
Leman Tech Leadership Podcast · 2026-07-02 · 1h 9m
Key moments - from our scoring
Substance score
49 / 100
Five dimensions, 20 points each
Francis Brero, VP AI Strategy at HG Insights, draws on his experience leading distributed teams across Istanbul, Manila, Paris, and the US to frame one of tech leadership's hardest problems: how to evaluate and manage mid-tier performers. Rather than the familiar A-player/C-player binary, Brero introduces a four-tier framework (task monkey, problem solver, system thinker, and rockstar) to diagnose where each team member sits and what that means for organizational health. He argues that the "messy middle" - medium-potential performers - represents a hiring mistake that leaders must address clearly, not through endless second chances but through explicit contracting on expectations. Most provocatively, he contends that with AI and automation rising, "task monkeys" who can only execute defined work without systemic thinking may lack a future in the workforce unless coached into higher-tier thinking. Leaders managing B-players will find his distinction between impact and potential especially useful, along with his warning that delaying difficult personnel decisions corrodes team morale faster than having an obviously poor performer.
Brero defines four tiers: task monkeys (execute defined tasks without thinking), problem solvers (given a problem, develop a solution), system thinkers (diagnose root causes and solve systemically), and rockstars (unprompted identify and solve systemic issues). The composition of your team across these levels determines whether you'll spend your time micromanaging or empowering.
Average performers create a 50/50 chance of success, which keeps hope alive but erodes trust over time. Leaders give them repeated chances to improve, creating ambiguity. Bad performers are obvious to everyone - no discussion, quick decision. The uncertainty around average performers breaks team mechanics and morale more than the clarity of a clear exit.
Brero argues that with AI and automation rising, task monkeys - workers who only execute defined work without thinking systemically - may not have a future in the workforce. If someone cannot think about the system, break down complex problems, and delegate, they cannot scale their impact or create systems that AI could eventually replace.
Radical Candor is the most misused book in Silicon Valley. Leaders invoke it to bash people openly while claiming good intent, but they forget the core component: you must genuinely care about the person receiving feedback. Without care, it's just public criticism.
Brero says that's a leader problem, not the employee's problem - 100% on you as a manager. But for medium-potential performers (likely a hiring mistake), you must be very clear on expectations and help them understand whether they can grow into system thinkers through coaching, or if there is no future fit.
Our reviewer’s read on each dimension, with quotes from the episode.
The guest offers a handful of genuinely useful frameworks - the task monkey/problem-solver/system-thinker/rockstar pyramid, the insight that managing on output is obsolete in the AI era and must shift to process, and the counterintuitive interview-score heuristic. However, long host monologues and repetitive elaboration significantly dilute density across the 69 minutes.
I think it's too easy to fake output. We used to manage based on output. Today we have to manage on process.
the absolute top players generally have like thirty percent of four ratings, but they'll also get a one and potentially or like generally a two or a few twos and maybe a one, And that's actually not disqualifying. It's often like people that feel threatened more than anything else.
The task monkey pyramid and the claim that average performers are more damaging than clearly bad ones are fresh framings with real practitioner texture. The even-numbered interview scale insight is a small but concrete original heuristic. However, the episode leans on familiar concepts (radical candor misuse, Netflix talent density) and the AI displacement narrative is well-trodden.
how an okay player actually has a very very much more negative effect than a bad player, because a bad player, it's obvious you leave them on the bench
the easiest way to never mess up is to not do anything right. So maybe we need to start changing the way we think about management.
Francis Brero is a genuine operator - co-founded Kudu, managed distributed engineering teams across five regions, named plaintiff in a US immigration lawsuit, and currently heads AI strategy at HG Insights. He speaks from real operational experience rather than theory, though he is not a widely recognised name and his current role title somewhat overstates the practitioner depth on display.
Also the founder of a of a company. During COVID Black Lives Matters, we had immigration policy issues with people stranded in different countries. We actually were the name plaintiff in a lawsuit against the immigration in the US, which we won.
I've led teams of developers in Istanbul, in Manila, I had when we founded Kudu, our engineering team was in Paris. I had teams in the US, in Latin America
There are several concrete anchors - Claude Opus 4.5 as a named inflection point, vast.ai for GPU rental at ~$10 for model distillation, and the specific 1-4 interview scale with the 30% four-rating threshold. These are useful and specific. However, most leadership claims go without supporting data, team sizes, timelines, or outcome metrics.
Something fundamentally changed with Opus four point five in November last year.
vast dot ai. They're basically a true row for GPU...for like ten bucks. You can distill like a pretty like big Mini Max or Quen model.
The host's questions are frequently incoherent, run for several minutes, and often contain the answer embedded within them, leaving little room for the guest to surprise. There is no meaningful pushback or challenge to any of the guest's claims. The host repeatedly inserts his own experiences and upcoming book in ways that derail rather than deepen the conversation.
So I'm super curious on your of house on this one because you mentioned that them, the back players are obvioused, the A players are abused, but the worst worrstponding quote. Of course, the most problematic, let's say, is the middle
how we as leaders, ourselves included, we can teach ourselves. We can learn the courage to be bold to you know, just to sometimes quit the logical thinking critical might be not critical thinking, but the all new using logical approach, analytical approach and analytical part of yourself
Computed from the transcript - who did the talking, and the words that came up most.
▶︎ #174 | A-Players, Task Monkeys, and the AI Reckoning w/ Francis Brero (VP AI Strategy @HG Insights) In this episode of the Leman Tech Leadership Podcast, Alex welcomes Francis Brero: VP of AI Strategy at HG Insights and Co-Founder of MadKudu, a French-born engineer who swapped the grandes ecoles of Paris for Silicon Valley and has spent 15 years building AI-powered systems for some of the world’s leading B2B revenue organizations. Francis brings a mathematician’s rigour and a founder’s scar tissue to every conversation about teams - and this one is no exception. What drives this episode is Francis’s deceptively simple framework for mapping any team: the task monkey who executes, the problem-solver who adapts, the system thinker who diagnoses root causes, and the rock star who spots the problem before anyone else named it. The conversation moves fast from there - to the hidden cost of keeping “okay” players on the roster, why brilliant jerks are rarely as indispensable as they think, and why the bar for what counts as a rock star has shifted dramatically with AI.
Transcribed and scored by The B2B Podcast Index.
