Software, German, and the parts of life that fit in neither.

Quick thought

When someone on your team brings you a problem, resist solving it right away. Ask what they would do first. Most of the time they already have a good answer and just need permission to trust it.

Quick thought

AI removed the bottleneck of doing. What is left is the bottleneck of choosing. When output is cheap and endless, time management stops being the skill. Prioritisation is the skill. The question is no longer 'how do I fit this in' but 'should this exist at all'.

Quick thought

Most of your problems can really be resolved if you just give up, stop resisting, and see the truth.

Quick thought

The more I work with AI tools, the more I value colleagues who say I don't know. A model will always give you an answer. A person who admits they are unsure saves you from trusting a confident guess.

Quick thought

Treat code from an AI assistant like a pull request from a confident new hire. It usually looks right and reads well, which is exactly why you need to check it. The bugs hide in the parts you were too relaxed to read.

Quick thought

Every yes to something small is a quiet no to something you care about. Before you agree, ask what it will take time from. If you cannot name it, you are probably about to give away your evening.

Quick thought

The biggest problem in communication is the illusion that it has taken place.

Quick thought

If you are stuck for more than half an hour, ask someone. Nobody remembers that you asked for help. People do remember the week you lost trying to look like you did not need it.

Quick thought

When a customer asks for a feature, ask what they were trying to do when they needed it. The request is their guess at a solution. The problem behind it is what you should build for, and it is often simpler.

Quick thought

If you cannot explain a task to an AI in a few clear sentences, you probably do not understand it yet. Writing the prompt is often where the real thinking happens. Sometimes I solve the problem before I even press enter.

Quick thought

Before you accept a meeting, ask what decision it should end with. If nobody can say, it is probably a conversation that fits in a short message. The best meetings I have been in started with a clear question and ended early.

Quick thought

The quickest way to trust yourself again is to keep small promises. Say you will read ten pages tonight, then read them. Each promise you keep makes the next one easier, and slowly the big goals stop feeling out of reach.

Quick thought

Your first price is almost always too low. The people who leave over a small increase are rarely the ones keeping your business alive. Raise it once, watch who stays, and you will learn more about your customers than any survey could tell you.

Quick thought

When you send a long message, put your question in the first line. People skim, and the ask hiding in the last paragraph is the one that never gets an answer.

Quick thought

Before you build anything, talk to five people who would pay for it. Not friends, not family. If none of them asks when they can have it, the idea needs more work, not more code.

Quick thought

Friendships rarely end with a fight. They fade because nobody sends the next message. If someone crosses your mind today, tell them. It takes one minute and it might make their week.

Quick thought

Before you close your laptop, write down the very next step for tomorrow. Starting is the hardest part of the day, and a clear first step means you spend the morning working instead of remembering.

Quick thought

A good apology has no but in it. The moment you add one, the other person stops hearing the apology and starts hearing your defense. Say sorry, stop there, and explain later if they ask.

Quick thought

Some of my best work was code I deleted. Every line you keep is a line someone has to read, test and understand later. Before adding something new, check if there is something you can remove instead.

Quick thought

When you thank someone at work, name the exact thing they did. A general thank you is forgotten by lunch. Telling someone their test plan caught a bug before release stays with them for months, and shows them what to keep doing.

Quick thought

When you feel behind, compare yourself to who you were a year ago, not to the people around you. You see their results but not the years it took to get there. Your own progress is the only comparison where you know the whole story.

Quick thought

Frameworks come and go every few years. The things that stayed useful for me were the boring ones: SQL, HTTP, and reading other people's code. When you pick what to learn next, spend most of your time on what will still matter in ten years.

Quick thought

Keep a short note of your big decisions and why you made them. Read it again six months later. You will see which choices came from good judgment and which ones just got lucky, and that is hard to learn any other way.

Quick thought

When a bug takes more than an hour, stop and write down what you know for sure. Half of what you believed turns out to be a guess, and the bug usually lives in one of those guesses.

Quick thought

Most of my best ideas came on a walk, not at my desk. When you are stuck, pushing harder rarely helps. Step away for twenty minutes and let your mind work in the background. Rest is part of the work, not a break from it.

Quick thought

Ten customers who love what you make are worth more than a thousand who barely care. People who love it tell their friends for free. People who barely care have to be won back again and again, usually with money.

Quick thought

Remember the small things people tell you, like the exam they were nervous about or their kid's first day at school. Asking about it a week later tells them they were really heard. It costs you nothing, and people rarely forget it.

Quick thought

When you learn something new, try explaining it to a colleague the same week. The gaps in your understanding show up fast, and fixing them over coffee is much cheaper than finding them in production.

Quick thought

When a friend tells you about a hard day, ask one question before giving advice. Do you want ideas, or do you just want me to listen? Most of the time they want to be heard, and the advice can wait.

Quick thought

When someone asks how long a task will take, give a range instead of a single number. One number sounds like a promise. A range shows what it really is, a guess, and it gives you room to share early when things change.

Quick thought

Being senior is not about how much code you write. It is about how much faster the team moves because you are there. Some weeks that means writing less and unblocking more.

Quick thought

Customers rarely leave over a missing feature. They leave when they feel ignored. A quick, honest reply to a complaint often keeps someone longer than your next release will.