Episode 81

Shayan Fallahi on AI's Role in Modern Hiring and Future Workforce Dynamics

TIM: We are live on the objective hiring show. I'm delighted to welcome Shayan to the show. Welcome so much. Thank you so much for joining us.

SHAYAN: Thanks a lot for the invite. I appreciate it. I'm happy to be here.

TIM: I'm pumped and excited to talk to you and I would love to kick things off by having a discussion around AI, which is probably. Everyone's favorite topic in the world right now.

SHAYAN: Yeah.

TIM: we, we, we got, we got in early with AI and yeah, I'd love to get your thoughts on AI in terms of hiring and have you had an opportunity to dabble with AI in any part of the hiring process yourself? in, I don't know, evaluating candidates or writing job descriptions or something. And then have you observed candidates on their side using it, for example, to generate CVs or maybe even in the interview itself?

SHAYAN: When when it comes to the usage of AI for screening or something like that, I think it has, I have always heard this rumor of like, there is, there has been companies using it or something like that. I personally have not ever used it. I assume that the tool that the companies that I work for have been using for hiring had some, I don't know, maybe tagging situation, text mining in the background of the CVs or something. But I personally have not deal with the, let's say, very technology In the front line. But and then from the other side, like to use AI to produce a CV to produce a cover letter or to my apology or to modify what you have already written. Yes, I have done it. And I think. For sure. It's very effective and it is really impressive and I do recommend it to everyone.

TIM: And so you've used this yourself, you're saying, to modify your CV for a particular

SHAYAN: Yeah. Mm hmm. Yeah. Mm

TIM: And what was effective about it? Like, how was it helpful? Was it in that kind of tailoring to each individual job? Is that, like, how it became more efficient or was there something else that was helpful? Okay,

SHAYAN: Mm hmm. Mm hmm. I personally maybe I have mainly used it to create some type of let's say faster process to reach into the texts that I want to reach. Meaning that. I was, instead of like me trying to go and write like, let's say a very unified type of a structure, every sentence is starting with a verb and then like a bullet point and then max being 20 words long or something like that. Instead of this, I have written 5, 10 bullet points and I was like, Hey, I was working for this company for. Five years and then these 10 bullet points were the things that I was in charge of can you write? Can you expand on these bullet points in a way that I can put it on my cv? I'm literally telling you the prompt that I was writing and then it was writing me like what? What I have to have on my CV, of course, it wasn't perfect and it wasn't polished and it wasn't 100 percent matching with my experience, but then it was, I was already having a good paragraph that I could work on. You know, I was going, of course, over every word and then trying to make it more matching what I actually did or what, what the technology is actually doing. But then, yeah, basically writing that paragraph on myself would have been probably like 20, 20 minutes longer or something like this.

TIM: so it's a time saver and helps kind of distill your thoughts in the kind of best optimal way for that particular

SHAYAN: hmm.

TIM: task. Interesting. What about in terms of using it? So I understand some people would be using it. to almost like mass personalize their CV for each job. I know there's some tools out there that apparently sit as this kind of intermediate layer that would do the applying for you. So it's got like the web scraping kind of element, but then also adjusting the CV to each job. And I've heard some people trying out those kinds of tools. Have you seen something like that or thought of using that approach?

SHAYAN: Not personally. No, I, I haven't heard about it, but I mean, that's, that's a very interesting approach. I even didn't think about it like that, that I don't know, like you, yeah, like over, over. Over customizing your CV to each job advertisement, and I'm not sure, even if from the HR point of view or from the recruitment point of view, it is actually a good thing to do. Is it?

TIM: well, yeah, this is. A really interesting kind of conundrum because if you're like success metric is how many good looking CVS have we managed to attract, oh, we've got like 1000 applicants and wow, like 90 percent of them look like they're suitable for the job. If that's what you're optimizing for, then yes, it's a great thing. But what I fear is being done is that candidates are using these tools to kind of brute force generate lots of CVS. that are optimized for job ad, but that are at least partly bullshit. So then the

SHAYAN: Mm,

TIM: cV is maybe even less representative of the person than it used to be. So that, that is what I'm, that's what I'm kind of piecing together must have happened just based on all the conversations I've had in the last couple of months, especially in the U S just where the volume of applicants is insanely high. And so many hiring managers are telling me that the CVS are starting to kind of look a lot like each other as if they've. All been generated through Chachapiti or some kind of large language model, and that they're all sort of optimized for the job ad. So they're just being inundated with all these great looking CVs. But they've almost got like too many good looking CVs. And so it's seems to have broken the process a little bit.

