Sunny

← All weeks

What I’ve been reading (W10, 2024)

You can simply decide to do big things · Escalate things politely

  1. You can simply decide to do big things. When you have an idea for a product, you want to build, for a community you want to form, for anything you really and truly want to make exist in the world, when you can’t stop thinking about making it exist, the tools you’ll need start appearing. New technologies, research, and even people. The world becomes a menu.
  2. The biggest risk for a system that evolves is that it may become weighed down by experts from an expired era. As a system evolves, expertise has a limited shelf life, and those most vulnerable to this risk are often the smartest and most intelligent individuals. While they may have displayed their brilliance at one point, they may struggle to acknowledge that their success is not always repeatable — The dumber side of smart people
  3. The easiest way to make things go faster is to do less work. Use a more efficient algorithm, refactor code to eliminate unnecessary operations, and move repeated work outside loops. There are many flavours, but very often, the biggest performance boosts are gained by simply solving the same problem through fewer instructions. — On code optimisation
  4. There is no AI Strategy without a data strategy. Training AI models require data. High-quality data is crucial whether you’re training existing models, developing models from scratch, or simply testing theories. Incumbents have customer data to train models as long as they maintain privacy. — AI startups require new strategies: This time, it’s actually different
  5. Slopping hiring, hands-off leaders, no expectations, not firing and not having tough conversations is how you kill a company — Jason on Twitter.
  6. You can go to hell without moving an inch; just focus on what you lack. You can taste heaven without leaving earth, just rejoice in what you have. — James Clear
  7. As usual, Matt Levine has an interesting perspective on Elon Musk’s desire to sue OpenAI. According to Matt, Elon believes that the most effective way to improve the world is through a for-profit company that only he can manage. In other words, he wants control not charity.
  8. Escalate things politely. Provide background, make your ask clear, tee up your ask, and be open to feedback.
  9. VisionOS creates zero lag, high-precision illusions that appear real to the brain. Simple Tricks and Nonsense
  10. A painter in Delhi gets his hands back with a hand transplant surgery. The surgery took more than 12 hours; it involved connecting every artery, muscle, tendon and nerve between the donor’s hands and the recipient’s arms.
  11. Bonus for , a new version of Eloquent JavaScript is out. It’s a super good Javascript learning material.
Substack edition

View the original post

  1. You can simply decide to do big things. When you have an idea for a product you want to build, for a community you want to form, for anything you really and truly want to make exist in the world, when you can’t stop thinking about making it exist, the tools you’ll need start appearing. New technologies, research, even people. The world becomes a menu.

  2. The biggest risk for a system that evolves is that it may become weighed down by experts from a expired era. As a system evolves, expertise has a limited shelf life, and those most vulnerable to this risk are often the smartest and most intelligent individuals. While they may have displayed their brilliance at one point, they may struggle to acknowledge that their success is not always repeatable — The dumber side of smart people

  3. The easiest way to make things go faster is to do less work. Use a more efficient algorithm, refactor code to eliminate unnecessary operations, move repeated work outside of loops. There are many flavours, but very often the biggest performance boosts are gained by simply solving the same problem through fewer instructions. — On code optimisation

  4. There is no AI Strategy without a data strategy. Training AI models requires data. Whether you’re training existing models, developing models from scratch, or simply testing theories, high-quality data is crucial. Incumbents have customer data to train models as long as they maintain privacy. — AI startups require new strategies: This time it's actually different

  5. Slopping hiring, hands off leaders, no expectations, not firing and not having tough conversations is how you kill a company — Jason on Twitter

  6. You can go to hell without moving an inch, just focus on what you lack. You can taste heaven without leaving earth, just rejoice in what you have. — James clear

  7. As usual, Matt Levine has an interesting perspective on Elon Musk's desire to sue OpenAI. According to Matt, Elon believes that the most effective way to improve the world is through a for-profit company that he can manage. In other words, he wants control not charity.

  8. Escalate things politely. Provide background, make your ask clear, tee up your ask, and be open to feedback.

  9. VisionOS creates zero lag, high-precision illusions that appear real to brain. Simple Tricks and Nonsense

  10. A painter in Delhi, India gets his hands back with a hand transplant surgery. The surgery took more than 12 hours, it involved connecting every artery, muscle, tendon and nerve between the donor's hands and the recipient's arms.

Originally on Substack