book written by this guy, Sinan Aral, around that time, actually, uh, right after the bubble burst, right after, the internet went belly up around 2020. He explores how hype cycles drive the way organizational behavior and hiring patterns emerge, independent of what the actual needs are. So what's ironic about this is two things. One is he wrote about it 26 years ago. It's still true today. But also, the other thing is, it keeps raising his ugly head every now and then, right? And yeah, we talked about Bitcoin. It happened then. If you didn't know about Bitcoin or, crypto, you're done for as a product manager. Likewise, look, it's not just product managers, right? But. primarily this podcast centers around that role, but. I think we need a counter in the upper right about every time we say blockchain. I think that's, yeah, yeah, yeah. Again, it should be a drinking game. If you don't know blockchain, you're not going to have a career in three years. Exactly. Blockchain. Right. So companies who were doing that at the time wanted to be seen as the leaders and forward thinkers, right? nothing different now. People are still wanting to do that. However, I will wager. my next paycheck, most of these companies actually do not understand what that means themselves. It's just that they're, they're looking at all the salmon, you know, swing upstream, and that's the direction they have to go in. So if you're a product manager, look at a swim upstream or downstream in this, in this way. Our position, at least my position on this is recognize where you are first. This is not the one -time thing that's happening suddenly all of a sudden. It's one of those things that's going to die out, the next thing will come up. So ride the wave, right? But the hype machine is real. So companies need to be seen to be forward thinking their financial statements that they come out with are, forward-looking statements. And then they're always underpinned by the disclaimer, right? Past performance is no guarantee of future results. That's true. right. So again, the whole stock market goes along with this, by the way. Everyone does this from IBM all over down to five people companies. So let me, let me try to summarize. Yeah, yeah, please do. It's a big diet tribe. I was on there. Some of your diet tribe. I enjoy a good diet tribe. I'm usually on the giving end, not the some summarization. Not the receiving end. Not the summary end. So companies need to signal AI competencies. to their investors and the general market even if the general market and their investors are dumb all right so sorry i know that was not the way he would put it but but no i probably would put it that differently so investors and shareholders same thing listen they they they're watching the news they're watching c s i miami they want the language they hear on tv they want to see it in the company reflected of like we're on the cutting edge, we're using all the latest hippest terms. We're probably ballistic, Brian. Yeah, we're six, six seven, six seven. I don't know. in the same breath, as I will say that, I would also say, we need to align our job titles and roles. to the edge of the market to attract the hottest talent, hottest, uh, with a hard age, hottest talent. I don't know what that means. and then, that, that helps him, that helps him, kick off and organize their, industry leading slash AI. That's what I really mean by industry leading initiatives. because they, you know, they got all those people in a bucket over here and they spin them all up and they're all attracted to this shiny object that the company is permanent because that, promoting because that's what they think everyone wants. These, these companies really, the perception is kind of what they want to make reality. So they want to hire people that have quote unquote pedigree. So they're looking, you know, they're looking at hiring people that say, on their LinkedIn profiles, X, Google, X meta, X, X, X, this, X, that. no, seriously, I've seen a trend of that, especially, especially now in the wave of, oh, these people excelled over there. That's not the, that's not the EX that they left with. That's going to be the thumbnail right there. Ooh. Right. So I just look at that and say, they're X meta and X Google probably for a reason. Ex -Amazon. Ex -Amazon. There's lots of people that are X Amazon right now. Sure thing, yeah. So the companies want to be seen as forward thinkers hiring people that are the best of the best of the best in the AI field. And that makes the shareholders and the stakeholders, you know, feel better, right? I mean, these companies are leading. right in this this new wave of working right working in AI so that's really the perception that they want to make reality I'm not have a hard time on this podcast defending the other side, uh, because I think you, you just did it, uh, right there, uh, signaling, uh, AI strategy now or, you know, VP of a strategy now. That's a big way that companies signal that they're headed in this bold new direction without actually, without actually changing a single thing at all or actually