the dumbest funniest most horrifying thing just happened to me and I'm gonna tell you guys about it in case you need a deep gutteral belly laugh

So, you’re a writer and you’re worried about writing outside your immediate experiences and identities.

Here is the thread of practical advice for you.

EXAMINE YOUR MOTIVES. Why do you want to write this?

Are you trying to solve racism/sexism/colonialism with your work?

As writers we love stories about heroic writers whose work has changed the world. And as such we like to look to our own writing to solve societal problems.

And I understand this completely, not the least because I’ve felt the pull.

But if you’re looking to play saviour with your words, it is unlikely that you will do the marginalised people you are trying to save justice.

And I understand this very often comes from place of good intentions, but there is a reason that most of the moralising plays written by white abolitionists are deeply uncomfortable to read.

It is incredibly easy for works looking to play saviour to become patronising or traffic in simplistic stereotypes that ultimately hurt the people they are looking to rescue.

Velimalai Kumaraswamy Kshetram
Kumarakovil, #Kanyakumari
#Tamilnadutemples 🚩

Velimalai Kumaraswamy Temple is a Murugan temple located in Kanyakumari District of Tamilnadu.

Muruga faces east and is said to be facing Sthanumalayan (Paramashiva) of Suchindram. Muruga is of

eight feet vigraham and Valli thayar is six feet tall. Eventhough Thiruthani is believed to be the place where Subrahmanya married Valli, it is also believed that Valli’s thayar's marriage took place with Murugan here.

There is a Kalyana Vinayakar sannidhi
who helped Muruga's

marriage with Valli. Kumarakovil is also known as Kumara Kshetram and Subrahmanya Sthala. The Tamil epic Chilapathikaram refers to the temple as Chera Nattu Aeragam

Wedding of Valli and Murugan is celebrated here in the Tamil month of Panguni (March-April).

Om Muruga
🙏🕉🙏

Pyramiding 2.0

A thread-

What is pyramiding?

It is simply adding to your winners.

Why use pyramiding?

1. It makes your winners bigger than your losers.

• Because you are adding positions when your trades are going your way, you are increasing your gains but your losers will be small comparatively.

2. You lose less during Draw-down time ( it saves you from losing big during whipsaws )

• Many times when the market is in sideways mode, Entering a trade with full risk might not be the best idea.

• So, you test the waters with a small position first, it could be ⅓ or ¼ of your total risk.

• So if the Breakout fails or whipsaws come, then you will only lose a small part of your overall risk.

• Then when position starts to go in your favor you can add to it.

3. It helps you to concentrate on a few trades,

They say that don’t put all your eggs in one basket, but don’t put them in too many either.

• See, Everyone can see charts and find breakouts, pullbacks etc.

A distillation of what @naval said on Clubhouse last night:

What you call chaos I call spontaneity.

A regimented life is like a heartbeat that's non-chaotic; it's a system that’s too ordered. It doesn't have any life to it. And real life has lots of ups and downs, some of them very extreme.

I over-plan, but planning is pretty useless. What tends to dominant life are a small number of Black Swan events in both directions, positive and negative.

Expose yourself to asymmetric upside and lots of good options: things that can become massively important for you.

And you want to avoid the asymmetric downside: anything that can end the game, whether it's through financial ruin, or reputational ruin, or physical ruin.

That requires a certain amount of chaos and spontaneity.

If nothing else, 2020 taught the power of optionality.

We think about planning as linear and controlled, but that's not how the world around us works anymore.

The world is dominated by nonlinearities, so understanding options value is far more important.

Important paper from Google on large batch optimization. They do impressively careful experiments measuring # iterations needed to achieve target validation error at various batch sizes. The main "surprise" is the lack of surprises. [thread]

https://t.co/7QIx5CFdfJ

The paper is a good example of lots of elements of good experimental design. They validate their metric by showing lots of variants give consistent results. They tune hyperparamters separately for each condition, check that optimum isn't at the endpoints, and measure sensitivity.

They have separate experiments where the hold fixed # iterations and # epochs, which (as they explain) measure very different things. They avoid confounds, such as batch norm's artificial dependence between batch size and regularization strength.

When the experiments are done carefully enough, the results are remarkably consistent between different datasets and architectures. Qualitatively, MNIST behaves just like ImageNet.

Importantly, they don't find any evidence for a "sharp/flat optima" effect whereby better optimization leads to worse final results. They have a good discussion of experimental artifacts/confounds in past papers where such effects were reported.