18 Matching Annotations
  1. Nov 2021
  2. Sep 2021
  3. d3c33hcgiwev3.cloudfront.net d3c33hcgiwev3.cloudfront.net
    1. if a model assumes that peopledo not optimize and then that model purports to produces some predictable pattern, thenthat pattern can often be exploited by someone who does optimize. Or at a minimum, bysomeone who as read the model. For this reason, many economists believe that any modelthat does not assume rationality lacks coherence

      Like overpowered character in games

    2. So, while one good test of a model is not only whether itcan generate the thing that it set out to explain, another is whether it can explain or predictother related phenomena

      It is obviously usefull for smmilar cases, but mroe interesting for cases in other fields, when we could use our model not for specific case in field, but moer based on keyword combination

    1. You need a different checklist and different mental models for different companies. I can never make it easy by saying, ‘Here are three things.’ You have to derive it yourself to ingrain it in your head for the rest of your life.”

      Use process to do confusing work

    2. Well, the first rule is that you’ve got to have multiple models ‑ because if you just have one or two that you’re using, the nature of human psychology is such that you’ll torture reality so that it fits your models, or at least you’ll think it does.”

      We frame everything in known to us models

    3. Well, the first rule is that you can’t really know anything if you just remember isolated facts and try and bang ‘em back. If the facts don’t hang together on a latticework of theory, you don’t have them in a usable form.

      Efficiency of facts = to efficiency of systems it is getting used

    4. If you get into the mental habit of relating what you’re reading to the basic structure of the underlying ideas being demonstrated, you gradually accumulate some wisdom.

      Means: extract abstract concepts from text

    5. The trick is to have your brain work better than the other person’s brain because it understands the most fundamental models: ones that will do most work per unit

      Fundamental models collect more stuff inside it. And they most basic elements for synthesis and combinations

    1. And whatever vehicle people pick it tells you where their fears lie. Did you pick your business? Money is your fear. Did you pick your wife? Losing your life is scary for you.

      It's our tools, we need to indentify what is gonna rust it

    2. Connection and Love Tony Robbins says that everyone needs connection and love. Love is a bit scary though so most people settle for connection which you can get with intimacy, walk in nature.. Again, you can meet this need in positive ways, negative ways or empowering/disempowering ones

      Will be connected in life, where able to show our inside, it's allow us to grow stronger and develop ourselves as we can access Feedback from other people, and these people are accepting us as we are, so we can become truly our self, here is no need to be in constant pressure to hide some information or lie to someone, also it's providing opportunity for deepest relaxation

    3. Certainty Anthony Robbins says you can get certainty in ways which are positive, negative, disempowering

      Good thing she creates stress but small amount of stress is positive he's motivating us putting out changes

    4. Tony Robbins says there are 3 levels at which people evaluate things: the targets (driving human forces: top 2 needs of 6 human needs), the guiding force (global/identity beliefs, so you know how they meet their needs); fuel of choice (the top emotions people experience).

      First it's what you required for survival, second is identificztion with some group by checking beliefs, and last evolution of current situation using emotions

    5. Anthony Robbins says Change often happens when we least expect it. Example, a tragedy strikes, and then we quickly change. And it happens more easily when we’re in a trance

      Condition similar to hypnosis

  4. Aug 2021
    1. This is because the output of the strided layer still does represent the same image. It is not so much cropping as it is resizing, only thing is that each single pixel in the output is a “representative” of a larger area (of whose other pixels were discarded) from the same rough location from the original input. So when the next layer’s kernel operates on the output, it’s operating on pixels collected from a larger area

      So we make pixels more influential on their territory, instead of sharing big cities together

    2. rom the start to the end of the network, while the number of channels grow deeper

      So in fact when making vision more far, moving from one level abstraction to another.

    3. because the kernel itself is applied across the entire image, the features the kernel learns must be general enough to come from any part of the image

      So kernels masters of all