having any AI strategy, I know you're going to be surprised though when I say this, but. Companies do this. Companies do this all the time with their chief product officer, where they get a salesperson or an operations person who was a big VP or whatever at the other company. They get them in. They've never worked product in their life. And now they're signaling, the signaling, we do product here. They're signaling AI competency through a job title. Yeah. But that is theater. That's, that's Marty Kagan's product theater. Front row seats. That, that's all that is. specialized titles that fragment your PM org, right job AI product manager, back of the house product manager, strategy product manager, database product manager, All that does, all that does is it introduces silos in your product management, made up one organization silos. for no reason. I mean, it might not be for no reason because of people you hired as the AI PM, they might have been people that, interviewed well for those positions that claim that they work, but they have never actually worked in, I mean, the higher up you go, the worse it becomes, right? I don't need to keep, I don't need to keep beating this dead horse on the podcast because it won't die though. But yeah, I get you. It won't die, but if you invent a job title. and then invite your friends to apply for that job title because, you know, they, they play a mean round of golf or whatever. Yeah, of course it's not gonna die. Right. You can make up whatever job title for your friends at all. Just call them something not product. call them something else. please save my career field and call them something not without product management in the title, without product or manager in the title. Special operations or whatever special projects. Sure. It used to be called, and it used to be called special projects. Yeah. Back in the day, yeah, but those titles have fallen out of favor because rightfully they mean nothing. Right. And everyone saw right through them. So they just retitled product and now products getting degraded because you're an AI product manager. That's completely different. We can't use any of the skills that normal product managers have. We need these, we need this special skill. But it's, it's, you attract better candidates by describing, clearly real problems that you need real skill brought by real people who have real experience rather than a collection of like a roll of d20 worth of buzzwords and throw it into the job description because you're your your your your HR people writing these job descriptions they don't know either way they have no idea you know by the way neither do the applicants that are using it yeah AI tools to apply for these anyway sure so so it's a it's a loss -loss situation right there that thing that you pointed out very early in this podcast is when the wave moves on and all these titles, and all these titles deflate when nobody wants to hire any blockchain people anymore. You did a lot of, you did a lot of rejiggering of your identity and your resume and your experiences to fit into this new mold and, But it would have been better just to just anchor on real problems, I mean, maybe this is not a great point. I think it's a great point. I mean, this is going to keep coming back up like a sign wave every so often. So I think it's a good point. No, I say maybe it's not a great point because the, like the growth people at like about 2016 or the growth people is, oh, I'm a growth product manager. they, they found a segment in the market that you can carve out to be product managers can't do. product manager and then they and then they changed they dropped the product manager because they segmented so well yeah they dropped the product manager and now they're just growth i think there's a parallel there to what's happening now though right i mean now you have AI product managers and they may just drop product managers and say you're now the AI strategist or whatever yeah i i specialist you know the AI wizard they could i kind of hope they would i kind of hope they would because then then they would stand alone right in their niche, with, with just their accomplishments and not hijacking another career field. I kind of wish they would, you know. that's not going to happen. Sorry. boy, I wish that would happen. before you create or if you're a hiring manager before you approve any AI specific. role no matter what that role is i have a couple things that i would caution you to ask okay number one would be it will this person do, will this person do this AI specific thing that the, that the product managers, the existing product managers can't do after their first 30 days of onboarding or after their first 30 days of onboarding will, will they continue to do everything every other product manager here in the organization does. So is it all the rest of this stuff that the PMs do plus a couple little AI things because in that case, you're paying somebody 50 % more to, do things that potentially are learnable. Number one. Number two, are we creating this title because of a real capability