1 - >
Speaker 1: Welcome to the Lemon Tech Leadership Podcast. My name is 2 - > Alexander Leaminskam. I'm a former violinist turned management engineer, turned 3 - > organizational psychologist, speaker and PCM leadership mentor and facilitator. In 4 - > this space, everyone is invited to the table where we 5 - > have real, eye opening conversations about tech leadership. Two keywords 6 - > utility and real life implementation. Knowledge tools, frameworks, ways of working, 7 - > dos and don'ts taken from my personal experiences as a 8 - > leader in the experiences of others is what you are 9 - > going to get from every single episode. This is the 10 - > Lemon Tech Leadership Podcast. Okay, so let's go and take 11 - > some action. 12 - >
Speaker 2: Okay and real life. Thank you for being here with us. 13 - > In another episode of the Lemon Tech Leadership Podcast, today, 14 - > I have a very special guests, another amazing guy from 15 - > Silicon Valley. So we are going to see how are 16 - > the differences, how the differences will look like between different 17 - > guests from Silicon Valley. Francis burrowers with me today. Hi, Francis, 18 - > thank you for journey. 19 - >
Speaker 3: Yeah, thank you for having me. I'm excited to offer 20 - > the European who moved to Silicon Valley perspective. 21 - >
Speaker 2: Yeah, Yeah, definitely, there's always super interesting to see how 22 - > the stories look like. So what's possible and what kind 23 - > of peoples you can actually do in your life and 24 - > courage to see what is in there and just push 25 - > yourself and to be the best version of yourselves yourself actually, 26 - > so let's see how your story looks like. So, Francis, 27 - > we always start the conversation with the question about your story. 28 - > So as just started to talk about about your story, 29 - > if you can just go back to the very beginning, 30 - > what happens at the very beginning of your career journey, 31 - > what held on over the way and for are you now? 32 - >
Speaker 3: Yeah, so I think, I mean the journey has been 33 - > It's never a straight liner when you look at it 34 - > in retrospect. You can always like tie the dots together. 35 - > But the funny thing is like, at least in my mind, 36 - > the way I think about leadership, it started even before 37 - > my work career. So back in college in particular, I 38 - > was the captain of our handball team, very famous European sport, 39 - > not very well known in the US, And it was 40 - > a very interesting role because you had to represent the 41 - > team in from the coach, the coach in from the 42 - > team and really try and understand what was going on 43 - > on the field, like what were things that were happening 44 - > with the team, and like what we felt comfortable whether 45 - > or not to help like form the strategy with the 46 - > coach and make sure that the coach was bringing the 47 - > right strategy to the game and not trying to, you know, 48 - > force something onto the team. And so you're kind of 49 - > a mediator of some sorts. And then like through my 50 - > professional career, what's been very interesting is I've led teams 51 - > of developers in Istanbul, in Manila, I had when we 52 - > founded Kudu, our engineering team was in Paris. I had 53 - > teams in the US, in Latin America, and it's been 54 - > a humbling experience and it's just been incredible to see 55 - > how a lot of things that we take for granted, 56 - > especially on the cultural differences, can come to bite you. 57 - > One of my favorite examples was again this like in college, 58 - > but we had this group of students and we had 59 - > a student from Tokyo Tech, so Japanese student in the 60 - > winter in Paris not necessarily very warm, and so we 61 - > had a runny nose and one of the other French 62 - > students like kept on giving in tissues without realizing that 63 - > blowing your nose in public is actually like really bad manners. 64 - > In Japan, sniffing is actually the right thing to do, 65 - > but in Europe that's actually really bad manners to just 66 - > be sniffing in public. And it was really interesting to 67 - > see that, like both of them were kind of offended 68 - > by both gestures that were happening at the same time, 69 - > and it took someone to kind of like, you know, 70 - > ask them what was happening to realize, oh, actually, for 71 - > this person, they think they're doing the right thing by sniffing. 72 - > And the story around this, or part of the learning 73 - > around this, is there's a book that's been misused, Like 74 - > the most misused book I think in leadership in Silicon 75 - > Valley is radical Candor, because we tend to use radical 76 - > candor just to bash people openly and say, oh, like 77 - > I'm bashing you, but I promise it's radical candor. The 78 - > core component that people tend to forget about radical candor 79 - > is the part that you have to care about the 80 - > person you're giving feedback to. And so what was really 81 - > interesting here is by caring about both people, I was 82 - > kind of able to understand what was going on and 83 - > to you know, then give the feedback that hey, you know, 84 - > in the future, maybe don't be so obnoxious about like 85 - > giving someone a tissue. Don't assume that your cultural norms 86 - > are the universal cultural norms. Very Western thing to do. 87 - > And so anyway, a long story short is I've I've 88 - > had a like a like a pretty wide variety of 89 - > managerial and leadership experiences. Also the founder of a of 90 - > a company. During COVID Black Lives Matters, we had immigration 91 - > policy issues with people stranded in different countries. We actually 92 - > were the name plaintiff in a lawsuit against the immigration 93 - > in the US, which we won. So we've been through 94 - > our fair share of all the challenges that that you 95 - > can go through, from training first time managers, trying to 96 - > get longtime managers to adapt to the world of AI, 97 - > and everything in between. 98 - >
Speaker 2: You know, I I would love to dig daper in 99 - > through the handful team first, because this is this is 100 - > not a part of and you said that this is 101 - > not very well in the US, but the US is 102 - > going to have like Olympic team from Hample from one US. 103 - >
Speaker 3: And so I know some of the players, some of 104 - > the players amazing small world wise in the US, it's 105 - > a very very small world. 106 - >
Speaker 2: Yes, yes, that's true. So because you said that you've 107 - > learned so many things when you you had this part 108 - > of your of your story, what what do you actually 109 - > use right now? Because you have many years of experience 110 - > in leading different as you mentioned, different teams, different organizations 111 - > in tech, So what is something that you use from 112 - > this part of your life. 113 - >
Speaker 3: Still, I think there's a couple of things. One of 114 - > them is I think the concept of talent density is 115 - > a concert that was made famous by It was a 116 - > famous Netflix paper on this. But I think top players 117 - > want to play with top players. We were mentioning handball 118 - > and like I've I've been fortunate enough to play with 119 - > like you know, some like elite players in my days, 120 - > and the elite players want to play with the lead 121 - > players and any kind of average player on the team. 122 - > It's more than just their impact. It drags down the 123 - > entire morale the entire team. But they are situations where 124 - > you can help fix that, right, Like, you can have 125 - > high potential fairly minimal impact, but if you have medium 126 - > to low potential, then there is a hard decision to 127 - > be made about retaining or not that that person on 128 - > the team. And the longer you delay that decision, the 129 - > harder and the worse the impact is going to be 130 - > on the team. That's probably one of my you know, 131 - > biggest learnings is how an okay player actually has a 132 - > very very much more negative effect than a bad player, 133 - > because a bad player, it's obvious you leave them on 134 - > the bench. The problem is the average player where they 135 - > could it could work and it could not work right, 136 - > So you kind of want to give them the ability 137 - > to take a shot to like you know, like do something, 138 - > and then it's always like a fifty to fifty percent chance. 139 - > But like slowly over time people start worrying and they're 140 - > they kind of lose trusted. It breaks down, like the 141 - > whole concept of trust within the team, and like the 142 - > whole mechanics don't work as well. Wors again, bad player. 143 - > It's obvious to everyone, there's no discussion. We move on. 144 - > So I think that's one that that I reuse consistently. 145 - > And then just the fact that you know, like a 146 - > good team sport, everybody shows up, you show up to practice, 147 - > you show up on time, We start practice all together 148 - > at the same time, and we do the same routines, right, 149 - > just pass the ball for hours, right, so that it 150 - > becomes like mechanical, Like there are things about the game 151 - > that you repeat, like your plays, Like you know, you 152 - > don't do plays because you know they're going to score 153 - > a goal or they're going to help you do them, 154 - > because you want the mechanics to be so deeply ingrained 155 - > in your brain that they don't occupy any mental bandwidth 156 - > and you can focus on the actual challenge, which is 157 - > reading the situation, and that's where your brain power needs 158 - > to be. Like some of those things I think are 159 - > are one of the reasons why I like to hire 160 - > top level athletes because they have that sense. They know 161 - > how to practice hard, and they understand this concept of 162 - > getting things to become mechanical and consistency. 163 - >
