35 karma · joined January 10, 2016
Claude code itself consumes lot of tokens when not needed. I have to steer it a lot while building large applications.
If they optimize their entire hardware line (iPhone, Watch, Mac Mini, Macbook) AI enhanced with local/remote LLM model, they will win big. Imagine someone running a business can manage their entire business with iPhone/Mac/iCloud without buying any other saas services (inventory, payments, customer service).
sometimes the best decision is the one you don't take.
If anyone is making money off the code they should pay annual fee which goes to contributors. Github can setup an escrow, manage licenses and distribute the money to contributors.
1. The best engineers are obsessed with solving user problems.
I think this problem is rooted in early education: students learn languages, frameworks, and tools first without understanding what problems they actually solve. Once engineers have experience building a few products for users, they begin to understand what matters to the user.
2. Being right is cheap. Getting to right together is the real work.
- Sadly most of the arguments are won by either someone in power or experience. Right decisions are made with consensus. You build consensus during creative process and leverage power and experience during crisis.
3. Bias towards action. Ship. You can edit a bad page, but you can’t edit a blank one.
- Every decision is a risk management. The smart people convert higher risk into lower risk. Most people struggle here to take the risk because of the fear of failing and just waste time arguing, debating and winning over each other.
Apple is moving into the subscription everything model slowly if you noticed. You can get new phone every year with Apple Upgrade Program, News+, Fitness+, iCloud+, Apple TV+ etc. Slowly you will be locked into the ecosystem and can't escape it. We might as well call Apple (Google/Amazon to an extent) an utility at this point.
India's has DEPA (Data Empowerment And Protection Architecture) framework that addresses the data consent problem. (e.g bank will ask your consent before sharing the data). The advantage here is it providers legal framework as well.
The solid project from Tim Berners-Lee (who invented world wide web) is an attempt to solve that. https://solidproject.org/. This is pure consumer owned but there is no legal protection from the government.
This would probably need different plane design like vertical landing and horizontal landing gears.
You just need these three things.
- A Text Editor - A Calendar - A Cloud Sync for easy access
If you need to history just backup to any cloud drives or git or home backup.
My prediction is anthropic, google or amazon will buy cursor. The next logical step to coding is building apps.
I have had couple of scars of mine. I feel like sometimes you become risk-averse and When you are launching new things you will face the fear of failure.
Having no car, or maximum 1 car per family is probably a best option.
My opinion is, it doesn't matter how/what tools student used to write the program, what matters are two things 1) do they understand the program they have written 2) have they done a good job
At the end of every day I started reflecting on it. I started looking for answers and realized once you hit your peak of stress level your mind can handle, you start to crack, show signs of strong negative emotions during meetings/discussions, you will show those emotions somewhere else (like in traffic, waiting for coffee, when your kids take time to get ready), you loose patience and your rational mind starts to break down.
I later learned from some of the self book on how to not end up like that. There is a book called "Essentialism by Greg McKeown", you probably know by the title. But honestly do yourself a favor, read it/listen and try to apply and see if it helps.
Quitting should be the last option unless you already have another plan instead do the minimal thing necessary to keep your job (just as it says in the book). Remember, it's ok to do just enough work as long as you enjoy doing that, nobody can force you to put extra effort(it's ok to say no). It's based on how much effort you can put without reaching your breaking limit. Know your limits. Use tools to increase that limit (that's a career growth .. sometimes slow growth.. sometimes fast growth)