gap. Because sometimes that is true. Right. That, that you just don't have the skill at all in your organization. I can see if you, if you're trying to hire an AIPM. whose job is to teach your whole organization how to use AI to their benefit and embed it through the whole whole organization and you know and it's like an internal tools type of PM like that as I like the way I got my start as a geo admin yeah yeah you're like for internal tools and you're like your job is like spent up because that's like what that's one of the places that AI has really um found it's niche in modern businesses, accelerating back office type of repeatable tasks. So if that's your job, then I understand it. And, and you're probably, there's probably a real capability gap in that you, you're not breaking off time for your internal staff to learn AI. So you bring this internal person on. They're kind of a product manager. They're kind of a coach. They kind of accelerate your, your move to adopt new skills. Great. That is great. A lot of people are not just doing that. You're that, you're that, but you're also a normal PM at the same time. And then the last thing I have here is, what's the, what's the roadmap three years down in line for this person's career? one of the wages that most organizations have no idea. Three years is, so what used to be medium term was five years back in the day and long term was 10 years or so. up to three years that was near term, right? Right now, three years is long term. Three years, yeah, and nobody knows anything that's going to happen. The distant future, the distant future, the year 2000. The first point you made, though, about, you know, what will this person, do that our current PM cannot do after, let's say, a month of being on the job, right? That's the question. It's is it real or is it Memorex? So maybe your... product managers might have been fine if you just gave them a little bit of training on the new tools, the new AI, um, uh, capabilities, etc. You just don't know that till you actually try it. So if you're a hiring person and you think, oh, we need to, you know, offer commensurate compensation to attract these new skill sets, just look at your current match and say, can we enable them a little bit first? you know, maybe just, you'll come out the same if you just reward them for it instead, but you will save a lot because onboarding costs are real, the opportunity costs that time spent, etc, that these are all real. So think about that for a bit. Yeah, think about how has your company created an AI version of any existing role, like, whether it's product or developer or whatever else. Right. And then, how's it going? is, is it the same role with extra money? Are they doing a lot more work? Are they producing better results for customers, right? how's it going? Let us know in the comments is what I'm saying, and, we'll wait. Yeah, yeah, but not very long. I mean, you know, hey, I've got all, I can do this all day is what I'm saying. So we've covered why companies create superfluous roles. Now we're going to talk about the person on the other side of the job posting. We're to talk about if you're a PM right now. Should you be worried? or should you look at rebranding yourself. That's right, Ohm, you can call yourself the blockchain. Agile coach. X, X, blockchain. Oh, sorry, X blockchain. So if you're a product manager, um, the pressure. Your you've probably felt LinkedIn feed is full of people adding AI to their headline. LinkedIn influences are selling AI PM certifications for$625. Hashtag, buy my book. That's right. Sign me up, sign me up. Sign me up. You buy my book. And then, so here is, here is our category. Should you play the game and rebrand and call yourself, um, blockchain Patel, or should you hold the line because love isn't always on time. That's, that's a to tommy. They're a lovely song. but yes, there are plenty of them out there. So what should you do? Stay tuned because we're going to get into this right now. Oh, I thought you were going to say, what should you do, be a thought leader. Oh, no. You gotta lead, you gotta lead thoughts, or thought. thoughts now never mind it's it never comes out clean when i say yeah i know you gotta be a leader in the shower oh follow me for more tips uh so um we got a great book inspired mardi kagan uh how to create tech products customers love by marty kagan uh he argues that strong product management domain agnostic skills are rooted in discovery delivery delivery and customer value which is funny he's been writing a lot of blogs, so, um, there's a lot of, like, things tacked on to the end of the book where they're trying to stay at the forefront of this, but, clearly, you know, do, do we need specialist product management? I don't know, I don't know what Marty Kicking's opinion of this would be. that's, that's what the book says, you know, the arguments against you'll hear are, hey, recruiters filtered by keywords home and not by whatever marty kagan says that's the job the free market home free market free market indeed so marty i know you listen let us know in the comments