Ben Dover

SWE | Wannabe Gyarados

[Aina ☆ @afcaina]

can’t believe the world cup is over, what a final between england and france

[Ben Dover @BlrIntern]

A
#NewXAndroidFeedback

[Arpit Bhayani @arpit_bhayani]

If you want to fundamentally challenge an assumption and innovate, try building things from scratch.

Building something from zero takes away every assumption you would otherwise inherit from an existing solution. You naturally start focusing on, "Does this even need to be true?"

That is the real reason greenfield projects (starting from scratch) often produce radical designs. Inherited code comes with inherited constraints. A blank slate has none, so you have to justify every choice from first principles. That, makes you think.

Kafka is a good example of this. Every message queue before it - JMS, RabbitMQ, you name it - agreed on one thing without really questioning it: once a consumer reads a message, delete it. A queue is a mailbox, not an archive.

When LinkedIn built Kafka from scratch, they were not trying to build a faster queue. They asked a more fundamental question: why delete data the broker already has?

So they went in the opposite direction.

On paper, that is a worse deal for a simple use case. More disk. More moving parts for the client to manage. But it made things possible that traditional queues simply could not do: replaying old messages, running multiple consumer groups over the same stream at their own pace, and treating the log itself as the source of truth.

Apache Spark is another example. Instead of optimizing disk-based processing, it did in-memory computation. Columnar databases made a similar shift by rethinking how data should be grouped and stored.

The point is, you do not stumble into designs like these by patching an existing system. You get there by starting over and refusing to accept what has been handed to you.

Hope this helps.

[Ben Dover @BlrIntern]

@ClaudeDevs Lmao

[Ben Dover @BlrIntern]

@striver_79 Solving bigger engineering challenges do not necessarily translate to better business / growth

[Garry Kasparov @Kasparov63]

Incredible Egyptian goal is disallowed because of a foul far away, then same situation a few minutes later and goal for Argentina not disallowed! No VAR, nothing? FIFA again looks like a corrupt joke, playing favorites for stars.

[Pallavi Pandey @pallavipandeyy]

MNC and IT employees in India pay up to 30% income tax for years. Those earning ₹25 lakh a year can pay roughly ₹25–30 lakh in taxes over 5 years.

But after a layoff, they get no meaningful income support or assistance like workers do in many other countries.

The government remembers salaried taxpayers while COLLECTING taxes, but FORGETS them when they lose their jobs.

[Narundar @NarundarM]

PM Modi finished his short visit to India and returned abroad.

[sphinx @protosphinx]

Sir, they have blocked access to the very AI models you said we should share instead of build

[Kishore Nallan @kishorelive]

Indian billionaires are excellent at rent seeking, gate keeping, regulatory capture, labor arbitrage, mercantile trading and motivational speeches. Innovation etc. are out of syllabus.

[RedboxGlobal India @REDBOXINDIA]

OLA ELECTRIC : CO IN TALKS TO SUPPLY BATTERY CELLS TO GLOBAL AUTO MAKERS: PTI

[chasing_highs @chasing_highs]

The biggest achievement of Modi in last 16 years is he has successfully convinced majority of the Indians that the responsibility of the governance lies on public's shoulders and NOT on the government.

[Kruz @Kruzadar]

Lookin like toddlers shitting their diapers 🤣

[Kebbin @PunishedKebbin]

He promised 3000 rupees of YOUR money every month to bribe every woman in Bengal to vote for him and now he’s asking YOU 🫵🏽 to cut back on your lifestyle. He gets to be wreckless with your money, but you must live within a budget, make it make sense.

[Ben Dover @BlrIntern]

@WarMonitor3 More like "The Top 10 most impactful moments in European history"

[Nathaniel Whittemore @nlw]

BUT YEAH THE LAYOFFS WERE DEFINITELY ABOUT AI.

[Anish Moonka @anishmoonka]

Warren Buffett spent thirty years calling airline stocks “a death trap for investors.” Then he bought every major US airline. Then he sold them all at the bottom of the COVID crash for billions in losses.

In a 1990 letter to investors in his company Berkshire Hathaway, Buffett wrote that humanity would have been better off if “a farsighted capitalist” had been present at Kitty Hawk in 1903 and shot Orville Wright down before he could fly the first plane. He’d already been burned once. In 1989 he put $358 million into US Airways. Five years later his stake was worth a quarter of what he paid for it. He called the whole thing a case of “sloppy analysis or hubris.”

In 2016 he forgot all of it. Berkshire poured $7 to $8 billion into Delta, American, Southwest, and United. By late 2019 he was Delta’s single largest shareholder.

When COVID hit he panicked. American Airlines stock had already crashed 63%. Delta was down 59%. He sold everything at the bottom in April 2020. Within weeks the airlines started recovering. A year later American and Southwest were up 80% from those lows. United and Delta were up 70%. If he had simply held on, his position would have been worth around $5 billion more.

