Pankaj Kumar

strong beliefs win strong men, and then make them stronger.

[Pankaj Kumar @pankajflows]

Single point of failure -
If a component fails, I can attach a watcher to it. But,
Who watches the watcher?
Answer is redundant watchers. And then an alert to human if all watchers go down

[Pankaj Kumar @pankajflows]

A frontend client (browser, mobile app) calls the DNS service and gets the IP address associated with the web address. It caches the IP and reuses it until it end of life (TTL).
Imagine if we can have plain web addresses like pankajrocks. No jargon. Can we make our own DNS?

[Pankaj Kumar @pankajflows]

@neetcode 150 done. But deep down I know, I will have to revisit it after sometime

[Pankaj Kumar @pankajflows]

An interesting application of so boring pigeon-hole principle.
- Hash table collisions
- Cellular network allocation
- Data compression limits

How beautiful it that you can prove complex things using such simple analogies?

[Pankaj Kumar @pankajflows]

Started implementing Lift management system or Elevator as classes and objects. Got confused defining entities and could not figure a way out. Instead of taking help from LLM, I think I should learn it.

[Pankaj Kumar @pankajflows]

@polsia has no customization option. If we use MCP server of other services, like cold emailing by @InstantlyAI , email by @TitanEmail , landing pages in @framer etc and control everything from a central dashboard which has full context it might really make sense.

[Pankaj Kumar @pankajflows]

Turns out that @neetcode1 150 already was covering #blind75 and its done

[Pankaj Kumar @pankajflows]

144/150 @neetcode1 150 challenge. About to reach there.

[Pankaj Kumar @pankajflows]

@system_monarch Counting number of encrypted texts received is different from actually reading it

[Pankaj Kumar @pankajflows]

@NewsFromGoogle @Android Why would I update my stock holdings here after doing a transaction at brokers dashboard already?

[Pankaj Kumar @pankajflows]

I actually misplaced the terms. I was referring to SL-M trade which gets one out the market immediately upon a triggered price.
I wanted to place a SL-Limit order with well defined range like [Limit price, Stop loss]

[Pankaj Kumar @pankajflows]

If you trade in stock market, the price at which you buy/sell a stock as limit doesn't guarantee execution at that price. One has to specifically mention stop loss. Thus, a order will execute in range [Limit price, Stop loss]

[Pankaj Kumar @pankajflows]

The law of large numbers.
As the sample size grows, the average results will get closer and closer to real results.
Eg: LLMs became more accurate when a lot of data from internet was fed. But that did cost a lot of money

[Pankaj Kumar @pankajflows]

In real world, A share price of stock usually fluctuates within its bollinger band (another standard for moving average based on normal distribution). But it keeps pushing up/ pulling down the mean by a little. That trend is observed in coarse time windows like 1hr, 1day etc

[Pankaj Kumar @pankajflows]

Extreme results, both positive and negative, should make you more conscious not more confident.

[Pankaj Kumar @pankajflows]

Then, how should a person improve? Well, raise the bar of your mean 😎

[Pankaj Kumar @pankajflows]

Same thing happens when you start getting less marks or even worse fail. Within a few tests your marks will jump up to your mean performance.

[Pankaj Kumar @pankajflows]

This the foundation of mean regression. An easy concept but easily overlooked.
For eg:- When you start getting good marks in school, you assume you have become a great scholar. But within a few days your marks will fall to your mean performance.

[Pankaj Kumar @pankajflows]

Now a confusing case. IT companies are firing because AI has entered workspace. That's a narrative which seems true because it does some of the work faster. But, mass hiring during covid when interest rate was low seems like better probable causes. There can be many others.

[Pankaj Kumar @pankajflows]

We, as humans, have tendency to assume correlation is causation.
Correlation is when two things happen together. (eg: I am feeling good because it rained)
Causation is when a thing which actually causes the other. (eg: food prices went high due to scanty rainfall)