I spent a good part of the early half of my career or so thinking that my job was to think as good as I could, to have the clearest thinking, to have the best thoughts, the smartest thoughts, the most complete and most considerate thoughts. And I somehow believed that that would translate into success in my career, especially as a software engineer, that somehow if I was to write the best code, which was downstream of having the best thoughts, that that quality would translate directly into success for me. And I think a lot of engineers fall into this trap because very often that is a useful illusion, which is kind of ironic. That's what we're talking about in today's episode. It's a useful illusion to get us somewhere. It's a useful illusion in the sense that if we follow that belief, then we will likely invest in improving ourselves. And starting from baseline, that's probably going to be correlated with some kind of success. Because if you begin to have better decision-making processes, if you think more clearly about the problem that you're solving, for example, if you can weigh options and make good decisions, then you're probably going to be more successful than someone who can't do those things. So this was what I built up in my mind as the primary part of my job. But very often, this thinking, this overthinking, this focus on analysis and on nuance and on exceptions to the rule, on being correct to the hundredth percentile, it often underserved me because I would overanalyze things. It would get in the way of me taking an action. And interestingly, in kind of a meta way, I learned that the best thinking doesn't produce necessarily the best decision-making. That decision-making is not just about gathering information. It's about gathering the most complete information. That decision-making is also weighing things like time to make that decision. That if we were to always be trying to keep in mind every alternative or every exception, then the limits of our human mind will be hit. We'll run out of working memory, for example. And so if theoretically having all of those things in mind may help us, but practically it doesn't help us, then that excellent thinking actually stands in the way of good decision-making. Because if we're trying to load all of this context in, and we're failing to do that, then it doesn't matter how good we are at it. It doesn't matter how good the context was. And so in today's episode, I want to go back to a phrase that we said earlier, the concept of useful illusions. Useful beliefs that if you follow that belief, if you use it as a basis for decision-making, even if it's wrong, even if the illusion is incorrect, it is indeed an illusion. That you're still going to get a positive outcome from this. Right? Okay, so what we're really talking about here, if you were to go and talk to a psychologist or a cognitive researcher, these are heuristics. And the concept of heuristics, you've talked about, who knows, hundreds of times on this show probably. The concept of heuristics focuses on the idea, the idea that our brains are mostly efficiency machines. We're driven to complete a particular task. We want to do that in the most calorie efficient way, at a kind of metabolic level. We're driven to conserve calories. And so the least number of calories, you can think about this like tokens, right? The least number of calories that we can spend, the better. And so what our brain does is it makes connections. It creates labels. It kind of caches the concept. And so it'll reduce the complexity. And you can think about this as like stored routines almost. And so if you've ever found yourself, you know, driving down the road and you're having a hard time, you're driving halfway to your destination, and you realize that you're driving, right? This is kind of a phenomenon that many people have experienced, where you're driving to work, you're driving to drop off your kids at school, or you're headed to the grocery store. You can even do this walking out your front door. It's not just driving. It's almost anything that you do repetitively. Your brain is going to have this kind of stored in memory so deeply, and it's going to be able to do that. That it has very little work to do to accomplish that task. Right? So we do this for those kind of automatic routine movements, but we also do it when thinking about more abstract concepts. And so, you know, this is one of the reasons, for example, that first impressions matter. The first impression that somebody makes on another person, they kind of create a stored memory, and this heuristic approach to relationship management that none of us can avoid, by the way, we have a hard time updating that memory. We have a hard time busting that cache that that first impression lasts, right? And so now each time we meet that person, we have some immediate association. So our brain doesn't have to, we don't have to recalculate this new person every single time we encounter them, right? Okay, so we kind of see the usefulness of heuristics. Although, you know, we're not necessarily, that person, average person listening to this show is not struggling to find calories most likely. And so they're less useful in the sense that our brain needs to conserve calories, and they're more useful in the sense that we can move through a lot of concepts, a lot of generally complex concepts, very quickly, right? We have kind of these bookmarks in our minds. And so heuristics can be very useful. What I'm going to talk about today is kind of choosing to exploit this heuristic concept, right? Choose to exploit this, and then we're going to talk about the short cutting that our brain does. Because if you're like me, there are a lot of behaviors that you want to develop. There are a lot of mindset adjustments that you want to make. And if you're also like me, when it comes to overanalyzing, a lot of software engineers do end up here, where you're thinking about all the caveats and all of the reasons why that particular estimation needs to be padded a little bit more. All of the reasons why that plan may not work, right? These are things that engineers are pretty well known for. We raise concerns as kind of our, we are essentially kind of trained, whether through experience or formal training or maybe something in between. We're more or less, we're trained to raise the concern, right? That we're thinking. And so we've identified the risks and we're raising the risks. And this can hold you back. in your career. Even though early in your career, this may actually accelerate you, and you may have even experienced the negative effects of not raising the concern, not sounding the alarm early in your career. Okay, so we have this concept of useful illusions, this concept of heuristics, of exploiting heuristics. We're going