below get in touch because uh you know we want to hear about what your pivot point might be today since the time you wrote this book back in 2018 -ish um yeah so I can understand people saying, hey, look, I got to do what I got to do to, you know, even get to the interview, right? So I'm going to sprinkle liberally that, ju -ju in my, you know, in my recipe. Oh, okay. You have to meet the market where it is. by virtue of branding yourself personally so that you do get tapped on blockchain for that for that interview so that's the against in this category all the blockchain yeah all of it that's right that's right blockchain crypto all of it all of it yeah i mean sorry Yeah, you're right. I mean, those are the against. I mean, chasing titles, like that treadmill never gets old. So you're gonna stay, listen, if there's, if there's something that I'm gonna argue legitimately for on this side, it's gonna be, if you're on the treadmill of chasing titles, it goes on and on forever. You're gonna stay fit, is what I'm saying. Oh, you're gonna say nice and fit, because you're gonna be on that treadmill running for, ever because these these cycles in tech go on and on and on and they don't stop like the ups and down cycles depend on whatever hype is happening on the market so i guess if you're saying i'm just gonna purely chase hype cycle ups and downs and i'll always be in demand and my skills will always be at the cutting edge i don't know like i you kind of got something going for you there If you're one of those rare unicorns, please do get in touch, because we'd love to talk to you, right? I mean, the hype cycle, yeah, you're gonna be fit, to your point. You're gonna be fitter than the fittest hamster, I guarantee it. Because, yeah, that wheel doesn't stop turning. That's right, that's right. And the cash keeps flying out of your pockets. I can envision this, you know, like gold coins, flying out of your pockets, because you're paying for all these certs every so often. Are they, are they? are they flying out of your pockets? I don't know. Is that where they're coming from? Where they're coming from? Sorry. Someone's getting paid. the hype following isn't, it doesn't represent deep expertise. the hype following. Okay. that, that was where, that was where I thought your pushback would be centered on, You're in the pursuit of trying to follow deep expertise, but you don't get it just because you keep getting certs and trainings and whatnot. So that was my point. Oh, boy. Anyway, the smartest move here, I think anyway, is to build AI literacy and to keep your identity rooted in PM fundamentals, PM fundamentals, which by the way, Marty Kagan wrote a whole book on it that we just referenced in the, in the opposite end of this podcast is inspired, how to create tech products that customers love. PPM. fundamentals and they haven't changed since 2018 since he penn the book so in your last statement you said the smartest move is to build AI literacy and keep true to your identity with PM fundamentals there's an and that's not you know comma or or it's and that both sides of that are on an equal footing in my book P.M fundamentals will never change, you know, until something very, very radical comes along, which we can't foresee at the moment. Oh, I was going to say the lean startup, but never mind. The lean, lead startup actually is still rewritten PM fundamentals. I understand. It was that, but, but, you know, building AI literacy, the side of the seesaw that is always like higher than the other side, that, that needs to have equal weight, I think, right? So that both, both sides are level. You'll always be building AI literacy. AI will change to something else. So build X, Y, Z, whatever, literacy. You'll always be doing that. You'll always be, you should at least, at least, you know, you should be hungry enough to learn how these tools enable you to fulfill PM fundamentals. That's the only way, so as I can see. So if you're feeling the pressure to rebrand. as an AIPM. You're out here on the market. Take a yoga glass. No, no, no, no. I mean, you should do that anyway because, yeah, it's great for your pusher and, your well -being and your health, okay. But once you're past that. But then when you get back home, then we introduce AI into, if you're gonna build, if you're gonna build an AI related case study into your portfolio, here's the way that I would recommend that you go about that. Are you ready for this? Riveting advice home? Ship something. Yeah, really. ship something AI enabled. Yes, AI ready. Yeah, so I think there's a transition. So if you're a traditional PM who's never worked with AI before, and you're asking them, hey, guess what, ship something, how? yeah, I'm in for the two week, three week, whatever learning period, but how, right? So one of the things you could do is, have a retrospect on what you've delivered already and figure out how you could have delivered it had it been AI enabled. Yeah. Assuming again, you've gone through your two week, three week period, and you've learned the fundamentals of AI to say, how could I have done it