Speaker 2: Right, because this is all about the reps. If you're 164 - > under your reps, this is as as you mentioned, this 165 - > is non mechanical. This is you need to focus all 166 - > the time of doing something and it's it's people just 167 - > quit it because this is too too much, it's too 168 - > hard for the brain and don't want to be in 169 - > that kind of a situation, and you just you just 170 - > don't want it to be gone. And you know, people 171 - > just not in position to do it all the time. 172 - > They think that this is Uh, this is what's happening 173 - > right now, because I see it and I have a 174 - > lot of conversations about it lately, especially with some leaders. 175 - > So I'm super curious on your of house on this 176 - > one because you mentioned that them, the back players are obvioused, 177 - > the A players are abused, but the worst worrstponding quote. 178 - > Of course, the most problematic, let's say, is the middle, 179 - > the messy middle, right, So the B players, because they 180 - > have potential, we recruited them, you know, in some context, 181 - > and sometimes they deliver, sometimes they're not. So we're like, 182 - > in this limbo of decision making, we don't know what 183 - > to do with them. Uh, And I have so many 184 - > questions from leaders that are saying to me, Alex, what should. 185 - >
Speaker 3: I do with them? 186 - >
Speaker 2: How long should I give second chances or third chances 187 - > and so on? So I'm you know, we always talk 188 - > about contracting and what we creed both of the sides 189 - > agreed to do. From from the transactional perspective, it doesn't 190 - > work with every single culture, as you mentioned culture differences. 191 - > But let's say this is quite fair that we have 192 - > some agreements or what you should deliver and now what 193 - > should you do for of how you should tre have 194 - > to change and what kind of outcome you should you 195 - > should deliver in your in your job, and what kind 196 - > of growth Father, we have what we have for you 197 - > create together. But sometimes people are okay ish, So you cannot, 198 - > you know, say that they're bad. But they're just like 199 - > nine to five, I'm doing my job sometimes not perfectly 200 - > on time, but they're okay, and uh it is it 201 - > is exhausting sometimes, and so what's you know, what are 202 - > your faults. 203 - >
Speaker 3: On this one? I think there are a couple of 204 - > things that I want to call out. One of them 205 - > is I try as much as possible to separate the 206 - > impact from the potential, because I think a someone that 207 - > has great potential and has a medium like okay ish impact, 208 - > that's your problem as a leader, Like that's one hundred 209 - > percent on you. However, then the people that are medium 210 - > potential generally that means that probably was a hiring mistake. 211 - > But once it's happened, it's happened, right. That's a story 212 - > for maybe later on on how to avoid those. But 213 - > the biggest thing I've found is being very clear on 214 - > what is expected of them. I there's this concept that 215 - > I've been using lately. I find with AI the bar 216 - > is getting higher, where usually it's just essentially think about 217 - > people through different layers of the pyramid. Right, So at 218 - > the bottom you have kind of your your doers. Right, 219 - > you give them a task, it's well defined, they'll do 220 - > the task. Seventy percent of the time, they do it right. 221 - >
Speaker 2: Right. 222 - >
Speaker 3: This one is kind of the I call it the 223 - > task monkey. It's a horrible name, but it's like you're 224 - > just like you're you're just like, you just do the 225 - > task right and you don't really think about it. Then 226 - > above that you have people that you you know, you 227 - > give them the problem and and they might come up 228 - > with the solution. And then above that you have the 229 - > people where you know they're system thinkers. They see the problem, 230 - > they understand the systemic source of it, and they're going 231 - > to find a solution not just to the problem symptom, 232 - > but rather to the root cause of it. And then 233 - > at the top of that you have the absolute rock stars, 234 - > the people you aspire to have on your team, who 235 - > just unprompted will come to you and say, by the way, 236 - > I've diagnosed this in our system, and here's a systemic 237 - > solution I want to propose for it. And so The 238 - > question then is if you use these kind of scales 239 - > and you start mapping your team against it, the question 240 - > is like what percentage of your team is in the 241 - > rockstar bucket the system thinker or kind of the task monkey. 242 - > And the thing is if you have if you're too 243 - > heavy today on the task monkey side, you're just going 244 - > to spend your life micromanaging and like guiding things here 245 - > and there. It's exhausting and your rock stars are going 246 - > to leave immediately. So again like the trend that there's 247 - > a status of like where you are to given point 248 - > in time. But someone who today starts as a task monkey, 249 - > but it is showing the potential to be a system thinker. 250 - > That's someone you want to invest in. That The challenge 251 - > I find today and this is it's a bit unfortunate 252 - > right in our workforce is you have people who spent 253 - > their career just doing the thing with AI and automation. 254 - > There is a question like what is their place in 255 - > a company. But if you're just literally just doing the 256 - > thing and you're not thinking about the system, that means 257 - > you're not able to take the thing you're doing, you know, 258 - > create a system out of it and delegate it to 259 - > someone else there potentially isn't a future for you in 260 - > the in the workforce, And the sooner you come to 261 - > that realization, and the sooner you help the person understand 262 - > this and you start coaching them on how to be 263 - > system thinkers, how to divide and conquer problems, how to 264 - > think about like breaking down you know, complex problems into 265 - > smaller task that then you can address one by one 266 - > and sorry thinking about it more systemically that if you 267 - > want to invest, if you can rather invest that time, 268 - > that's what I would recommend. But otherwise, unfortunately, they're just 269 - > going to be it's a it's a debt and attacks 270 - > that is, yeah, that you have to pay on the team. 271 - >
Speaker 4: And uh, you know, I'm very much a fan of 272 - > of what actually Kam Scott is saying what you have 273 - > mentioned pradical candor so use it that you cannot have 274 - > only eight players in your team because there are people 275 - > need it from maintenance as well. 276 - >
Speaker 2: So maybe that's the monkeys because we need for them 277 - > to be a little bit different because as you mentioned, 278 - > the changes in in a structure of a team. This 279 - > is something that I've talked about in the previous episodes 280 - > of the Guests Staff Engineering, interest and we were talking 281 - > about how expectations about certain levels and teams changed over 282 - > the last three years or so. About these years and 283 - > then particularly they moved even faster geometrically, I would say, 284 - > so them were we were talking about how the all 285 - > juniors right now are new meats because the expectations changed, 286 - > and then the expectation level of a leader of the 287 - > team quite quite for changed because of how the infrastructure 288 - > is changing and what kind of things they can actually 289 - > do with are all different, a different, different tech stack. 290 - > So so this is connected with the other thing I 291 - > want to ask ask you, because we talked about the 292 - > middle and what about the situation when you have in 293 - > a team one or more people who are brilliant jerks. 294 - > They're amazing eight players, theylieve an amazing job. Sometimes they're 295 - > very very expert, have an expertise so narrow that in 296 - > one one piece of tech for example. So they work, 297 - > they they have their kps s right, everything is great, 298 - > But they are jerks to others, so they comment, what 299 - > is organizational organization is doing, what kind of decisions, what 300 - > kind of pavots are happening in organizations. They have a 301 - > toxic influence on others, but they delivered the job, so 302 - > the question is what to do with them what they think. 303 - >
Speaker 3: In my experience, you try to try to get them 304 - > to realize that their role in the team is beyond 305 - > what they delivered directly, the same way that a star 306 - > player in a team is meant to be an inspiration 307 - > for the rest of the team. So, I mean, for 308 - > those who've watched ted Lasso, I think a lot of 309 - > the show is about this, right where ted Lasso is 310 - > trying to coach one of the top like this like 311 - > young striker. Like then that's a very typical thing in football, right, 312 - > Like the striker is really kind of super cocky and 313 - > and he's trying to coach him to understand that his 314 - > behavior is what then others are going to mimic, and 315 - > his role is not just to strike, it's also to 316 - > drive to lead the team. And that's why I think 317 - > a concept that most people don't realize, and that's what 318 - > I love about the engineering has this dual track of 319 - > you can be an ic or you can be a 320 - > manager all the way to you know, the top levels, 321 - > and some people are just excellent leaders without being managers. 322 - > I think getting the person to realize that you're at 323 - > this point if you do the job really well, but 324 - > you're a jerk and you're you know, not necessarily helping 325 - > to improve the system. You're just an excellent task monkey. 326 - > You're not a rock star. And you might think you are, 327 - > but you're not because you're not improving the system. You're 328 - > actually destroying the system by demotivating the rest of the team, 329 - > by not helping them, by not putting in place solutions 330 - > that would remove your frustration. Right, because if you see 331 - > something broken, the right system thinker is going to be like, oh, 332 - > I can fix this, rather than just pointed to say, oh, 333 - > this is broken, I'm not happy about it. And and 334 - > so that's again the role of the manager is and 335 - > and the leader is really to help this person understand 336 - > that their role goes beyond beyond that. And I've had 337 - > players like this and in my teams in the past, 338 - > like I've had a like a former national Boast Bosnian 339 - > player and he was excellent, like best player in the 340 - > US for for a couple of years, but horrible attitude, 341 - > and we never managed to really perform at the elite level. 342 - > And once he was gone, funnily, like the you know, 343 - > the level of the team was a bit lower but 344 - > we performed a lot better because then we were all 345 - > pushing each other and there was like something positive. And 346 - > so you try and some people actually transform. I've had 347 - > players the same thing that transformed and became incredible leaders 348 - > within the team, and it's it's magical to see them. 349 - > They can come and save the day and at the 350 - > same time they're going to, like, you know, pick up 351 - > the player that's struggling, and it means a lot to 352 - > them when they show up. Their impact is just it's 353 - > way beyond just what they can deliver themselves. So I 354 - > think you try to get them to understand what leadership means, 355 - > and if that doesn't work, you have to cut your losses. 356 - > I think this is another one of the hard decisions 357 - > to make, is letting go of someone who's excellent at 358 - > what they do, but it's detrimental to the team. It 359 - > seems it seems hard, but like that's why we do 360 - > cultural fit interviews, and that's why culture fit and you know, 361 - > impact on culture is actually an important metric to evaluate 362 - > people on. 363 - >