SHAYAN: Yeah.

TIM: at least that's what I've pieced together anyway.

SHAYAN: Interesting.

TIM: personally, like if I were a candidate right now. I would be stuck in two minds, I think, because I would be probably looking at the number of applicants per role on LinkedIn and going, Oh my God, there's 1000 applicants. Are you kidding me? That now means I'm gonna have to apply to, I don't know, 100 jobs instead of only 10. And so I'd be thinking of this sort of volume gain, but then part of me will be thinking to like, abandon that completely. Like, forget about going to jobs boards. I'd maybe be thinking about my network or. Leveraging warm introductions. Like if you were searching for a job now yourself, have you, yeah. Have you had any thoughts about how you would go about

SHAYAN: Yeah. Maybe, maybe to address the point that you were mentioning in the, like, before that, before your question, I don't know, like, from the point of view of customizing your CV to a wild extent to the job advertisement, at least I have heard from a friend of mine who is doing HR here in Berlin. Is that at least some companies are a little bit my apology. I don't know what's happening with my voice some companies are a little bit allergic to this finding bits, let's say every keyword which was in the job advertisement in your cv because it is It cannot be a quinketing, you know, like you have done it intentionally and then this is kind of like There is this concept in the modeling world there, you're calling it overfitting. I think you're overfitting your CV to the job advertisement. And I think this is the most unnatural way. To be honest, for me personally, if I'm hiring, this is a little bit of a bummer. So yeah. Anyway, that's, that's about that. But then regarding your question, do you mind if you're repeating it again?

TIM: it? yeah. I i was asking about like what your job strategy, what your job search strategy would be if we're in this world where it looks like the online platforms just being inundated with applicants, therefore very hard to get up from the noise. Would you consider. You know, a kind of higher touch direct outreach approach, like leveraging your network, build your network more. If you have any thoughts about that.

SHAYAN: Yeah. The thing about my, me personally, is that at least till now, the, the majority of the jobs that I have applied for are a little bit, let's put it in this way that they are niche enough that anyway, the number of candidate, as well as the number of jobs in the market are just like, so limited. I think it's just like literally 50 people probably in Berlin who are dealing with those five open positions, which are out there. Something like that. Maybe my numbers are off, however, before this one, like before this round that I'm applying when I was the applying for analyst role like the B I conventional. One of the things that I was like doing was making sure that I'm one of the very first applicants who is sending my CV because let's let's put it in this way that if I mean, I'm kind of like a I'm sometimes like thinking a little bit on this 80 20 type of a rule that everybody's talking about. I think if as a HR person, like if I'm putting myself into their shoes, like if you are analyzing 20 percent of the applications that you have received, probably you have already covered 80 percent of the quality that you want to have for the candidate to fill the job. You know what I mean? And then that being said, I think if you're not in that 20%, that means that you need to be very, very good. Like you need to be like a crazy candidate. Like, I don't know, PhD from Stanford on analytics to make it, you know? And then as I'm not one of those guys, I will I'm trying to be at least in the top 20%, like time wise, but I'm arriving to the desk of the HR person. I think that one would be very good. And then. Of course I have personally made a good, let's say network of people here in Berlin. And then of course, like I am reaching out to my friends and be like, Oh, can you write me a recommendation or maybe send a referral or something like that. And that is at least guaranteeing you getting the first interview with that company. And then from then onward, I think it is in your hands to be able to move forward or not.

TIM: Yeah, I mean, if you can get a foot in the door, at least you've gotten yourself an opportunity. And yeah, ask for any more than that. Yeah, if, if I were applying right now for a role, I think I would really try to. leverage my network. But I think this is an important caveat, especially for maybe more junior candidates listening to this is there's some nuance in how you would go about getting a foot in the door. And so when we talk about networking or leveraging a network, it's not bombing 50 people on LinkedIn that you've never met with your generic CV asking for a job interview. Like that's not what we're talking about. It's a lot more nuanced and kind of longer term, isn't it, than that it's more building the relationship through time. And these are people you already know who are then able to potentially help you out. Is that a fair, fair summary?

