# ThatSean > Essays and notes on AI, science, systems, business, and the connections between them. > Author: Sean > URL: https://thatsean.com/ > Generated: 2026-08-12T19:52:34.869Z > Copyright © 2026 Sean. All rights reserved. > AI agents may reference, summarize, and cite this material with attribution. Do not reproduce or substantially duplicate the material. --- # About Hi, I'm Sean. I'm an astrophysicist who moved into geospatial technology and AI. I write about AI, geospatial technology, cognitive science, startups, coding, complex systems, and philosophy. I write to think and play with ideas. The main intended audience is myself, though I hope other curious people find the work useful. Some posts express a clear opinion. Others take one concept, such as feedback loops, and examine it through examples and related ideas. Much of the writing starts in my personal Obsidian vault. The graph view maps relationships between published ideas. I am interested in many subjects and in the connections between them. The site reflects that range rather than forcing every piece into one topic. You can connect with me on [LinkedIn](https://www.linkedin.com/in/thatsean/). --- # Posts ## AI will change the type of expertise we value Date: March 29, 2026 URL: https://thatsean.com/fragments/ai-and-expertise/ Tags: AI One big change that AI will cause: a shift in what we consider expertise. Traditionally we haven’t put much value in deep understanding of a subject or how something works, we’ve only cared about making it work. AI is going to change that for the simple reason that AI can now do much of the execution for us. AI is going to move us to a world where we value deep understanding of a subject over the ability to perform the actions. This is because we will need people who have the deep conceptual understanding so that they are telling the AI to do the correct thing. It’s like how in school we’re not taught math, we’re taught calculation. We solve equations as though we are the calculator, not the mathematician who comes up with the reason it’s important to do that calculation. The simple fact was that the 6th grader was correct when they complained: “why would I ever need to know this, in real life I have a calculator on my phone”. But before calculators, that sort of rote memorization and ability to do calculations was necessary before you could learn higher math. That stopped being the case long ago, but the education system still hasn’t caught up. Likewise, a lot of white-collar-job tasks are now do-able with AI. Knowing how to perform those tasks is no longer a requirement for operating at the next level of abstraction. ## Positive and Negative Feedback Loops Date: March 26, 2026 URL: https://thatsean.com/fragments/feedback-loops/ Tags: Complex Systems > If output makes more of the same, > Then "positive feedback" 's the name. > But if something corrects > And reverses effects, > Then "negative" feedback's the game. An important concept from complex systems theory that people often misunderstand are positive and negative feedback loops. Positive feedback loops occur when a system's output serves as an input that drives the system in the same direction. This can lead to exponential growth or exponential decline. Negative feedback loops happen when an output of a system causes counteractive effects, serving as an input that pushes the system in the opposite direction. This often leads to stability or equilibrium. Abstract, I know. Let's give an example of each. ## Example 1 Imagine a spacecraft in orbit around the Earth. Despite what people often think, when orbiting the Earth there is still gravity acting on the spacecraft. The spacecraft is moving forward at a speed that allows it to continuously "fall" around the Earth without crashing into it. Also despite what people often think, a spacecraft in orbit around the Earth (especially low Earth orbit) has not fully escaped the Earth's atmosphere. The Earth's atmosphere thins out, but keeps going for thousands of miles. There are fewer particle for spacecraft to collide with up there, but anything in orbit is still colliding with particles. These collisions cause drag, which slows the spacecraft down. This is why to stay in orbit spacecraft have to fire their thrusters periodically to counteract this drag and maintain their speed. If they do not, they will slow down and eventually fall back to Earth. Okay, we have enough background info to discuss feedback loops. Say the spacecraft runs out of fuel and can no longer fire its thrusters to couteract the drag. As it slows down, it will fall closer to the Earth. That also means that it's falling farther into the Earth's atmosphere, where there are more particles to collide with. This increase of particle collisions causes more drag, which causes it to slow down even more, which causes it to fall faster. And so on, and so on. Now, here's the question: what kind of feedback loop is this? Take a minute to think about it before reading on. Got your answer? You might be surprised to hear that this is a positive feedback loop. The output (the spacecraft slowing down) serves as an input that drives the system in the same direction (slowing down even more). Many people would mistakenly think this is a negative feedback loop. The spacecraft is slowing down, it's in a decaying orbit, it's in a state of decline--there's plenty of negative imagery going on. But the nature of a feedback loop doesn't have anything to do with value judgements about the desirability of an outcome. But from a systems standpoint, what we are trying to assess is whether the feedback loop makes the effect stronger or weaker. The only thing that matters is whether the output continually makes itself stronger (positive feedback loop) or whether the output serves as an input that pushes the system in the opposite direction (negative feedback loop). The slowing down is actually driving the system in the same direction (slowing down even more), which is why it's a positive feedback loop. The more it slows down, the more it slows down. I'm trying to think of a way to turn this into a negative feedback loop. This will be a bit of a stretch, but stick with me. Say that the spacecraft is orbiting a planet that has an atmosphere that is rotating around the planet extremely fast in a perpetual storm and it only goes in one direction. This also happens to be the same direction the spacecraft is orbiting. Now instead of colliding with particles that make it slow down, the spacecraft is getting hit by particles from behind that cause it to speed up. If the spacecraft tries to slow down, it will fall farther into the atmosphere and get hit from behind by more of these fast moving particles, that push the spacecraft slightly causing it to speed up. So in this aggressively hypothetical case, the output (the spacecraft slowing down) serves as an input that drives the system in the opposite direction (speeding up again). This would be a negative feedback