Speaker 4: Do you know that? 364 - >
Speaker 2: For example in Poland and you cannot say that you 365 - > can't you know, you're not offering a job to a 366 - > candidate record because they are not a culture fit. It 367 - > is actually league. So of course we are we are 368 - > talking about recruitment and onboarding check right now than the criteria. 369 - > But this is this is important from the perspective of 370 - > of of the audience because a lot of people are 371 - > listening goal around the globe and then the bigle. 372 - >
Speaker 3: Stuff is also ing and I think often the problem 373 - > is like the cultural fit is kind of this this 374 - > like wishy washy kind of thing of like for a 375 - > long time there was this test of like would I 376 - > want to go have a beer with this person? I 377 - > think that's the absolutely that's like the kind of cultural 378 - > fit that should be illegal. It doesn't mean anything, right 379 - > because then you're just hiring your buddies from college. But 380 - > I do think there's something about the like how you 381 - > fit in the team, like the kind of the EQ. Right, 382 - > So we can measure EQ. We can measure how do 383 - > you react to feedback? How do you react to being 384 - > challenged on something by someone who isn't your peer and 385 - > it is potentially considered junior. How do you learn? 386 - >
Speaker 2: Right? 387 - >
Speaker 3: Like those kind of questions I think are are questions 388 - > that measure more the EQ. And when we say cultural fit, 389 - > essentially what we're describing is like, what is the kind 390 - > of EQ that we're looking for in addition to the 391 - > type of IQ, and. 392 - >
Speaker 2: We can actually measure the communication intelligence. So then another 393 - > part of the of the story and this is what 394 - > I what I actually teach in't and well this is 395 - > my expertise area. And what I what I say to you, 396 - > to the to the leader is like, hey, you cannot 397 - > do the culture fits per se, but this is something 398 - > that the person needs to marge the team. They cannot 399 - > be copy and piece because this is this is actually 400 - > you mentioned you are hiring your bodies or you are 401 - > actually hiring you all the time because it's easy, we 402 - > have a flow in the conversation and we are talking. 403 - > We can talk for hours. Yeah, of course, because this 404 - > is the same person as you from the structural perspective. 405 - > And then yeah, surprise, surprise, you say, Oh, then novative, 406 - > they're not engaged as I would love for them to be. 407 - > It's troubles reprised because you manage the the recruitment in 408 - > a certain in a certain way. Yeah. And coming back 409 - > to to the to the brilliant, brilliant jerk, this is 410 - > the situation. This is something that I Honestly, I have 411 - > this conversation with different theaters repetitively, and this is happening 412 - > all the time. And as you mentioned that there are 413 - > different scenarios, I would say that this is. You know, 414 - > every single scenario has a conversation per beginning. So you 415 - > need to have this conversation with the sampoint because this 416 - > is part of your team and they're destructed to your team. 417 - > So they either understand and change. As you mentioned the 418 - > examples that you have, actually that they change. The rock 419 - > stares writing right now and they understood because never nobody 420 - > never said to them before, and this is something that 421 - > we assume that they should know, but we know how 422 - > assuming is ending most of the time. And then the 423 - > second situation is like they the here but they don't change. 424 - > And the third situations that they quit. And I think 425 - > that a lot of rrors are scared that they're a 426 - > players quote unquote because they delivered the results, they're going 427 - > to quit when we have this conversation or this is 428 - > you mentioned that this is not intuitive, but I think 429 - > it doesn't make any sense to fire them right because 430 - > it's like, hey, they delivered the job, but long term, 431 - > the cost is so much higher than the short term delivery. 432 - >
Speaker 3: So now, yeah, and that's ah, I find that'say. Just 433 - > in general, like, the one of the biggest challenges with 434 - > hiring someone or you know, putting them on a PIP 435 - > is this idea that even people who aren't delivering much 436 - > are doing something right. So there's always the fear, Oh, 437 - > but if I get rid of this person, somebody like 438 - > the rest of the orc has to absorb the work 439 - > and you cannot afford it, right right, exactly, We're already overwhelmed, 440 - > so like, how are we going to deal with this? 441 - > And again, the of your system thinkers and your rock 442 - > stars is pretty frequently what they'll realize is a lot 443 - > of the work we're doing doesn't need to be done, 444 - > and they'll come in and apply a couple fixes here 445 - > and there that just like then reduce the overall workload 446 - > for everyone. But you were never you never really looked 447 - > into how we were doing things because you kind of 448 - > had the people that the task monkeys doing this, and 449 - > and that's where you start creating. You know, you just 450 - > have like a big and fat organization right that just 451 - > doesn't move as quickly. There's just a lot of inertia 452 - > because we're just like this is how we do things 453 - > right instead of thinking about how do we fix the system? 454 - >
Speaker 2: That's true switching gears a little bit, you said that 455 - > I'm with you on this one. Eight players just to 456 - > just want to work and play with out our eight 457 - > players because this is fancy and we want to be 458 - > an a team, because this is well, this is the 459 - > best one that we can be. What we as leaders 460 - > right now we are recording this episode at the very 461 - > beginning of June twenty twenty six, right now, in this 462 - > world that we experience in right now, what we need 463 - > to do, who we need to be as leaders in 464 - > tech to attract all of those A players to our teams. 465 - >
Speaker 3: What do you think? I think the main thing is 466 - > to well have something think about them, right. I think 467 - > it's not about It's never about us, It's it's about them. 468 - > So it's thinking. You know, if you've met A players 469 - > and you ask them like, what are you looking for? 470 - > What gets you excited? 471 - >
Speaker 4: And then. 472 - >
Speaker 3: Based on that and based on what aligns with who 473 - > you are, putting together a clear story for them. I 474 - > think it's funny when I read most of the job 475 - > descriptions and offers that are out there, they always say 476 - > we're looking for They never write it from the perspective 477 - > of if this is you, this job is for you, right, Like, 478 - > we don't care that you're looking for this like HP 479 - > or whatever company. Right. I'm reading this like taught to me. Oh, 480 - > if you are a blah blah check, yes I'm that 481 - > person and you like too, yes that's me, and you 482 - > want to yes that's me, then this position is view 483 - > Oh okay, that's appealing. So I think really thinking about 484 - > it that way and then designing a hiring process that 485 - > showcases that you're different, showcases that they are going to 486 - > be interacting with with rock stars and they're going to 487 - > be doing rockstar work is one of the biggest things, right, 488 - > So that's that's one. I get it that this is 489 - > at the point where people are already applying to, you know, 490 - > to be hired or looking at your job posting, which 491 - > isn't necessarily the case for most people. I do think 492 - > there there's a lot of This is one of those 493 - > things where there aren't many shortcuts, right. You have to 494 - > you have to try and determine what does a rock 495 - > star profile look like, how do I go hunt them, 496 - > how do I go find them? Where are they? And 497 - > the thing is like rock stars tend to hang out 498 - > with the rock stars, and so there are different pools 499 - > of people we realize good like find these people, I 500 - > talk to them and like where it spreads and and 501 - > it works well to get their attention. But unfortunately I 502 - > have not seen any shortcuts, like you know, sourcing agencies 503 - > all that kind of stuff. Like the top players are 504 - > not looking for a job. The top players are crushing 505 - > it at their job today. The only thing that's going 506 - > to make them move is like an incredible opportunity to 507 - > work on something exciting within an exciting team. So again, 508 - > think about it from like the sports analogy isn't perfect, 509 - > but there's a lot of good stuff. Right, if you're 510 - > a top player at one of the top European teams, 511 - > you're not looking to go, Like the only way someone 512 - > can get you is if they have a project that's 513 - > really exciting. Right. I'm a big fan of the the 514 - > Kilsche handball team, right, and you don't get talent du 515 - > Chebayev to come and coach that team for years without 516 - > having like something to offer. And it was this whole 517 - > idea of like this is like a very like historically 518 - > meaningful team in Poland and there's like a big project 519 - > to try and bring it to the front of the 520 - > of the European scene. And that's something when you're you know, 521 - > a coach of the caliber of talent, you're like, yes, 522 - > I want to do this and I'll go to Poland 523 - > When he could have been coaching any of the German 524 - > like top teams or Spanish or Barcelona and now he's 525 - > going to be coaching the French teams. That's very happy 526 - > about that anyway. So I think having a clear idea 527 - > of what is the project, what is exciting about it, 528 - > and why are they the good fit for it, and 529 - > then surrounding them with the right kind of uh, making 530 - > sure you're surrounding them with the right kind of talent 531 - > and that you're ready to make the the hard decisions 532 - > that have to be made to you know, to keep them. Yeah, 533 - > I think it. 534 - >