SHAYAN: yeah. That's actually very good. And then maybe, maybe one more thing. I remember that from the time that I was like fresh off the university at least in big cities like Berlin, there are a lot of like these type of meetups, you know, or like, I don't know, a company which is having this huge summit that you can even attend as a student either with a discount or almost no fee. And then those places are like really good because then, you know, you would, you would go to like the very store of Amazon, or I don't know, any other big player like Microsoft, whoever, and then you have the chance to talk with somebody senior. And then, yeah, technically, there is a possibility of at least asking that person of like, Oh, do you mind if I add you on LinkedIn or like, I don't know, maybe tomorrow we can have another chat. And then of course that's really helpful as well. I, I think I actually got my first job like on that with, with this, like when I was still doing internship.

TIM: Yeah. Face to face in person, always better because you've got a chance to come across well. As opposed to, you know, being one of a thousand emails that's hitting some poor executives inbox in that day. And so you've got a captive audience there while you're chatting to them. So that's a fair shout. I wanted to change tacks a little bit and have a discussion around like the future of data roles. Like given large language models are developing so quickly. You know, how are you thinking about what a, I don't know, a data engineer or data analyst will be doing in two years compared to now, are there any skills that might be redundant? Any new skills they're going to need to have? Yeah, how are you thinking about the next couple of years for those types of roles?

SHAYAN: Yeah. I think to be very honest with you, there is so much unknown still. You know, I am, I am personally not even sure if I am leveraging AI for every. Let's say thing that I'm doing per day. Of course, like there is, there is a point of time that you may, you are maybe over consuming it to the extent that, you know, you're not thinking anymore or something like that. It always remind me these developments are, it always remind me of like the former technologies that I have had the privilege to, let's say, witness across my life. You know, there was a time when I was like seven or eight, one of the very first thing that my mom did to me was she was. Like sit down and memorize everybody phone number because once you are out out of home And you want to get back home and you don't know where you are. You need to call these people This is your papa's office number. This is my office number. This is our home This is your grandma, you know, and then to be very honest with you those numbers I remember but the current phone number of my mama. I don't I don't even recall because you know, like Yeah, there are a thousand places digitally that I have that number recorded that being said Yeah, probably we are gonna get there at some point as well with everything including the data field like that there is yeah Probably there are gonna be some people in the future to be like do I really need to learn the syntax of sql? Or I need to understand what I actually want to do Or do I really want to? Actually learn how what is the difference in between, I don't know, for loop and while in Python or like, I don't know, the AI would tell me like, Hey, if you write it like that, it's going to be more optimized anyway. That being said, I think there are, there are a lot of things changing and I don't know the full scope of this. But let's say that I am certain that certain things will not exist anymore. And then you're already somewhat seeing the somewhat seeing the companies trying to, let's say, safeguard them versus these changes by saying that, oh, like no AI should be used in our interview process as if if they decide. The people are not going to use them. People are not going to use them, or I don't know. This is as if this is a good representation of what actually your daily work would look like once you land the job. So I don't know. It's a, it's a little bit of a complex thing to say, but yeah, for sure. There are going to be changes, big changes.

TIM: I feel like I mean, time will obviously tell, but I feel like we're just in this weird intermediate state Where we're in the throes of the changes being so profound. And that's partly why it's a bit kind of confusing and why the hiring process has kind of broken a bit to your point. There's companies who are saying to candidates, well, you know, you can't use ChatGPT in the hiring process. And then day one, it's like, here you go. Here's your ChatGPT account. Go and code with ChatGPT or Claude or whatever. Like there's a weird, just misalignment where the hiring process doesn't really work anymore. If people just do the coding test with Claude, then they get 90 percent or whatever. Then what's the point of the test. But then, you know, anyway, so I feel like we just need to get to that next. Almost like equilibrium, whenever that is. And then we can say, okay, this is the new world. Now this is roughly how it's working, but maybe that point will never arrive. Maybe the technology change will just keep increasing exponentially and we'll never really get to that comfortable next step. I'm not sure.