differently? Right. At a minimum, that's going to give you some talking points at your interview, if you're interviewing for these AI PM jobs, but. Also, it will inform how you should be working next on your next product or next, you know, release of the same product, whatever it might be. So think about that. Look back because it's low risk. It's actually no risk because you've already shipped something. So at least have that discussion and be smart about it and have a discussion with your peers as well, right? Because it's a learning opportunity. The other one that I have is to learn of your business domain slash product right yeah well enough that you can at least have a 30 learn the vocabulary minute technical conversation with like a real machine learning engineer or end or data scientists yeah you know where you're informed and not left behind the dust that would be good Yeah, that would be good. Again, look, we're saying 30 minutes, again, depending on your domain, 20 minutes is fine. Because these guys are, are beyond you when it comes to, you know, the technical domain of AI. But you're beyond them when it comes to the impact that technology has on your customers. So if you can talk to that, it's like cogs on a wheel to me like these guys will enable you to deliver better value to your customers so you need to be able to articulate what that value looks like and given the technology and the two of you can or however the two parties can be multiple people can come up together with a solution potentially that might actually get you there and then my last tip here is when you're asked Hey, tell me about your AI PM experience or hey, tell me about, you know, being an AI PM or what stuck out to you. Something along lines of the tell me about yourself kind of job interview question. You need to be prepared with, some kind of response that says, hey, I'm a product manager who has shipped AI powered features or has shipped an AI enabled product and then to be able to talk about the impact of that product, right? and talk about what went into it and, and how it helped and how it impacted the customers and how it changed the customer's viewpoint on things or, you know, help the change their perspective or help the, you know, in their workflow, whatever it is. Yeah, I, so two things. One, yes, I agreed 100%, but also it's not just the customer's viewpoint on things. It's also your vendor companies, enable delivery of how did AI help you that value quicker? faster, better, more efficient, more effective, etc. also, if you haven't done that, so if you've done that, it's fine, slam dunk. But if you haven't done it, look back at your product, have a discussion point in your mind that says, here's how I could have used AI and use that to illustrate that in your answer, right? You know, are you an AIPM? The answer always is yes, and I could have delivered my last product had I. you know, AI tools at my disposal in this way. If you haven't, but if you have, then that's the easy part. You can say, here's what I've done and here's what I could do next. Yeah. So what do you think at the end of this category? have you added AI to all of your job roles on LinkedIn, in the time that we're talking. And be honest, I want to know. Yeah, especially if you've got AI parent on you. No, never mind. I mean, listen, whatever helps you get the job. So if you found this useful, stick around for our next debate. we're going to talk about hiring managers and what they actually should look for. So we've told product managers not to panic rebrand. Right. I think, I think that's what we've said. but what about the other side of the table? So now we're going to talk about If you're a hiring manager for an AI product and your job posting says AI product manager and you've listened to everything that we said and you're oh, I could be in trouble here. You might be filtering out. your best candidates and attracting the worst ones With blockchain like spray painted out and AI product manager spray painted in. Is it spray painted in? Sincoled in? I don't really know what it is. so we're going to go through what to look for instead. So in escaping the bill trap, that's right, that's right, we're quoting Melissa Perry on the podcast. Um, nobody, like, put all your reservations aside. Nobody can besmirch thine Melissa Perry. All right, so I don't know why I was, uh, went Shakespearean all there for a second. in escaping the bill, bill trap, Melissa Perry emphasizes hiring for product thinking and outcome orientation over domain specific labels. That's right. got two heavy hitters so far on the podcast that are saying hey higher for product thinking not anyone specific thing that happens to be flying any paper airplanes that happen to be in the air at the moment that you happen to look up last category that we're going to talk about yeah i want to zip through this one as fast as possible because are going to be, hey, we need someone who already knows since this is the AI. We, we don't have time to ramp up. The, the, the arguments here are the arguments here going to be the most corporate, arguments