Speaker 2: Requires a lot of courage from the leadership perspective, because 535 - > it's like, what I see is a lot of people 536 - > who are leading teams are not very well preferred to 537 - > do it, you know, and they never were taught how 538 - > to make decisions that are unpopular. We talked about it 539 - > a little bit before we hit record, and I think 540 - > that it costs a lot of frustrations, burnout, misunderstandings, wasted time. 541 - > And I see a lot of tech teams. I work 542 - > mostly with the engineering community. I am a data science 543 - > so this is my area as well, and they see 544 - > the pattern here as well that people are saying to me, Hey, 545 - > you know, I'm leaving this team. I had this conversation 546 - > like last week, very fresh and my head, my boss 547 - > is like, is avoiding decision making whatsoever. We are talking 548 - > to him that we need this and that and that 549 - > and even very much open because they're just like, hey, 550 - > I'm beyond being nice and being nice, and I'm just 551 - > going to say it out loud because this is what 552 - > I need to be a good leader for my team. 553 - > And they are not making decisions. They are just trying 554 - > to avoid it as much as possible. It is super uncomfortable, 555 - > and this is costing me so much quote unquote, And 556 - > those people are earning so much money and they're super 557 - > high in organizational structure that you again you would assume 558 - > that they're supposed to be those people who actually are 559 - > courageous enough to make bold decisions, sometimes super uncomfortable decisions, 560 - > because this is a part of the job and last 561 - > this week actually today, very very early in my morning, 562 - > I talked to him, one of him, one of one 563 - > of them, and that he told me that he didn't 564 - > even know that it was part of his job to 565 - > do that. Then that and you know, not going into 566 - > too many details. So my question about it to you 567 - > is how we can if it, if it's even possible, 568 - > how we as leaders, ourselves included, we can teach ourselves. 569 - > We can learn the courage to be bold to you know, 570 - > just to sometimes quit the logical thinking critical might be 571 - > not critical thinking, but the all new using logical approach, 572 - > analytical approach and analytical part of yourself, and and just 573 - > you know, to be the best version of yourself as 574 - > a leader for your team, because what you need, what 575 - > you mentioned is they're building. We need to build for 576 - > those people the environment where it's they are going to 577 - > be the most important things, not us. So is it 578 - > even possible for us to learn how to be more 579 - > courageous or what do you think? 580 - >
Speaker 3: Yeah, I think it's a it's a great question. And 581 - > uh it's you know, like leadership is like multi layered, right, 582 - > but this one, in my opinion, is is one of 583 - > the highest levels of leadership. So it potentially goes all 584 - > the way up to the CEO, and it's around what 585 - > what do we value? What do we reward? Right because 586 - > at a lot of companies, we you know, we the 587 - > stick is always present, Right, you mess something up, We're 588 - > going to come at you with a stick. But we 589 - > don't reward failing at a moonshot, right, Like sailing at 590 - > a moonshot means that you had a moonshot idea that 591 - > you went all the way to trying and at some 592 - > point it didn't work, but at least you tried. And 593 - > this is like, this is where like this is cultural. Right, 594 - > there's an element of how do we reward being bold? 595 - > How do we reward being courageous? It's like the you know, 596 - > the famous saying nobody gets fired for buying IBM, and 597 - > the thing is, maybe we should start firing people for 598 - > buying IBM uh and and and it's if you're not 599 - > making any bold decisions, then you're slowing the company down. 600 - > I think the we're at a time, especially in tech, right, 601 - > we're at a time where AI is accelerating everything. Velocity 602 - > is now the name of the game. The question is 603 - > like how quickly can you move? And it's not just 604 - > the speed at which we're seeing the AI native companies 605 - > move is not this kind of like linear. We just 606 - > work a little bit harder and we'll get there. These 607 - > are massive step functions that are happening one off to 608 - > the next, and the only way that happens is you 609 - > make a gigantic bet or multiple gigantic bets, hoping that 610 - > you know you're going to be able to sling yourself 611 - > from here to there in a very short period of time. 612 - > That's the only way you can get that kind of 613 - > exponential result. And if you don't encourage this, if you 614 - > don't create time space rewards systems to promote this, then 615 - > it's not going to happen. And then you are going 616 - > to end up with a lot of people that are 617 - > well paid, very comfortable, and they're just chugging along. Right, 618 - > They're just doing the thing, and that's what they're asked for, right, 619 - > What we're asking them to do is to not mess up. 620 - > They're not messing up. The easiest way to never mess 621 - > up is to not do anything right. So maybe we 622 - > need to start changing the way we think about management. 623 - > That's like, to me, is a massive one that's changing 624 - > with AI and also thinking about what is the reward structure, 625 - > how do we value and how do we promote these 626 - > bold choices. I don't think most I've seen most companies 627 - > don't do it yet. That's just what I just wanted 628 - > to say that. 629 - >
Speaker 2: Actually, my book is coming so fingers crossed for everything, 630 - > and I I wrote a part of it because this 631 - > is the book is from Code to People and this 632 - > is about everything that we are talking about here, about 633 - > that being the leader that actually people want to work 634 - > with and the whole communication intelligence for the structures, frameworks, 635 - > et cetera. And there's one piece that actually stuck with 636 - > me like months ago, and I thought about it. I 637 - > thought about it that people are failing in leadership in tech, 638 - > not because they are bad potential for leadership or they 639 - > are bad people and they just want to screw everything up, 640 - > but they never made a shift from the rewarding perspective 641 - > because as I see it, they were rewarded for the 642 - > delivered work, the tech delivered work tasks as we mentioned before, 643 - > and they moved to team lead position, manager position and 644 - > then for further in your organization, and they never they 645 - > never had this switch that right now they're rewarded not 646 - > for their their job delivered the tasks actually crossed out 647 - > from from the checklist, but from they are. They should 648 - > be rewarded for what their team is delivering. And this 649 - > is the difference. So this he as well. No, I'm 650 - > super curious. 651 - >
Speaker 3: Yeah, no, absolutely, And I think the most most organizations 652 - > have terrible first time manager onboarding. That's okay, it's it's 653 - > a very common one. And uh, there's so many things 654 - > that change when when you become a first time manager. 655 - > There's the idea of you know, your performance is not 656 - > measured on what you do, is measured based on what 657 - > the team is able to accomplish. There's this that the 658 - > notion of like the first team principle, which I find 659 - > always very interesting to say, like a manager's primary loyalty 660 - > is to his peers, not to his direct reports. That 661 - > is something that I see go wrong with almost every 662 - > I mean actually with every single first time manager, where 663 - > they're constantly fighting for their direct reports rather than fighting 664 - > for their peers. And yeah, and that's because we're not 665 - > training them to understand why. You know, like your role 666 - > is the cross alignment and then using that cross alignment 667 - > to bring focus to your team rather than your goal 668 - > is to be the voice of only of your team 669 - > in this in this organization. So I think that the 670 - > lack of training there is is huge and I think 671 - > we often make the mistake of taking And this is 672 - > why again I love the dual track where I see 673 - > versus manager. I do think a lot of the best 674 - > players engineering aren't meant to be managers, and there's a 675 - > lot of value in having managers who aren't necessarily the best, 676 - > the most technical people. The respect they're going to earn 677 - > from the team is not necessarily that there right better code, 678 - > or they scored better than them at lead code or 679 - > whatever it's it's rather that they help bring focus on 680 - > the right things. They help make the job easier, and 681 - > they are good at the cross functional alignment that is 682 - > required again to make the life simpler for all of 683 - > your direct reports. 684 - >