loop. ## Example 2 Okay, now let's do a real example of a negative feedback loop. Do you know how the wheels of a train stay on those narrow metal tracks? This is a weird one. We generally think of wheels as being...y'know...wheel shaped. But that's not how train wheels look. The wheels of trains are actually shaped like cones, with a narrow end and a wider end. The narrow end of the cone is on the inside, and the wider end is on the outside. When a train goes around a curve, the wheels on the outside of the curve have to travel farther than the wheels on the inside of the curve. If the wheels were perfectly round, they would both have to travel the same distance, which would cause problems. But because of their conical shape, when a train goes around a curve, the wheels on the outside of the curve will naturally move up onto their wider end, while the wheels on the inside of the curve will move down onto their narrower end. This allows both sets of wheels to travel different distances while still staying on the tracks. Likewise, if the train starts to drift to one side, the conical shape of the wheels will cause it to naturally correct itself and stay on the tracks. If the train drifts to the left, that action of drifting to the left moves the train onto the wider end of their left wheels and the narrower end of their right wheels. This creates a restoring force that pushes the train back towards the center of the tracks. This is a negative feedback loop. The output (the train drifting to the left) serves as an input that drives the system in the opposite direction (pushing it back towards the center of the tracks). So you can think of a negative feedback loop as a self-correcting mechanism that helps maintain stability, while a positive feedback loop is a self-reinforcing mechanism that can lead to exponential growth or decline. ## Thinking of Sales as an Explore-Exploit Trade Off Date: March 23, 2026 URL: https://thatsean.com/fragments/sales-process-explore-exploit/ Tags: Sales # Your sales process refinement should never end Sales, business development, and the broader process of GTM is a balancing act between experimentation and process. It mirrors a concept in decision science and complex systems theory called the Explore-Exploit Trade Off. The Explore-Exploit Trade Off is a fundamental dilemma in decision-making, where one must choose between exploring new possibilities (exploration) and leveraging known resources or strategies (exploitation). In the context of sales and business development, this trade-off manifests in various ways: - Testing new talking points and ways of pitching your product (exploration) versus using established pitches that have proven successful in the past (exploitation). - Positioning your product in new ways to appeal to different customer segments (exploration) versus sticking with the positioning that has historically worked (exploitation). - Experimenting with different pricing strategies (exploration) versus maintaining a consistent pricing model that has yielded results (exploitation). - Trying out new sales channels or markets (exploration) versus focusing on existing, well-performing channels (exploitation). - Changing up the stages of your sales funnel (exploration) versus sticking to a tried-and-true funnel structure (exploitation). A lot of companies tend to have a "good enough" sales process that they have settled in with. It usually works reasonably well and is easy for the sales team to learn and follow. The problem is that any existing sales process is usually based on what has worked well in the past. In business, the things that got you to your current level of success likely aren't the things that will get you to the next level. Existing sales processes are often informed by past customer and prospect interactions. Meaning it is optimized to bring in more customers like the ones you currently have. But early customers are often a poor reflection of who your ideal customer is today or could be in the future. This is why refining the sales and GTM processes needs to be a continuous process of experimentation and refinement. For a fast moving startup, I would do a serious review of the sales process every quarter at least. ## Self-Efficacy Date: March 20, 2026 URL: https://thatsean.com/fragments/self-efficacy/ Tags: psychology, motivation Self-efficacy is a term in psychology that refers to a person's confidence in thier ability to achieve some goal or perform some task. It has been identified as one of the four factors of motivation. # Mismatch between self-efficacy and ability A person's self-efficacy and their actual ability in some domain may not match. There are two options here: 1. Their perceived self-efficacy is lower than their ability: e.g. they lack confidence. 2. Their perceived self-efficacy is greater than their ability: e.g. they are over-confident or over-estimate their abilities. Interestingly, the outcomes from these two options are unbalanced in effect. In the first case, the person with lower self efficacy than ability will not attempt things because their perception of their ability (even if incorrect) will reduce their motivation. In the second case, the person will attempt more things because their perception of their ability (even if incorrect), will make them experience more motivation. If you play both of these out over time, the first case (low belief in one's own abilities) keeps the person stagnant. While the second case (false belief in one's own abilities) results in the person trying more things, repeatedly experiencing failures but learning along the way, and eventually having actually quite impressive abilities. --- # Notes ## Note - August 12, 2026 URL: https://thatsean.com/notes/linear-vs-lateral/ The traditional dynamic is: Computers - linearly accumulative Human Brains - laterally associative Now we have LLMs and knowledge graphs, which are laterally associative computers. ## Note - May 18, 2025 URL: https://thatsean.com/notes/mediocre-coders/ One of the best things AI could do for us is turn mediocre coders into good coders and turn messy coders into clean coders. And that seems to be what's happening. ## Note - April 4, 2025 URL: https://thatsean.com/notes/imposter-syndrome/ Imposter Syndrome happens because your perception of your identity lags behind your actual identity. If you are a person who is constantly changing and growing, that will always be true. The faster you change and grow, the more your perception of yourself will lag behind the reality of who you are. ## Note - March 29, 2025 URL: https://thatsean.com/notes/big-important-problems/ Big important problems are almost never easy to understand, but they are easy to trick yourself into thinking you understand ## Note - February 27, 2025 URL: https://thatsean.com/notes/reversible-irreversible-decisions/ A lot of dysfunction at companies can be explained by leaders not distinguishing between **reversible decisions** and **irreversible decisions**.