that I can think of, which is we don't have time to train somebody home. We're willing to pay 50 % over the average market to get an AI PM that already knows what they're doing, even if they don't. anyway that that we that we assume they already know what they're doing So they can come in house and teach all the rest of our people, the best practices of AI. And then we don't have to spend time teaching our people and sending in the class or bringing somebody in house to coach them, et cetera, et cetera, this is typical corporate thinking would be the first against in this category. Number one, before you jump on that one, the other one would be, If I get somebody who's already had experience building AI products at other companies, that reduces the risk of hiring someone. who may, conceptually have these skills or have learned these skills in their spare time or whatever. I'm gonna get somebody who's already been using it at other companies. I already know your pushback's gonna be like, this is a whole chicken and an egg, like, we can't all be like, we only hire chickens here. We don't ever want to hire eggs. We don't want to wait on them to turn into chickens, even though we know. They will. They will. That's exactly right. That was going to be my argument, exactly. I mean, we need someone who already knows AI so that we don't have to train them. Ten years of experience, don't we want to. But, yeah, right, right. When AI has only been around. Generative AI. Yeah. unfortunately, this is rife in our industry, right? We see job advertisements like this all the time. The downside to you as a product manager applying for something like this is it speaks volumes about the company. They don't invest in training. They expect you to do more than just product management because you expect to train all the other people that are there who might not just be limited to product managers, by the way, could be anybody. you're comfortable with? So that's one thing. I know these is that something companies will say, But oh, hiring AI-specific skill sets is a risk reduction strategy, right? At least we know we're getting people with AI skills. Right. What are their skills in exactly? I mean, you got a deep dive here a little bit. You know, instead of just looking at their resume that passes muster with your screening tool, you know what I mean? I do have a good, but like, I do agree with one of these pushbacks. don't know. I, I do agree with one of these pushbacks cynically. That's, that's a good way to say it. but maliciously. I we need someone who can quote, speak the language to our technical staff. Like that, the AI product management. is it's too technical for any one of our generalist PMs, especially at larger companies where the PMs all these people I see on LinkedIn learning who's like, I'm a product manager and they, like, technical understanding at all they are the people that like a year or two ago we watched on TikTok that go to you know they had no lunch half the day and then have an hour and a half long handoff with their technical lead who is their proxy, their go -between to the technical team and they never talk to a team. the technical lead goes and implements everything in the style thereof. That's what they do all day. Yeah. this is not the normal product manager that I am, that I am used to, but there are apparently, apparently TikTok to be a lot of these people that work on very large companies like LinkedIn and Microsoft and other very large companies. And for them, if we're gonna, if I'm gonna actually defend the against points here, for them, the against point here holds very firmly of they have very little technical expertise at all, let alone building AI products, meaning product products that are like leverage in LM and the response back sometimes is not quite correct and then you have to do some of the things and there is technical considerations that a normal product, you wouldn't have to jump through those hoops. Yeah. And they may not understand why all of the business. requests that they are bringing forward cannot be served or have to be served in specific ways, or maybe there's extra UIUX has to go into leading the user to help them understand. Anyway, it, very technical, not something I could put on one of these folks. if I'm trying to prop up, the against points here in a way that it actually defends them, these folks are out there. They work in real product management roles. Yeah, and some of these folks could also fall file of some of the illities that Marty Kagan has defined, because they don't have the technical jobs to your point, right? Absolutely. But what do they bring to the table instead? to the table? What do they bring to the table? is that worth having also? So are you better off not dismissing these people altogether and, and simply tapping them and What do they bring say, look, we need you to learn about these new things, AI related, right? And we'll give you, you know, a month or two. We'll even pay for some of your training. here's how that can help you. help us if you can do that i think you have the best of