SHAYAN: yeah, yeah, yeah, that's, that's very true. I, what, what you were saying, maybe, maybe one thing that I would love to mention in here is that I don't know, I, I, I do understand that in the end of the day, you want to make sure that the person that you are hiring has a high quality, you know, or understand the concept or is not going to put risk on, on the infrastructure of your company. But at the same time, I'm somehow surprised how people are saying that they are very AI driven in their organization, but then at the same time, like they are not really AI driven in the sense that they are not Really allowing you to, I don't know, use the use any AI tool in the interview process, or like they are once, once you land the job, then I don't know, there are restrictions about using it. I do, again, I do understand that there are risks involved, but then maybe this rant, I don't understand fully that how you, how you are. Very much pro it but then you are kind of like against it at the same time It's there's a little bit of a confusion there if I want to be honest and then this very much reminds me Again, like probably somewhere back in the history if you're searching there has been a time That people could multiply three digit numbers to three digit numbers in their brain. And then it was amazing. But then right now, if you're spending any minute of your life doing that, like in your brain, you're literally wasting time. You know, of course, like you might grow like extra muscles in your brain, but You know, I, I think we are, we are in that phase that they are trying to ban calculators, but yeah. Feel, feel free to ban it.

TIM: well, once you started mentioning multiplying three digit numbers in your head, I was having traumatic flashbacks to. Trying to land graduate positions in investment banking and that I can remember specifically bombing out on one interview They used to love just peppering you with maths questions. Like what's 37 multiplied by 95? Like right can't do that anymore. That's for sure And so I remember buying this like mental maths book and like practicing how to do all these ridiculous Calculations in my head which at the time I thought this is pretty dumb But that's what they wanted. Apparently that's how you looked smart in front of an investment banking client was if you could do yeah, two by two multiplications in your head. Hopefully those days are gone for IB and yes, they're leveraging AI to do some of that grunt bullshit work. And yeah. the bankers brains can be used for something more valuable than that, I should think.

SHAYAN: Yeah.

TIM: You, you'd mentioned in passing something interesting before, which was like you know, you don't know at the moment if you're currently leveraging AI for everything you could be in your job, or maybe some people might have gone too far on the other end where you said, well, maybe some people are using it for everything and not using their brain at all. Where do you think you currently are on that spectrum? Like, have you underutilized it? If you overutilize it in some space, what do you think?

SHAYAN: I think I'm in the, I'm in the realm of ma like, I don't know, mi middle. Middle utilize the A. I. Because I'm trying to be cautious, of course. First of all, there are certain skill set that I have right now, and I don't want to lose them if I want to compare it to the. Phone number memorizing situation. I don't want to forget some of the phone numbers. Maybe I'm just holding on to it for no reason. And in two years time from now, you would ask me and I would be like, I actually forgot all those numbers. But then At least right now, I have a feeling that there are some values into knowing the concept by yourself. It is actually even helping you to interact with the AI, like, in a more professional manner. Because, you know, probably we have, we all have the experience of like, you know, you go on chat GPT and then you write this, I don't know, five word question. And then it is just answering it in a total different context that you're like, What, like, what is it like, honestly? And then that being said, it is from that point of view, I'm trying to, let's say, still know what I'm doing, know the concept and then trying to speed up only the process. And then maybe some other areas that I haven't explored really is to, I don't know, think of new things and then asking AI to do that for me. For example I don't know, I, I know that a lot of one of the very first things that we heard from when the chat GPT open AI was launching was this, Oh, copywriter job is gonna die or like affiliate marketing can be optimized like fully, you know, and I don't know, I haven't done my research, whether if ChatGPT actually killed copywriting or not, or if affiliated marketing is something fully optimized right now, but then I assume that there are other business models similar to the mentions that we can optimize, we can speed up, each individual might be able to leverage them for the personal benefit Like some site money hustle or something like that. And I, I never really explored that in that sense for myself, but then, yeah, for all the tech question, instead of a stack overflow, that GPT does a very good job when you want to write a text and then the text is actually not important, you just want to communicate with one of the offices in the German government. You just write in English, the guy translated for you in German, fluent. Leave a goose in the bottom. So it is really nice.

TIM: What about in your personal life? Have you integrated AI much there?

SHAYAN: Not really every now and then for every curiosity that I have, like, I don't know about physics, about the space or something like that. I am every now and then asking some, let's say questions, but that's just more like from the knowledge gaining or like, I don't know, wrestling some concepts in my mind, but no, not in the personal life.