to talk about how this kind of actually plays out with some specific heuristics, some specific useful illusions, right after we talk about today's sponsor, Unblocked. I'm actually wearing my Unblocked shirt today. Your coding agents have access to your code base. And probably much more than that. Maybe you've connected other tools, tool calling, MCPs. There's probably multiple agents running right now as you're listening to this. But access doesn't mean that they have good context. Agents can't easily reason across MCPs without more guidance. They don't know your architectural decisions, your team patterns, or why the API is shaped the way that it is. What does that particular field mean? What does that... acronym mean? They may fill in the blank with something that makes sense to them, but we know that it doesn't make sense in reality. So agents look in the wrong place. They deliver bad outputs, and then you spend your time correcting it, wasting time and wasting tokens. Unblocked is the context layer that your agents are missing. It synthesizes your PRs, your docs, your Slack messages, JIRA issues into organizational context that agents actually need to understand. So if you're looking for a way to get your agents to understand, if you've thought about spec-driven development, this is part of that process. So they make better plans, they write higher quality code, they use fewer tokens, and they require fewer correction loops. That means there's less human in the loop and more work done. If you're running Cloud Code or Cursor or any other agentic workflow, go and check out Unblocked because you can get a free three-week trial at getunblocked.com That's getunblocked.com. Thank you again to Unblocked for sponsoring today's episode of Developer Tea. So we're talking about useful illusions and the exploitation of this trick that our brain does of generating heuristics and then acting on those heuristics. The trick is that our brain acts as if the heuristic is true. That's the important thing. It's easy to accept the heuristic as true and so your brain acts as if it is. And I can't remember who said this quote. Maybe my note taker will look it up for me, my AI note taker. But all models are useful or some models are useful. All models are wrong. I think it goes the other way around. All models are wrong. Some models are useful. The concept is simple. These heuristics are kind of the same way. All heuristics by their nature are incomplete. They're an incomplete picture, but they are directional and they may be directionally useful. Right? So this is the, the useful illusions part, the useful part of that, of that phrase. And so let's talk about a few useful illusions that are particularly relevant to our work today as software engineers, as engineering managers. One useful illusion is that the skill of coding by hand is going away. The skill of coding by hand is going away. Now, this is useful because it will likely encourage you to adopt new workflows that the industry is going to have in demand. All right. So if you were to, you know, think about kind of two engineers, one engineer who agrees with this and immediately goes and starts learning agentic workflows, learning cloud code, about, you know, maybe local LLMs, maybe learning about how LLMs work at their core, about the cutting edge, you know, understanding how the industry is changing. And there's a whole kind of open field for this person to go and learn more about. And then there's another engineer who kind of picks this argument apart and says that, well, no, not all coding is going away. And then they spend their time kind of defending that argument. Maybe they, you know, are a little bit slower to adopt those tools that we were talking about from the first engineer. And because they're slower or because, you know, somebody may, again, using heuristics, believe that they are a detractor, that they don't want to adopt something, right? That they may gain labels that they don't really want, or that aren't even necessarily things that they want. Like a skeptic, right? So this person may at their, you know, in their real belief, may believe, or, you know, kind of in their, in their clearest thinking, if they were to write down, you know, everything they believe about these tools, they may say that, well, yeah, you know, most hand, you know, by hand coding is going away, but not all. And that there's specific circumstances where by hand coding may still be relevant or may still be necessary. And, you know, that there will still be people who are practicing it as a hobby, maybe, or they're, you know, taking something that's not very efficient and making it efficient, and that it's not going away yet, right? Maybe those are the arguments. Well, those arguments get lost in the consequence, right? And so some of this is about this useful illusions concept. If you were to act as if the thing is, was true, then a lot of the inefficient use of your energy goes away. Remember, this is about heuristics, right? So efficiency of our effort should increase. In other words, this person who just assumes that like, that's true enough to believe it. That's true enough for me to act on it. I don't necessarily have to subscribe to every part of that belief. I'm just going to act as if it's true. I'm going to act as if it's true. I'm going to act as if coding is going away. Even if I have my, you know, specific kind of nuanced arguments to that, I'm going to act as if it is. And we'll follow that path until, you know, something changes. Okay, this is a useful, useful illusion. There are other useful illusions that I think, you know, especially different phases of your career will serve you well. And one of them is that hard work pays off. Hard work pays off. Now, if you are like me, you rail against this idea at its core, because, you know, the things that immediately pop up in my mind is that not everybody's going to appreciate the hard work. Sometimes working hard in the wrong direction is going to burn you out. You know, this mindset of hard work is very culture based mindset. It doesn't apply everywhere. And what we really should be focused on is value delivery, right? All of these things are valid, you know, contests to the heuristic that hard work pays off. They're valid. Does that mean that if you were to follow that heuristic, you wouldn't have good outcomes? For most people who follow that heuristic, they end up gaining a positive reputation as someone who can get things done, as somebody who's reliable, right? Very few people listening to this podcast would like a reputation of being unreliable. Few people listening to this podcast would like a reputation of laziness or of excuse making. These are not the kinds of things that we want to be attached to our names. And so, if you were to use this