Hot Tips

(Editing notes in square brackets)

Generally Applicable

- Keep your feedback loops tight; faster iteration = better product.
- Answer the question. [Corollaries: don't extend the question, don't answer something that the question does not ask]
- Fishing for ideas in random academic papers is an eternal convex bet.
- Nobody else reads the papers; just doodling around on Arxiv and Sci-Hub can get you further that your peers.
- Always be on the lookout for ways to maximize comparative advantages in relationships.

Organizational

- Logistics rules all.
- Brooks' law and Mythical Man Month concepts generalize outside of software. [I will eventually write about this and try to nail down the key characteristics of projects/processes that are affected by this]
- What is good for a human may not be good for a company; humans need to be generalists, but companies are better off being specialists.
- Commoditize your complement.
- General purpose methods that scale > low-cost but specific methods ("the bitter lesson").
- Conway's Law: your product design will always copy the communication structures of your company.
- 90% of excelling at any laptop job is just reading comprehension.

Personal

- If you want to learn, you gotta READ! [The bar is on the floor; read a language spec over the weekend and you're the company wizard, crack a book on game strategy and you're nigh-unbeatable]
- Fix whatever little things bother you in your environment; they cause mental load, whether you like it or not.
- Practice noticing how people make you feel; take notes of the behaviors of people you want to emulate.
- Start with the feeling and reverse engineer the behavior.
- Before going for productivity nootropics or dope, see if there's excess time waste in the other parts of the day (e.g., doomscrolling on the shitter).
- Create depreciation schedules for personal equipment to better chart and plan your medium-to-long term recurring costs.
- If you wake up extremely early (0300 or so), it can feel like you have an extra day to get work done.
- If you can't get the answer the normal way, start at the answer and derive the question.
- If you're the underdog, maximize variance and always go for the win. Heuristic: go for the option that can make you lose by the most points.
- Always do the math, if you can (esp. with respect to the above).
- Gamble wrt odds, not feelings, but participate wrt feelings, not odds.
- Just ask for stuff, you might as well.
- You can teach yourself anything, if you want to.
- You can set your password to X5O!P%@AP[4\PZX54(P^)7CC)7}$EICAR-STANDARD-ANTIVIRUS-TEST-FILE!$H+H* to see if it's stored in plaintext.
- Having a sick jacket leads to sick results.
- The only way to improve is practice, but if you practice consistently you can improve rapidly.
- If you put in ≥3 hours of solid effort, you can be in the upper echelon of effectiveness for pretty much anything.
- The idea of multiple intelligences is pretty much crap. If you're smart in one field, you can be equally smart in others.
- All money is fungible; 5 mediocre meals = 2 nice meals, and so on.
- You don't need a permission slip to live your life; just do stuff.
- If you self-identify as "smart", you increase your risk aversion and make your life worse. Looking dumb is the price of greatness.
- The more crude/humorous/sweary a mnemonic, the easier it sticks. The punchier and shorter, the better.
- For productivity, nothing beats your phone notes and alarm apps; cook up a spreadsheet for extra mileage if you really need it.
- Keep your office in a state where you can leave it forever in 30 seconds.
- The easiest way to make home-cooked food taste better is to use a ton of butter.
- You can use reinforcement and operant conditioning to form good habits and kill bad ones.
- Never miss a chance to sleep, eat, or expel waste.
- Do not hold an ideology that holds you in contempt.
- There are no edge cases; edge cases are just a uncommonly occurring part of the core problem, not something special.

Home