v0.1
e.g. Need to have some idea about Double Entry Bookkeeping before talking about building a bookkeeping system.
e.g. Biz vs Tech. (Except for cutting-edge tech driven biz). Need to launch something quickly vs build it properly. e.g. AliCloud.
Also refer to Yitang PM.
Often time we can't properly define a problem without introducing a certain level of implication in the form of solutions. Think of this as finger pointing to the moon.
easy peasy lemon squeezy
still somewhat ok until you need to teach someone else
Learn what? To what extend? Bloom's tax. 1/10/100 principles. (iterative)
It's okay that I just learn enough to solve the problem at hand (Ties with the recursive approach)
Pitfalls: not MECE. Missed a critical component. Missed a better option.
Verify that I've changed from I know I don't know to I know I know
Most dangerous
e.g. Recommender system
Give context and background, check the math example in Chapter 2
No learn is going to be wasted. So don't be afraid of learning
- improve learning skill
- enhance understanding of other matters
- need to learn anyways
- give you more options
Why? What are the possible causes? How to verify or rule out? Don't jump to fix too quickly.
More in ch 5
Why? Can we not do it? (Lean) Can we solve another simpler problem?
Usually one or two levels deeper can be enough.
short term vs long term
lo-fi, hi-fi, prototype, etc
ROI (1/10)
2/5/3 principles
not just think. sometimes needs to take action