illusion, use this heuristic that, you know, when all else fails, as long as I'm showing that I'm working hard, good things will come to me. This is a, an untrue belief, but it's true enough that if you follow it, you're likely to have positive outcomes, right? What are some other untrue, untrue but useful beliefs? There is, a general belief that learning is a good thing. That learning is, as long as I'm learning, I'm growing. Maybe that's the useful, the useful illusion. And this belief carries over from most of us when we were growing up. We were rewarded for learning in some way. Rewarded in school, rewarded at home. And even in the early parts of our careers, much of the kind of entry level, jobs that we take are dependent on our ability to learn things. So I was able to learn how to code and therefore I got an internship. Then I learned a lot of my internship and therefore I got my first job, right? And it continued to roll, continued to roll forward. And eventually this becomes less true as we've talked about in recent episodes. But if I continue acting as if it is true, I will continue learning. And as a general rule, learning is still associated with flexibility of mindset. It's still associated with staying kind of in that growth mindset, right? And so you're probably going to have more positive outcomes than you are negative ones. If you continue learning. The alternative belief, the alternative view that learning is, you know, sometimes is, is unlikely to harm you. But it could be a not very useful belief, right? Because if you are convinced that learning is a good thing, then you're more likely to continue learning. If you're convinced that, you know, you're a lifelong learner, for example, and you're more likely to actually spend the effort and energy to keep that a part of your life. All right. So there are probably a thousand more, a thousand more useful illusions that we could talk about. You know, some quick ones might be that clean code is better. Clean code is better. This is a useful illusion. That maintainability and performance are always worth your time. Right? That always thinking about the user. This is a really good one. Always thinking about the user's experience is a good thing. You might be a little bit heuristic. You might hear my dog barking in the background. Doesn't happen often. But when it does, she's she's fairly loud. So thinking about the user's experience or thinking about the impact on the user. It's not always the the absolute optimal thing to think about. But if you are always thinking about it, it's likely that you're going to end up producing value for the end user. Directionally, it's not a bad thing to do. Right? You know, to think, okay, should another useful illusion is go with the tool you know, right? These are almost sayings, right? You can kind of think about this as quick things that you pull, that you know that, okay, if I go with the tool I know, right, this is a useful illusion that going with the tool I know confers a bunch of benefits to me, right? So, you know, you probably have your own list of useful illusions that things that are helpful to you, even if they're slightly wrong or if they're incomplete, they're not considering the whole picture, you know, that this is, you know, a critical part of career growth is recognizing what kinds of heuristics can I lean on in order to promote the kinds of behavior that I care about and avoid wasting time in toil. I mean, this is especially true for things like performance reviews season, right? A good, a useful heuristic, another useful illusion is that, you know, more work volume is correlated with more success in any given company. This is certainly not true from a generating volume or sorry, generating value perspective, but overall, if you are considered someone who has delivered a lot of things, then that goes back to that kind of hard worker heuristic, same kind of thing, right? And so when you're doing your performance review, pay attention. How much am I actually doing? How much work am I putting out? The volume of getting things done is going to be a hugely like overpowering heuristic. So if you have an opportunity to do work where you can show the volume of work that you've done, it's more likely to help your career. Again, this, these are, uh, these are illusions, right? It's not true that you've generated more value for the company necessarily than somebody who put out less total things, right? But it is true that most people will think that way. And so, uh, you know, in order to shape your decision-making day to day, having these useful illusions in front of you, listing these out and saying, you know what, and this is the important part. I'm not asking you to believe these things, right? Uh, instead we're exploiting this system. We're exploiting the system in order to have more efficient action in our day to day. So we're not wasting our time arguing with ourselves, arguing with our manager, arguing with our peers over whether this is actually the right way or whether it's actually the, you know, the, the most optimal path. It actually, it doesn't matter most of the time. Uh, we're choosing to exploit this heuristic in order to drive more behavior that we care about more behavior that takes us wherever we want to go. Right. Uh, this is going to help people, uh, who are listening to the show. If you adopt this, it'd be more decisive, ultimately, um, more decisive, more action oriented, more able to kind of set aside arguments that have very little utility. Right. So many times the, the mindset of a good, a very good engineers is, uh, the apologist, right. Um, kind of trying to, trying to get people to understand where they've gone wrong. Um, but the truth is that most of the time, this isn't the, the best way to impact decision-making and it's not the best way to impact your career. I hope you enjoyed this episode of developer T. Um, if you enjoy this, if you have listened to this show for very long at all, you know, that this show has been, uh, going for over a decade. Every episode is free. Um, and we've just recently started our YouTube, uh, more in earnest. So you can find this episode on YouTube, go and check it out. Uh, this developer T on YouTube, of course, you can subscribe and whatever podcasting app you're currently using. Thank you so much for listening to the show and for, for watching us on YouTube and engaging with this community. And I hope that these concepts will help you in your career. Thank you again to today's sponsor unblocked. You can get started with three free weeks. I get unblocks.com that's G E T unblocked.com slash developer T get unblocked.com slash developer T. Thank you again to unblocked for sponsoring today's episode. Thanks so much for listening. And until next time, enjoy your tea. Thank you.