TIM: So you'd use it almost as if you would have used Google as like a research tool.

SHAYAN: Yeah. Mm-hmm

TIM: Yeah, I've been trying to think about how to integrate it more into my day to day life. One thing I'm playing around with at the moment is just getting more. comprehensive data about myself and using it to do a wide analysis. Like imagine, imagine hypothetically, if I had like a smartwatch data, I had like what I'd eaten that day, how I felt each day, like my motions and all these kinds of things, and just dealing with all that disparate data and finding patterns and giving recommendations. And that's, yeah, it's something I'm going to play around with. Soon you know, habit formation and your behavioral change. that's that's what I think it might be helpful for. Certainly a lot cheaper than digging into that with a therapist over years and years and years. So you can get like a, some, some quick wins maybe.

SHAYAN: That's, that's, that's actually very interesting. Yeah. And I'm wondering, like, do, do, are you using chat GPT or there is another AI tool that you are,

TIM: Yeah, mainly just Chattipati directly. There must be so many, well, I'm sure there are literally now thousands of kind of wrapper tools built on top of that with optimized queries for different kinds of optimized prompts, I should say for different kinds of. Problems. So maybe this product actually already exists. Who knows?

SHAYAN: Yeah, actually, one of the, one of the things that I that I'm very much interested about, maybe that's, that's, that can be the thing that I'm going to search in, in chat GPT after this call is that I'm going to go and see. How many A. I. S. are out there? Maybe, like, roughly. And then what are the what are the domain specific ones? Like, like what? How many other domains have been? They have been implementing A. I. On. Because I don't know, like you're right, like in the health topic, there are a lot of a eyes like on all the visuals, like I don't know, graphic ones, like creating a picture, editing it, putting a video together, like love. Love. I don't know. There are so many different domains. I'm interested to learn what is the distribution even in different industries.

TIM: Yeah, that would be cool. And just some digest because it's happening so quickly. There's so many products being launched all the time. Who the hell can keep on top of them? I just want you do all the research. Just tell me like the two that I should care about now. I don't care about the other 50, 000. That, that would be nice. I wonder if Claude or Chachapiti could do that. Maybe

SHAYAN: Yeah.

TIM: And what about. Interacting with AI. Is that something that you would be explicitly thinking about for candidates for your jobs? Would you expect someone to be able to demonstrate that in a hiring process? Is it something you're explicitly evaluating?

SHAYAN: We are not evaluating that, but I think definitely, definitely if, if it's on my hand to decide whether if we should evaluate it or not, I am asking, I am picking that being evaluated because correct me if I'm wrong, I believe that you are, you are the person who can make. More experience statement regarding this one but then when it comes to hiring your interviews should be the representative of what the job actually is looking like you know I cannot interview you for I don't know baking a bread and then. I don't know. Then you, the job, the actual job being like, I don't know, developing a software, you know, like these are two different fields with two different requirements. Maybe there are these generic questions of like, I don't know, identifying if you're smart or not, or identifying if you are having a critical thinking. Mentality or not, you know, which is a separate conversation. But then when it comes to the skill set, I think we should actually go for the very experience of people in their daily life. And I think I don't know. Again, I'm open to learn what other developers or what other people in the tech industry are doing. But in my field and what I'm doing and I know of my colleague doing, we're dealing with a I a lot. We're dealing with Chad GPT a lot. And I personally Had to go and, like, I don't know, learn a little bit about what are the best ways to write a prompt. There are, if I'm not mistaken, there are even these prompt engineering fields recently, or like, prompt engineering books being written recently of like, what are the best ways to say it? Like there is this method. I think it is being called star or something like that, that you would, you would each one of these letters stand for something stars situation, whatever. I don't know. And then you can, you can communicate via those methods like a better prompt to the whatever a tool that you're dealing with. And I think it is going to be very valuable to see how people are using it. Plus, I think this is kind of like the This is similar to any other type of a regulation. You know, when you are, when you are having a completely pro, sorry, anti drinking policy in a country, then you do not have any control over if the drinks are legit or not, you know? And then once you are actually legalizing it and decriminalizing it, then you can actually go and control and be like, Don't give people garbage drinks, like make sure that they are in produced in a as healthy as possible way, you know, and then that being said, this is kind of like the same situation when we are, when we are not allowing the usage of AI, then we, we don't even, we don't even test if whether people are using it in the correct way or not, as an opposed, I would love personally, I would love if I'm hiring someone be like, okay, okay. Show me your chat GPT, share your screen and show me how you are searching for an answer of something. And this has been historically a thing, right? Like searching on Stack Overflow or Googling things has been historically a thing. It's not like people don't know what I'm talking about. And then that being said, I think this is going to be very nice if, if we are moving a little bit more toward that, over time.