At the 2020 shareholder meeting he said it plainly. “Our airline position was a mistake. Berkshire is worth less today because I took that position than if I hadn’t.”

Buffett’s other most repeated line is “be greedy when others are fearful.” He sold at the bottom of a crash while everyone else was selling too. He broke his airline rule and his rule against panic-selling in the same decision. Two of the most quoted principles in modern investing, both written by him, both ignored by him when it mattered most.

The most cited investor alive could see the trap clearly enough to warn the world about it for three decades. He still walked into it twice.

[Vyom @HelloVyom]

I had a Jane Street interview that felt less like an interview and more like I accidentally wandered into someone else’s thought experiment.

It was round four. Not even final.

The interviewer walks in, looks at me, then rotates the chair 90 degrees away from the whiteboard and sits facing the wall.

No introduction.

He says, “We’re going to play a game.”

I say, “Okay.”

He says, “You are a pigeon.”

Long pause.

I wait for the rest. There is no rest.

So I say, “Alright.”

He says, “You are a pigeon in Manhattan. You have perfect information about all crumbs in a three-block radius, but you can only remember the last five crumbs you ate.”

I say, “Got it.”

He says, “What is your strategy?”

I start talking about greedy algorithms. Always go to the nearest crumb, minimize travel time, maximize caloric intake per second.

He interrupts.

“You’re optimizing locally.”

I say, “Well, yes.”

He says, “Pigeons don’t get paid to optimize locally.”

I pivot. I say maybe we model crumb arrival as a stochastic process. Then we balance exploration and exploitation. Occasionally ignore known crumbs to discover new clusters.

He spins his chair halfway toward me. Progress.

“Define occasionally.”

I say, “Maybe epsilon-greedy. Ten percent exploration.”

He nods, then immediately says, “Too high. You’ll starve.”

I say, “Five percent?”

He says, “Too low. You’ll stagnate.”

I say, “So somewhere in between.”

He says, “That’s not a number.”

I panic and say, “Seven percent.”

He writes 0.07 on the board and circles it like it means something.

Then he says, “Now introduce competition.”

He adds:

other pigeons: infinite
crumbs: finite
memory: still five
He says, “What changes?”
I say we need to account for adversarial behavior. Maybe other pigeons front-run known crumb locations.

He says, “They do.”

I say then we should randomize paths to avoid being predictable.

He says, “They randomize too.”

I say, “Then… we need better information.”

He says, “You already have perfect information.”

That one lands.

So I say the edge must come from execution, not information. Faster arrival, better pathing, maybe committing to a niche region and dominating it.

He fully turns around now.

“Define dominate.”

I say, “You consistently arrive first to crumbs in your region.”

He says, “At what cost?”

I say, “Opportunity cost of missing crumbs elsewhere.”

He nods.

Then he says, “Now assume crumbs decay in value over time.”

I say, “Like… they get eaten?”

He says, “Or stepped on.”

I say, “Okay.”

He says, “What’s your pricing model?”

I laugh a little because pigeon pricing model sounds absurd.

He does not laugh.

So I say each crumb has a time-dependent value function. The longer it sits, the less it’s worth. So we discount based on expected delay.

He says, “Write it.”

I write something hand-wavy like value equals base times e to the minus lambda t.

He stares at it for a while.

Then he says, “What’s lambda?”

I say, “Decay rate.”

He says, “No. What’s lambda.”

I realize he wants a number again.

I say, “Depends on foot traffic.”

He says, “Assume Midtown at lunch.”

I say, “High.”

He says, “That’s not a number.”

I say, “Okay… 2?”

He writes 2 next to it and underlines it twice.

Then he says, “Now you have a choice.”

He writes two options:

guaranteed crumb now worth 5
expected crumb in 10 seconds worth 12 with probability 0.6

He says, “Which do you take?”

I say expected value of the second is 7.2, so take the second.
He says, “You hesitated.”

I say, “A little.”

He says, “Why?”

I say because of risk.

He says, “You’re a pigeon.”

I say, “Right.”

He says, “Do pigeons hedge?”

I say, “No.”

He says, “Exactly.”

Then he erases the board completely.

He turns back and says, “Final question.”

I brace.

He says, “How many pigeons are there in Manhattan?”

I start doing the usual estimation. Area, density, food availability.

Halfway through he stops me.

“You don’t care about pigeons.”

I say, “I thought I was a pigeon.”

He says, “You are.”

Pause.

Then he says, “But you trade pigeons.”

That’s when I knew I was completely off.

I try to recover. I say then we should estimate supply and demand for pigeons, maybe based on tourism, food waste, urban density.

He nods slightly.

Then he asks, “Are you long or short?”

I say, “Long pigeons?”

He says, “Why?”

I say, “Because there are a lot of crumbs.”

He says, “Crumbs are decreasing.”

I say, “Then short pigeons.”

He says, “Too late.”

Silence.

He writes something down, stands up, and says, “Thanks for coming in.”

I never heard back.

all of this is cooked up