Speaker 2: This this is your job, right And I have a 685 - > lot of different conversations about being authority and in teams 686 - > in tech, and we are talking a lot above do 687 - > I need to be the best engineering a team, the 688 - > best data scient is the best cybersect person to actually 689 - > be the best manager for them? And right now it 690 - > is the market is shifting and so we see hr 691 - > people more getting people salespeople meeting tech teams in the 692 - > other way around, because you don't you know, it's extremely 693 - > hard for you to not micromanage people because you actually 694 - > know how to deliver a certain task, how to code, 695 - > hard to do analysis, do a power va, and every 696 - > every single thing. And I have these conversations all the time, 697 - > and there are two camps. I would say. One camp 698 - > is that, hey, this is impossible to lead a team 699 - > if you are not an expert, because how on earth 700 - > you should control them or know what they do and 701 - > build authority as an expert. And then the other camp is, hey, 702 - > this is better that you don't know how to code, 703 - > or how to how to do the marketing campaign or 704 - > how to do the sales, because this is this is 705 - > not your job. Your job is to create an environment 706 - > for them so they can deliver their job. You need 707 - > to know what they you know the process is that 708 - > they're actually working with them, so you can be a hand. 709 - > But it is not necessary for you to be an 710 - > expert in your team's area. So which count you are 711 - > end because I'm hearing that who can be in the 712 - > second one? But then I just wanted to check. 713 - >
Speaker 3: I think it's it really depends on how you set 714 - > up your structure, right. I think having a like having 715 - > a non tech person leading a tech team only works 716 - > if they are principal distinguished like top ice players who 717 - > can do that, who can look at the the technical part. Again, 718 - > just to go back to the analogies of sports, right 719 - > when you think of like a big soccer because not 720 - > everyone is that familiar with henball, right, But in soccer 721 - > you have the main coach, right. The coach isn't the 722 - > coach of the goalies and the coach of the strikers, right. 723 - > The coach's perspective is what is the right team set up? 724 - > What is the right strategy based on the players have, 725 - > based on who we're playing against, like how are we 726 - > going to set up to maximize our chances of winning? 727 - > And then when it comes to making sure the goalies 728 - > are the best at doing goalie stuff and that the 729 - > strikers are the best at striking, we then have you know, 730 - > specific coaches for that, and then we have a fitness coach, 731 - > and then each of these are the experts at helping 732 - > individual players understand by the way in your form, like 733 - > this is something you need to change when you're running 734 - > you're not doing this like we're going to have to 735 - > train on your like you know, whatever like zone two 736 - > or whatever kind of challenge you want to work on. 737 - > And the main coach doesn't have to be the expert 738 - > at all of this because he's put in place a 739 - > system that allows to delegate to specific experts the expert 740 - > level discussion. So the same way that most CEOs are 741 - > not technical people, right, and they're the ultimate leader of 742 - > the engineering organization. But what they do is they place 743 - > a ETO who's excellent at technical stuff, and then actually 744 - > pretty quickly you have the CTO who's generally not a 745 - > people manager, and then you have the VP of engineering, 746 - > who is the people manager and potentially isn't the most 747 - > technical person in the organization. But what they're excellent at 748 - > is understanding how do we place the team in order 749 - > to win, Like what are the right systems to put 750 - > in place? Where do we need DevOps? Where do we 751 - > remove DevOps? Like how do we deal with like soft 752 - > two compliance slowing us down because like manual reviews for 753 - > like all of our controls, like these are things that 754 - > like this person will handle without necessarily being the most technical. 755 - > But then when the engineering team has a very technical, 756 - > like very very deep question. That's when you need the 757 - > CTO to be able to be there. And this works 758 - > only if the CTO and a VP of engineering are 759 - > in lockstep, right. The worst thing that can happen is 760 - > if these two don't work well together, then the engineering 761 - > team stops respecting the VP of engineering because they're considering, Ah, 762 - > he doesn't know what he's taught or she doesn't know 763 - > what she's talking about. I'm just going to go to 764 - > the CTO because they do. And that's a failure of 765 - > like CEO leadership, of not making sure the structure is 766 - > designed to make sure everything works well. So I would 767 - > say in probably camp, you don't have to be technical, 768 - > but I do believe there is a level of understanding, 769 - > Like you need to be able to understand the process. 770 - > That's like one of my big things with AI. I 771 - > think it's too easy to fake output. We used to 772 - > manage based on output. Today we have to manage on process. 773 - > So you have to be able to ask your data scientists. Cool, 774 - > so you built this model, walk me through the steps, 775 - > like what decisions did you make? Like how did you 776 - > come to building this model? Why are you using you know, 777 - > boosted decision trees or whatever it is. Like you need 778 - > to be able to entertain that conversation and to understand 779 - > it well enough to be able to ask the right 780 - > kind of questions in order to even help them see 781 - > some of the blind spots they might have. So I'm 782 - > kind of in the middle where I think you need 783 - > to have a good enough understanding of what's going on 784 - > that you can ask questions and understand like the answers 785 - > and understand where when BS is coming out. But you 786 - > don't need to be a data scientist or you know, 787 - > a production engineer, you know, writing code for forever and 788 - > like innovating on whatever uh on small protocols. 789 - >
Speaker 2: Yeah, no, I agree. I saw so many success stories 790 - > where people actually were great needs in areas that they 791 - > didn't have any expertise whatsoever before. As you mentioned that 792 - > you need to understand what is happening and understand the 793 - > process and understand what's with that, what the what the 794 - > goal of the of the team is. But I see 795 - > honestly more advantages done disadvantages because I see so much 796 - > of a market management and not delegating and being burned 797 - > out because it is not that you don't have the team, 798 - > you just don't have the skills and your mindsets and 799 - > I see and you just you just cannot gain just 800 - > give it, give it away, and it creates and it 801 - > brings more harm and good to team, you know, And 802 - > this is something that I would I would love to 803 - > do more experiments like that and see the you know, 804 - > a b testing to to see two teams that are 805 - > working in similar space in the same organization with a 806 - > tech lead that is actually a tech person and that 807 - > non tech person but actually speaks the language. Because this 808 - > is essential and what is the difference between performance engagement 809 - > level and you know all the KPIs we can actually 810 - > measure how the team is going on and on the 811 - > psychological level, but also on the performance level because this 812 - > is an organization and meteorytur also yes, this is this 813 - > is this is interesting. The one last thing because before 814 - > we go into the linking crown, because last week I 815 - > was at the word computers actually so WordPress community and 816 - > it was a big event some twenty I think that 817 - > over two thousand people twenty five hundred or so were 818 - > there's a huge event. And at the end of the 819 - > event was it was a panel that was that on 820 - > the panel was the one person from the same level 821 - > and lead architect and I had a product. Actually, so 822 - > this is a panel and there were discussions that there 823 - > were discussions what is happening in WordPress and all of 824 - > the things that are happening. And in one moment there 825 - > was a question from the lead architect I believe or 826 - > that I had a product to the to the audience 827 - > that was their listening, and they started to talk about 828 - > word Press seven point all and the contribution. And then 829 - > there's a question, just please raise your hand if you 830 - > still coote manually. And I just look around because I 831 - > was sitting in a front row. Over I would say 832 - > over a half of the room, rose your hand. So yes, 833 - > I still cold with my hands manually, and I was. 834 - > I was. I started to think, okay, interesting, this is 835 - > an open source so maybe this is different in a 836 - > different space. But I started to think, what is going 837 - > to happen to those people and what should we do 838 - > as leaders right now? Again in middle twenty twenty six, 839 - > you you know, you had AI slash machine learning and 840 - > your profile before it was fancy started two thousand and 841 - > nine or something. So I said the story and I'm 842 - > asking you, what do you see that this is happening 843 - > right now in teens intact because of what is so 844 - > what is happening with AI for the last three years, 845 - > let's let's say, and what is going to happen next 846 - > two And what is going to happen to those people 847 - > who just roll their hands and to say, hey, I 848 - > do it manually all the time, not using any any 849 - > any any any automation whatsoever. So yeah, I'm opening this 850 - > loving one. 851 - >
Speaker 3: I think there are two, at least two things have 852 - > come to mind. One of them is there are a 853 - > lot of people who abandoned AI at CHATBT or like 854 - > GPT three point five, like the outputs are not good enough. 855 - > I'm just moving on. Something fundamentally changed with Opus four 856 - > point five in November last year. 857 - >
Speaker 2: It was just. 858 - >