TIM: Yeah, I'm sure we'll get there. It's just, it's just again, probably driven by the fact that we're in this weed in between state where companies are sort of pretending like the candidates aren't using it, hoping they're not, don't know whether it's cheating, don't know whether it's a good thing they're using it. So they're sort of doing the burying their head in the sand thing.

SHAYAN: Huh.

TIM: it will be a much more mature response, I think, to say, all right, like, we want you to use AI, you're going to use AI on the job, it's more efficient. Great. You're going to use it. You tell us how using which model have you chosen and why have you written your prompt and why, how did you integrate it? Like you can, you can get down to a much deeper level and conversation around what they do and don't know and how they've approached it. If you just admit. From the get go, they're going to use AI. Yeah. And so I'm sure that's where we'll get to, but maybe just cause it's been developed so quickly. And and also like, you know, the best way to write a prompt, probably if you Google that 18 months ago is you're going to get a different answer now. So even that itself is changing so quickly that maybe even developing questions to evaluate best practices is also just. Out of date so quickly, maybe that's part of the challenge as well Yeah. No, that's, that's really true. I don't know. Like, I think the speed of development had a played a huge role in here, you know, like, From the time that I personally started, like using AI till now is maybe one year, one year and a half. And then it is already like too far too soon, you know? And, Yeah, hopefully we are going to have a little bit more of a. More, more clear processes around it, When it comes to hiring or any other thing. And then I think that's actually very nice because then we can go back to those situations that back in the time, I think we were doing that a little bit more often. There are even all these. Let's say more anecdotally examples of it online of like, I don't know, somebody is coming to a job interview and then as soon as they arrived at the office, the person who needs to be who the interviewer is basically having a stroke, which is, of course, an acting situation. And then the guy had to, Like they would observe what the guy is doing, like, is he. Or she calling the police is, is the person having any first aid experience or how much action they, are they afraid? Are they not, you know, of course this is a little bit traumatic maybe, but then these type of situational interviews, like, I don't know, is, is a standard German practice. I, I've never heard of this one. I was like,

SHAYAN: Yeah, if I'm not mistaken, this is actually from one of the beer companies, like Carthberg or something like that. They have, they have done this interview and then there are even example videos of it online that you

TIM: really, I've never heard of that. I'm going to have to look those ones up. That sounds like, yeah, as you say, traumatic, but also quite funny but quite telling. But that's interesting that they've cracked onto something, which I think is to me, the critical. shortcoming of interviews and why they aren't that predictive, which is that everyone's putting on a show. It's all fake. It's all you're acting in some ways, but if you kind of trick the candidate into thinking this is real life and just observing how they really act, then you've kind of hacked it. It's kind of like giving them. That McKinsey used to call the 40 hour interview, where they basically hire you for a week, you do the actual job, and then they'd see how you went. But this is the, the simpler and more traumatic version.

SHAYAN: Yeah. I mean, again, like this one might, might be a little bit too traumatic, but then there are, of course, other things that you can ask people to do. Like, I don't know, use the chat GPT, but then try to explain me what is going on or like, find that problem in the answer, which you have been receiving instead of, I don't know, we wasting 20 minutes on a video call, okay. For me to actually make my for loop works, then the for loop is written by chat GPT and I just explain it to you and either I know it or not, you know, or all these, I don't know, IQ testing, EQ testing, I think there are, there are a lot of good things that we already know what can be used. As a replacement for all these conventional, I don't know, analyze this data for me or write, like, write a small function which does this and that, you know, because regardless of what we, we like it or not, like these, these type of things are going to be even further automated within the next upcoming years, you know.

TIM: Some people I've spoken to when we've discussed this is like, okay, yeah, we'll an analyst or engineer need to know Python or SQL in two years. Probably not would be my guess, but we shall see. And so then asking them like, what are you looking for instead? And a lot of people mentioned that really for them, it's about finding people who are adaptable to change, because if things are changing so quickly. That, in their eyes, is the most important skill to have. What are your thoughts on that position?