both oils sure sadly most corporations don't do this yeah yeah i mean I mean, the trade off here is, you'll never know until you give them a chance to learn. even I am, I'm coming down harsh on this class of folks, but, you give them a chance to learn, they will surprise you, you if you just hire for AI keywords, you're going to miss people that legitimately can do the job, want to learn, want to expand their skills, their skill set, that kind of thing. and you're going to miss out. you know yeah yeah people want to expand their skill set and you never know what you're looking for, you know, and also the engineering team, they're experts at engineering. They don't need a PM that, you know, can double check their engineering. that's what they have each other for. Yeah. And they don't need another data scientist. that's but you never know, what they have data scientists for. They need you to be the hook that hooks to the business. And, that's the main thing. Any customers. Yeah, yeah. You're the conduit for them. Yeah. and in that, I, I kind of align with, Melissa Perry and Marty Kagan on that one is hey, sure, you can know a bunch of technical stuff. maybe you even know how data models are built. Great. not all that is a PM scale. Yeah, I agree. I think those are mostly in the technical domain quadrant, if we had a two by two, which I don't for this time, but, yeah. So look at us without our two by two. We don't have a two by two today, folks, but next time we'll make up for it with a three by two or three. Um, take away, the takeaway here, write your job postings around problems to solve. Not around. Not around technologies. Technologies, yes. Yeah, or blockchain. Oh, blah. Yes. so if you are writing a job posting for an AI PM, blink twice if you're being asked to write a job description against your will, we have some suggestions for you, which is replace AI product manager with a product manager for whatever product or problem. or whatever it is in the space that you're trying to solve, put that in there instead of AI, right? Right. And that'll attract PMs to your business domain or your problem domain that you're actually looking for. And then the AI stuff is just like icing on top of the cake at that point. Okay. Number two, in the requirements, you put your three years of AI ML experience, like instead of that, you put demonstrated ability to learn complex technical domains. And, and if you want to sprinkle in AIML a plus or whatever, throw that in there as well. Because the real machine learning folks that are going to be in your career field, they're going to have machine learning. They're going to have ML in their resume because they probably have an updated resume in four years and in the interview, uh, you can give, candidates a real AI product scenario. You can say, hey, our model accuracy has dropped 5%. What should we look at? What should we do? And you can, walk through the scenarios. Now, everyone's using AI tools in the interviews anyway, so good luck with that question. But remember, the devil's in the details. So the more details you dig into and the more you guide them. The bigger the devil. That's exactly right. The bigger the devil. I was going to say that the deeper the, the, the more tokens you burn from their AI tool. Right. That's right. That's right. Oh boy, oh boy, oh boy. so all right. I, I think that was, that was it. Let's land this plane. That's it. if you're a hiring manager, if you're hiring for AI roles, if you're hiring for product management roles, you know, let, let us know, if there's anything here that you found useful. and remember, like him subscribe because every like helps a podcast. So let's, let's wrap the whole podcast. So we, we talked about how to spot the difference between a quote. real AI PM, whatever those are, and a rebranded job description. We talked about knowing which AI skills actually matter versus what product specific skills really matter. We talked we talked about evaluating, AI in job descriptions and what you should put in there and what you should look for. I mean, AI is real. I'm not trying to, I'm not this is not the AI is a fad that will blow over like blockchain podcast, even though I said blockchain 87 times. I, I do believe these tools are here to stay, uh, and the more that you integrate them with, with your normal business processes, the better they will be. But that, again, note, integrated with your normal business processes and tools and products is the way that these tools are accelerating the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, or deployed any other ways or methods, et cetera that helped you that we didn't cover, we'd love to hear from you. Yeah. that is it for this podcast. if you have any other topics that you would like us to cover in the future or if you have strong reactions to this one, let us know in the comments. fire off angry tweets at us. Let us know on YouTube. Let us know on LinkedIn. Let us know either way. Either way, it's good. And, uh, if you want to donate some of your Bitcoin, you know where to do that. Thank you. All of it, that's...