Speaker 3: Something like fundamentally changed. Like I was still you know, 859 - > I was fortunate enough to get early access to get 860 - > help copilot. It feels like it feels like ten years 861 - > ago now at this point when it would just autocomplete 862 - > and I was like, oh my god, I'm like writing 863 - > so much faster. There's like so many how does it 864 - > know this is what I'm trying to do. But I 865 - > was still like writing a lot of it manually and 866 - > then I would use JGPT here and there. It was 867 - > very buggy, and so I ended up like using it 868 - > really to write specific functions or to write tools, but 869 - > I never really got it to function as a outsourced engineer. 870 - > And then Opus four point five was really the moment 871 - > where I realized, Okay, now we have something where if 872 - > you have the right guardrails, if you have the right 873 - > system in place, it can actually do proper work. And 874 - > so I think there's a bunch of people that you know, 875 - > abandoned at that point, and I think for those people, 876 - > we need to get them to actually try it, try 877 - > the new models and realize, Okay, this is actually genuinely great. 878 - > It's not perfect, but it's great. And that's going to change. 879 - > You know, maybe like twenty percent or ten percent of 880 - > the people that we're still raising their hand are no 881 - > longer going to raise their hand. I think that the 882 - > biggest challenge is and again, like this task monkey thing, 883 - > hopefully we can put in the show notes like it's 884 - > I took this from a Twitter post that I thought 885 - > was really interesting in like thinking about system thinkers and 886 - > all of that. The people that are crushing it on 887 - > using AI are not vibe coding, right, It's like AI 888 - > engineering you're thinking about like how do I architect a 889 - > system that allows the AI to produce code with the 890 - > right guardrails, with the right context at any given point 891 - > in time, that has loops to verify what it's doing. 892 - > That is system designed, right. It's really thinking about when 893 - > I'm building a system, if I have to build a feature, 894 - > if I break down those tasks and I allocate some 895 - > kind of like what would be the outsourcing costs for it? 896 - > So's say, think about the architecture of the system. That 897 - > is something if you were to outsource it, you'd pay 898 - > a lot of money. Now, if you have all of 899 - > the architecture, you have the interfaces, and now you just 900 - > need to write the code, probably wouldn't pay very much 901 - > for that part, right, And that's the part you can 902 - > give to the AI. And so what I find is 903 - > that the people who haven't figured it out and just 904 - > can't figure out how to think in systems and to 905 - > realize it's not about producing the code, it's about thinking 906 - > about the architecture and think about like bringing to consciousness 907 - > everything that you were doing without even thinking about it, right, 908 - > and understanding that that's something then you now have to 909 - > give to someone else same thing. That want of challenges 910 - > when you're first time leader is that you've been used 911 - > to doing things and you're great at doing them, but 912 - > you never realize why you were doing them, and now 913 - > you have to teach someone to do them. Like that 914 - > first step of teaching is so interesting. Thing when I start, 915 - > you know, coaching handball for like young kids, the servilizing, 916 - > Oh my god, Like even just like throwing the ball 917 - > isn't something that's natural, Like they don't get it, Like 918 - > what is the movement of the risk? Like when do 919 - > you like use their risk to whip the ball? And 920 - > like what's the kind of like shoulder movement like arm up, 921 - > armed down? Like if people like kind of push the 922 - > ball like this, Like things that just like are natural 923 - > to you, Like you start thinking about them and you realize, okay, 924 - > like there is a process to this, and being able 925 - > to document the process allows you then to offload it 926 - > to an AI. So the long story short is, unfortunately 927 - > for the majority of those people, I don't know that 928 - > there's much of a working future for them because if 929 - > you're not able, like the bar of where automation can 930 - > go in this pyramid all the way system thinkers, the 931 - > bar is now higher. If all you do is executing tasks. 932 - > That's where I is today, and it's doing it at 933 - > a accuracy that is higher than like most people. And 934 - > so if you're not able to think in systems and 935 - > to have the AI do the work for you, I 936 - > think it's yeah, it's going to be a rough couple 937 - > of years. 938 - >
Speaker 2: And you know, the skills is just one thing that 939 - > the mindset shift is in mass. 940 - >
Speaker 3: The mindset is the main thing. Right, If you're not 941 - > able to think in systems and decompose everything you do 942 - > into smaller subtasks, that's going to be rough. 943 - >
Speaker 2: Yeah, optimistic, Hey, but this is what it is, right. 944 - >
Speaker 3: And I guess what it is and I think there 945 - > that's where there's one thing, which is like individually, we 946 - > all need to make sure that we you know, we 947 - > just the bar has gone up, so now we have 948 - > to think about like how do I make myself? How 949 - > do I put myself above the line. And that's really 950 - > the practicing system thinking, practicing, just practicing thinking, right. I 951 - > think this is like it's it's hard to say this 952 - > in an era where we're all consuming like everything in 953 - > like five second, like TikTok videos and stuff, but critical 954 - > thinking is more important than ever because we're going to 955 - > spend more time critically thinking about the architecture we're putting together, 956 - > Critically thinking about how did the AI implement this, Critically 957 - > thinking about are my harnesses good enough to make sure 958 - > that this thing is not going to go off the rails. 959 - > That deep thinking and like spending time on a problem 960 - > ahead of the execution, I think is a skill that 961 - > we've started losing a little bit and is going to 962 - > be incredibly important in coming years. Second part to this is, 963 - > I think from a policy perspective, not necessarily the topic 964 - > of this conversation, but there's a lot of work we 965 - > need to do collectively to bring this to the forefront 966 - > of our political leaders to understand. Look, it's great that 967 - > Entropic and Open AI keep on saying that they're going 968 - > to create mass unemployment with their AI. Shouldn't that be 969 - > a warning sign that we should think about what happens next, 970 - > Like if they actually reach this thing, what happens, like 971 - > how are we making sure that we are setting ourselves 972 - > for success rather than another set of revolutions where people 973 - > are just going to storm the data centers and burn 974 - > everything to the ground. Right, it's a pretangent, but it 975 - > is something that we should be spending more time thinking 976 - > about and at a local scale. I've been spending a 977 - > lot of time thinking about redefining the roles and responsibilities 978 - > of every job position in the organization in an era 979 - > where more of the work is done by AI than 980 - > what are we expecting of you? What is your role? 981 - > What are your responsibilities? 982 - >
Speaker 2: Like? 983 - >
Speaker 3: What are you accountable for? That changes what I expect 984 - > of every employee, But it also changes who I'm hiring, 985 - > because now I'm potentially hiring fewer super super deep experts 986 - > at something and more generalists, more people that are excellent 987 - > at learning quickly. They understand the concepts, they understand the 988 - > laws of thermoidynamics, and like basic like advanced math and physics, 989 - > and they can jump from one thing to another. These 990 - > people tend to crush it in the AI era a 991 - > lot more than the super deep experts who are very 992 - > good at doing one thing but struggle to translate what 993 - > they're doing. On the note of a European in the US, 994 - > that's one of the things that I found fascinating where 995 - > the European or at least the French engineering institutions are 996 - > a lot about like horizontal learning. We're very we're Jacksonville Trades, 997 - > master of none, like become very good at like just 998 - > learning complex stuff. And for the longest time I thought 999 - > that was stupid. I came to the US and I 1000 - > went to Stanford for a little bit, and I met 1001 - > people who are excellent at what they were doing. They were, 1002 - > you know, getting masters in mechanical engineering, and they were 1003 - > struggling with some of the algebra that I was doing 1004 - > the year after I graduated from high school. And that's 1005 - > because we were learning just that, like it nothing applied, 1006 - > like it had no idea why you were doing this. 1007 - > And what I'm seeing is that this like ability to like, 1008 - > you know, learn a bunch of things and shift and 1009 - > not be fitted in a mold is has created some 1010 - > incredible talent that I think is going to be the 1011 - > the top talent of this coming era. I think we 1012 - > are entering the era of the generalists. 1013 - >
Speaker 2: Yeah, and this is I think this is a subject 1014 - > for another episode, you know, because how to create teams 1015 - > right now, what is going to happen and what is 1016 - > already happening on the market, and when we can actually 1017 - > observe it, and how we as leaders and as I 1018 - > sees as well, what we should do to stay in 1019 - > the market and not being kicked out of it because 1020 - > we are not adaptive enough or we just say out 1021 - > it worked for so many years, So what actually can 1022 - > happen with the whole approach? And I think we both 1023 - > see those people all the time using the example of 1024 - > the right hand right, So this is something that I 1025 - > think that yeah, you should do, I know, part two 1026 - > of the conversation, because you know, we just scribe the surface. 1027 - >
Speaker 4: So this is it. 1028 - >
Speaker 2: But right now, because of the timing, I'm aware of 1029 - > attention span of our audience, I will go to the 1030 - > liking ground. Frances every ready, Yes, Okay, where's go. So 1031 - > what is the worst leadership advice that you ever? 1032 - >