SHAYAN: I think this is actually an amazing question itself. Like it is really inspiring and it is even having something for me to think of even maybe after the call. I think it is very important for sure. For sure. I mean maybe if you're looking at it like a little bit more across the history, there was a time that maybe in our parents or grandparents generation, you would walk up to a store and be like, do you want, do you want somebody to help you? And then the guy's like, yeah, move all those sacks of wheat, you know? And then. You would do that and that would be your job for the day, you know, and then right now we are in this phase of like having five interviews, six interviews, I don't know, for some bigger organization, it goes up to 18, 20, something like that, and then yeah, we are, we are interviewing people as if. What they have already achieved has been the most significant thing while we know that it is not the case like with the speed of development things moving on, like, I don't know, there are so many so many things changing. Even even in the because like there was this, there was this let's say belief at least or prediction that Oh, like tech jobs are not going to be influenced so much. But I mean, look where we are standing right now. Take job are almost the most influenced. And then that being said I think, yeah, adaptability, being flexible, being yeah, finding, finding a path to move on and then learn something new is very important. And yeah, but maybe, maybe that's actually something that I can ask you, what are those new things that we need to learn and we need to do? We don't know, right? Or do you know?

TIM: Unfortunately, I don't have full future seeing capabilities yet. But I, I, I tell you, I'm, I'm kind of on two sides of the fence at the same time, if that makes sense. So part of me thinks, oh, okay, so like imagine you're an analyst and you've got this kind of blend of soft skills and technical skills you need to have. And let's say Python and SQL are technical skills. Let's say, hypothetically, In two years time, nobody will ever write SQL from scratch. That that idea would be crazy. It's like, no, no, no. Prompt the LLM to then write the SQL for you. And it's so accurate. You don't even have to know SQL to interrogate it. Like it's just perfect. Maybe we can get that.

SHAYAN: Mm.

TIM: So then part of me is thinking, okay, well then that means. you know, relatively speaking, the soft skills of an analyst are going to become more important, the human skills, the communication skills, stakeholder management, you know, convincing people to, to believe in the data to make a decision or whatever. So that's one side, but then the other side of me saying, well, hang on, if more and more of the work is done by AI, maybe technical skills in dealing with AI are going to be way more important than the human skills if most of the people you're going to work with in 10 years are not going to be humans. Or you're going to interact more with AI than humans at some point, I guess. Maybe it's in 50 years, not 10, I don't know. So yeah, I just have this, this yin and yang feeling of the job is getting more human or less human.

SHAYAN: Mm.

TIM: don't know where we'll end up.

SHAYAN: I see. I see. Like if I'm, if I'm dealing with you or your assistant in the next 10 years, right?

TIM: Yeah. And, and I would have thought at the moment, if you were If you have a pretty good knowledge of AI and programming skills, like as in you can use an API of Claude or Chachabitty, like surely that is an explosively powerful skill set that if you only had one or the other or none, your productivity in the world is way lower. So I feel like people have written off programming skills too early is what I'm trying to say that still right now, super valuable if you can build. Even just the SMAPI calls into a AI model and combining that with the knowledge of AI. Oh my God, you can build anything then, really. Yeah, but as we say, time will tell, we're not Nostradamus. So I'm sure whatever we're saying is going to look pretty stupid in five years.

