[dminik]

It may be defeatist, but it is correct.

Seriously, how do you even realistically approach taking on ASML. They spent decades and billions of (investment) dollars to do insane moonshot research and it paid off. But it also closed off the door behind them.

Entire countries (Russia, China, ...) have been trying to reproduce it. They have not succeeded yet.

[apsurd]

why does one need to take on ASML? I had to look it up, semiconductors.

ASML market cap is ~500B. Meta market cap is ~1.5T.

i'm no facebook fan but it was started by a dude in a dorm room.

So I think saying "well those times are gone now" is defeatist.

(fwiw i personally have no interest in building a trillion dollar company from my basement, just talking philosophy here)

[dminik]

If we're going to be like that, we should go for Nvidia. Their market cap is currently ~4T.

I can see that going very well.

[apsurd]

I'm genuinely asking why are we goaling on overtaking the most valuable companies in human history?

Startup advice, as i understand it, is about innovation: expanding the pie. I sound like a VC shill. Don't mean to be, i know it's riddled with rich people passing money around pretending like value is being created.

It's just I don't get what's so wrong with the HN crowd here trying to be better at building a successful company?

[aboardRat4]

>I'm genuinely asking why are we goaling on overtaking the most valuable companies in human history?

The most valuable companies are expected to be the largest, and, as a result, the most inefficient hence the easiest to overtake.

[aboardRat4]

>They have not succeeded yet.

Reproducing an ASML machine is a piece of cake. Okay, not a piece of cake but definitely doable. The problem is that you cannot sell your reproduction in rich countries because the US government will threaten you with sanctions and US companies will screech "patents!".

[Jensson]

> Reproducing an ASML machine is a piece of cake

No its not, you have to be extremely precise when making the machine, and only ASML knows how to do that. China already have a big government funded project to reproduce ASML machines and they have failed so far.

[IAmBroom]

Nonsense. Patents are a locked glass door; anyone can get in if they knock hard enough. And Chinese engineers wear heavy gloves and eye protection when going door-to-door.

I used to work for ASML as a design engineer, and I asked my manager why we didn't shred our paperwork, like some other companies I worked at.

"With all the trouble we've had, getting our designs to work? We should publish them to slow down our competitors!"

He wasn't wrong. “Nothing that's good works by itself, you've got to make the damn thing work.” — Thomas A. Edison.

[liotier]

> Reproducing an ASML machine is a piece of cake. Okay, not a piece of cake but definitely doable. The problem is that you cannot sell your reproduction in rich countries because the US government will threaten you with sanctions and US companies will screech "patents!".

... An argument that may not convince China and Russia, who have a track record of ignoring it - I doubt it is a significant reason why they have not achieved semiconductor manufacture tooling parity.

[aboardRat4]

>I doubt it is a significant reason why they have not achieved semiconductor manufacture tooling parity.

It's not enough to reproduce just the work of asml, you basically have to reproduce their entire supply chain, because, same reason, their suppliers will have trouble obtaining export permissions.