Speaker 3: Titles don't matter. Oh, this is a big debate we 1033 - > had with my co founder. He genuinely doesn't care about 1034 - > titles and he does it so he's like, I don't 1035 - > get why people want titles whatever, it doesn't matter. Just 1036 - > as long as they get the job done, we're fine. 1037 - > And the funny thing is like Conway laws, Conway's law 1038 - > comes to bite. So it's this idea that your product 1039 - > matches your org. It doesn't matter how small your this 1040 - > principle is always going to apply it might seem meaningless 1041 - > to spend time thinking about like what are the titles 1042 - > and what is the ORG structure and all of this, 1043 - > but the reality is like those decisions as leaders will 1044 - > dramatically impact what your team is able to ship. It's 1045 - > very funny because right now I'm in an organization where 1046 - > I'm part of a CTO office. I'm working on innovation, 1047 - > redefining the software development life cycle and building products, and 1048 - > we're moving very, very very fast, and we're at the 1049 - > point where now some of the things that we've built 1050 - > are being sold and now we need to bring them 1051 - > back to the engineering team. And the engineering team's like, 1052 - > oh my god, like this is not how we work. 1053 - > And so because we've created organizationally this thing where we 1054 - > have a relatively slow moving team and a like rewarded 1055 - > on as moving and learning team, we now have two 1056 - > different products that are going to be hard to bring 1057 - > back together again. And not saying it was the wrong 1058 - > thing to do, I just saying that it always happens. 1059 - > And so spending time on the ORG layout and the 1060 - > ORG layout and the titles, like titles mean things to people. 1061 - > They're actually more important than you might think. As a startup, 1062 - > you kind of want to give see titles and vps 1063 - > to everyone. It's going to come back to buye you. 1064 - >
Speaker 2: It's might not be important for you, but it is 1065 - > not about you anymore when you are so yeah, what 1066 - > is the one thing that you wish you knew at 1067 - > the very beginning of your leadership journey? 1068 - >
Speaker 3: We touched upon this, but I would say it's the 1069 - > cost of okay players I've worked with. I still work 1070 - > with some truly truly exceptional people. Like the the energy, 1071 - > the momentum, like it just it feels magical. It feels 1072 - > like nothing can stop us when we're working together, and 1073 - > you know, you wake up you want to work with them, 1074 - > Like I'm I'm excited to go to work and to 1075 - > work with these people. So I know they're going to 1076 - > challenge me. I know we're going to have a good time, 1077 - > like you know, pushing each other and then we're you know, 1078 - > getting the best out of one another. But as soon 1079 - > as you introduce like an okay person in the mix, 1080 - > everything slows down, Like you know, it's you know, the 1081 - > train only moves at the pace of the slowest wagon, right, 1082 - > like all of a sudden, like everything just feels slower. 1083 - > It's more painful, you have to explain more. You're not 1084 - > challenging each other, and so there isn't that resonance like 1085 - > there can be in a team of you know, top players. 1086 - > The thing related to that to make it maybe more 1087 - > tangible for people listening, so I use an interview scale 1088 - > where every it's like one to four. One is like 1089 - > I'm strongly against hiring four, I'm strongly for hiring. Very 1090 - > important in my mind to have a even number. Indie 1091 - > scales never an odd number, because if you go one 1092 - > to five, everybody's going to be a three. When you 1093 - > have to determine between two or three, that's a decision 1094 - > like do you hire or not hire. And I used 1095 - > to think that the you know, we want to hire 1096 - > people that didn't get a two in any of their 1097 - > interviews with the team, because that was the kind of 1098 - > this disqualifying factor. What I've found is that the thing 1099 - > that matters is actually like how many fours is this 1100 - > person getting? And I found that the absolute top players 1101 - > generally have like thirty percent of four ratings, but they'll 1102 - > also get a one and potentially or like generally a 1103 - > two or a few twos and maybe a one, And 1104 - > that's actually not disqualifying. It's often like people that feel 1105 - > threatened more than anything else. But if you hire people 1106 - > are just like a bunch of threes, there's no twos, 1107 - > no fours, you end up again. It's like this idea 1108 - > that you know, if you don't you're just not going 1109 - > to move. Nothing's going to change, and you're just going 1110 - > to be static. And in the age of AI, I mean, 1111 - > you know, being at a standstill is basically a death 1112 - > sentence as a company. 1113 - >
Speaker 2: Amen to that. And what is something Because we've talked 1114 - > about learning transformation and that's you know, staying in the 1115 - > same place and not being kicked out of the market. 1116 - > So what is something that you do for your growth? 1117 - > How you learn? Maybe see you or read something lately, 1118 - > or listen to something or watch something that inspired you 1119 - > to thought to something that we can all use a. 1120 - >
Speaker 3: Lot of things. The number one is I have a 1121 - > network of people that I highly highly respect, these kind 1122 - > of top players. Some of them are you know, in 1123 - > the company or in my team. Some of them are 1124 - > even direct reports. Some of them are outside of the team, 1125 - > and I go to them for inspiration, like what are 1126 - > you working on? What's something cool you've seen this week? 1127 - > And the funny thing is I find that LinkedIn Twitter. 1128 - > I hate to say it, but I found more value 1129 - > you lately in Twitter. Then yeah, then on LinkedIn. It's 1130 - > just a cesspool of I call it tech porn. People 1131 - > are just pretending that they're doing things. They're like showing you, Oh, 1132 - > look at my five Mac minis that are rigged together 1133 - > with like open claw running you know, at nauseum, and 1134 - > you're like cool, but like, what are you doing with it? 1135 - > Like what's like, what do you have to show for it? 1136 - > Generally it's nothing. When you talk to people that you 1137 - > know and you respect, they'll tell you very insightful and 1138 - > very detailed things like, oh, like I ran into this 1139 - > particular issue with like the Claude harness and this is 1140 - > how I fixed it and and yeah, so that that's 1141 - > one big thing. 1142 - >
Speaker 2: And then. 1143 - >
Speaker 3: Building. I think I I'm pret diligent about carving out 1144 - > time every week. I know every week can seem like 1145 - > a lot of other people. I'm part of CTO office, 1146 - > so it's it's fairly easy for me. But even as 1147 - > a co founder, I would always have one afternoon per 1148 - > week where I'm just building and it doesn't matter if 1149 - > it relates to the company. If it doesn't. But feeding 1150 - > that curiosity and having to build is a great forcing 1151 - > mechanism to go see what are other people doing, how 1152 - > other people solve this problem? What are some of the 1153 - > things that are out there. I recently discovered this tool. 1154 - > It's not sponsored, but vast dot ai. They're basically a 1155 - > true row for GPU. So if you need an H 1156 - > one hundred, very expensive, very hard to buy these days, 1157 - > you can rent one. So people have these GPUs sitting 1158 - > at home. They're connected to a network and you basically 1159 - > send your code and it's going to run on their GPUs. 1160 - > If you want to do fine tuning or distilling, or 1161 - > like some kind of heavy process that doesn't fit in 1162 - > like the Google Collab GPUs, you can actually do this 1163 - > for like ten bucks. You can distill like a pretty 1164 - > like big Mini Max or Quen model. And that's something 1165 - > that I discovered by just like lollocating time to doing this. 1166 - >
Speaker 2: Yeah, I think that this is an extremely valid point 1167 - > because you just not learn from reading or watching before 1168 - > you start actually doing something like that. You're just amazing 1169 - > you brain with a lot of information about the question. 1170 - > So what so this is this is a great advice. Francis, 1171 - > Thank you for the conversation. I loved it, and I 1172 - > think that was uncomfortable from time to time. So this 1173 - > is this is extremely well designed process that happen over 1174 - > over us, I would say, So, I hope that your 1175 - > audience will get a lot of value out of it, 1176 - > a lot of real talk, not non fluff, but just 1177 - > you know, real examples and the truth. That's sometimes we 1178 - > are just not courageous enough to say it loud. So 1179 - > thank you. And the question is if somebody wants to 1180 - > connect with you, check out on you to see what 1181 - > you're doing, where they can find you, will tell us 1182 - > all the places. 1183 - >
Speaker 3: Yeah, I would say LinkedIn is the best place. Even 1184 - > though I hate the platform, it's where I'm the most 1185 - > active trying to make it better by posting some actual 1186 - > content that I would want to see on there. So 1187 - > I'm generally freely responsive there, so that's the best place 1188 - > to find me. 1189 - >
Speaker 2: Great. Yeah, we're going to put the link in the 1190 - > show notes so everybody can just click and then check 1191 - > you out. So yeah, thank you. It was a blessed 1192 - > to have you on the show. Yeah, thank you for 1193 - > having me, of course, and thank you all for listening 1194 - > because we are doing it for you. Of course we 1195 - > are enjoying conversations, but this is for you, so we 1196 - > don't know what's what resonated the most, what you hated, 1197 - > maybe you'll disagree with something. You know, we just love 1198 - > to have you in our DMS. Some LinkedIn and we'll 1199 - > also that and comments. So just go and leave some 1200 - > and shut the link with somebody who might actually use it, 1201 - > because you know, there are so many things that are 1202 - > happening right now so we can all be better together 1203 - > as a community. So thank you, Anthlesy you soon in 1204 - > another episode of the Lemma PEG Leadership Podcast by Everybody 1205 - > Lehman Tech Leadership Podcast
Other episodes covering the same guests and topics, from across The B2B Podcast Index.