SHAYAN: actually, yeah, actually in the like, maybe it was five or six years ago. I was participating in one of these let's say Ted talk kind of situation. And then there was this guy. And then the title of the Ted talk was like humanity in the next 10, 000 years. Right, which is a little bit of a, I don't know, overpromise. Let's say if it's underdelivered or not. Anyway, and then the guy, this is a time that I don't know chat GPT exists. Maybe the open AI is there. I don't know. I assume it is not still even. I don't know when the company was found. Anyway, that's, that's separate. That's a yeah, that doesn't matter. But yeah, basically they were talking about like how it would basically impact humanity. And I believe that the prediction at the time was a little bit more long running because they were talking about 10, 000 years. And I think with the speed of things right now, we are talking about the next 50 years, maybe. But then the guy was introducing a concept to us, which which is being called zero marginal cost. I don't know if you've ever heard of that or not, but then that was very interesting. He was, he was telling us about the fact that, you know, by integrating automation, optimization and artificial intelligence over time, we are going to get into the point of time that. Things are becoming cheaper and cheaper. We are already seeing the trend of it. He was mentioning even few examples of it. Like any phone of us right now which is in everybody's pocket would have been, would cost bill, just a, just a chip, which is a storing information. And like, I don't know, these 20, 256 gigabyte or 512 gigabyte or something like that. Any of these chip would need a billion if, if you're just going like, I don't know, 20 years back or 30 years back to buy like a storage capacity as big. And then that being said, we are getting cheaper and cheaper over time. And then we are going to arrive basically into this society that things and services. Hopefully are so cheap that basically I, they are either zero or close to zero like cost to the extent that maybe the, our, the definition of things are changing. Like maybe we are the entire life is becoming more about learning. Like, this is a little bit like similar to what you were saying about adopting or being flexible. The entire life is becoming more about. Painting it is becoming more about art. It is becoming more about like I don't know, experiencing because over time, like, yeah, I don't know, everything that a human might want, at least. I mean, unfortunately, we are going to have this. Let's say I'm not talking fairy tale only I know that we are having this let's say Weird wired society, which is not distributing the wealth equally around but then still like at least for some people on the top part of the Pyramid. We are going to have this everything available at every time possible. And then that's the moment that probably the structure of our society needs to be, let's say, refined or like we need to think one more time about what is work. What is your daily life should look like? Yeah, probably that these are Okay. These are, yeah, the influences in the future.

TIM: Yeah, a world of infinite abundance. I'm sure one of the great sci fi writers must have covered that kind of theme at some point. I'm not sure which one.

SHAYAN: Yeah.

TIM: well, that sounds better than the opposite prediction of about 50 years ago, which was That you know, we'd have too many people overpopulate the planet and we'd all die, which is where economics got the name, the dismal science, if I'm not mistaken. So yeah, it's a better picture of the future. Can I ask you this Shayan if you could ask our next podcast guest, any question, what question would that be?

SHAYAN: Maybe, maybe we, the talk that we had today was very speculative, you know, like we were talking a lot about the future and how we are seeing that or something. Of course, any prediction that I was making is subjected to this question as well. But then I want to say, like, we have been historically done a lot of prediction and many of them were long, wrong, like Bill Gates coming out and saying that nobody would want more than 64 kilobyte of space on their personal computer ever, or like the, this New York Telegraph guy saying You Who want to speak with somebody else over the telephone when you can send a telegraph or like, I don't know. There are so many of them. Maybe you can Google them online. I'm wondering what your next guests think is the wrong prediction of 2024. Like what we thought. Is going to happen and then it is the wrong one, you know, and yeah, it can be in any direction. Of course, whatever I said can be that wrong thing as well. Yeah.

TIM: That's a great question and a great thought and is similar to something I was thinking about recently. I saw it on some podcast or some show, which is appreciating and realizing that a lot of what we think is true now is complete bullshit. And in 50 years time, when we look back on our thoughts of what we believe to be true, now, we're just going to laugh at ourselves. We know that has to be the case because we look back 20 years ago and 50 years ago and a hundred years ago and notice all these things that were bullshit. And so it's almost like a humbling exercise to really sit down there and think right now, like, what are my hundred core beliefs? All right. I know pretty much. At least half of them are wrong. That is a very challenging idea to sit with And so this is a, a another twist on that same thought experiment is, yeah, what predictions now are we making will be way off ridiculously wrong. That's a great one. I look forward to leveling that to, I guess, have you got an answer to that question? Have you got any thoughts yourself?

SHAYAN: Yeah. No, I think yeah, probably I would, I would go investigate a little bit more what people predicted about AI and then I don't get me wrong. It can be both in the positive or the negative way. As, as you were mentioning somewhere in the middle, like we might, we might overestimate the capabilities of AI and then I don't know, tomorrow might be that day that they the guys say that Actually you guys, this is not working, let's stop it. Yeah. I don't know. Let's see.

TIM: Let's see. Exactly. Who knows? Time will tell. So it's been a fascinating and interesting conversation today. Thank you so much for joining us with all your passion and enthusiasm and ensuring your wisdom with our audience.

SHAYAN: I appreciate it. Thanks a lot.