Random Thought Thread

I adore when something weighs like half a ton, a ton, or even more, costs a fortune, and yet is extremely easy to break.

Sometimes you can get a good one at auction, it's not crazy expensive, but it's still a five digit number and if you drop it even one inch, it's not any good anymore... 😞


There are special considerations for moving one too. This is one reason I like to go pick stuff like this up myself. You can't trust a rigger or hot shotter on stuff like this.
 
I have definitely picked up bigger more expensive things

Resized_20210617_152749_392106877334276.jpeg

That one's 10 metric tons

(I actually had to rent a forklift for that one. Our biggest lift is only a 16k)

But the only machine I've ever dropped was a grinder

Resized_20211027_174017.jpeg

Screenshot_20211027-203213_Messages.jpg

Screenshot_20211027-203316_Messages.jpg


When you're bumping up against the limits of your tools and capabilities and experience, dangerous and costly mistakes can happen.

Good times

I have learned how to find almost anything funny.
 
Sometimes you can get a good one at auction, it's not crazy expensive, but it's still a five digit number and if you drop it even one inch, it's not any good anymore... 😞


There are special considerations for moving one too. This is one reason I like to go pick stuff like this up myself. You can't trust a rigger or hot shotter on stuff like this.
Damn, unlucky...!

I’ve got a rig with 12× RTX 5090 cards.

Sometimes I have fun with custom AI models. They can do all sorts of things — including the fun stuff.

My wife accidentally came flying in on her office chair and smashed straight into the rig (it’s sitting right next to my desk).

It’s fully water-cooled and everything. And what do you think happened?

She was totally fine (I mean my wife, obviously).

But the whole rig just shut off, then came back on — only because a contact had come loose.

Though I’ve seen an NVIDIA B200 server — that thing weighs like 150 kg. Installing one is a whole procedure!

And your grinder weighs many times more and is even more fragile.

I’ve dropped video cards, whole computers, and all kinds of stuff tons of times…

And nothing happens. Whatever.
 
I have learned how to find almost anything funny.
I also find it funny when something extremely expensive and important — the kind of thing where people gather specifically to transport, load or unload it — just suddenly drops and falls.

For some reason I just find it hilarious.

I have a weird sense of humor, but god, I laugh my ass off in situations like that.








I guess I’m an idiot. :rolleyes:
 
I think deep down I don’t really like AI and computers.

It’s not that I don’t like them at all.

It’s just… when I look at a computer or something like that, I don’t feel the same emotion as when I see, for example, a machine tool or similar equipment.

Seriously, even just looking at some big machine, mechanism, or heavy equipment gives me a genuine feeling of admiration.

I’ve seen huge servers and data centers many times and used their power, but I feel zero admiration for them.

That’s why if an extremely expensive server broke right in front of me… I’d just shrug it off. Whatever.

But if it was some big machine tool? Real physical equipment? Oh man… I’d actually feel sad.
 
But if it was some big machine tool? Real physical equipment? Oh man… I’d actually feel sad.

Hmm.

The typical American football field has 150 million to 200 million blades of grass.

Apple’s M3 Ultra SoC (3nm) has 184 billion transistors, the number of grass blades on about a thousand football fields.

Apple doesn't disclose the M3 yield; personally I'm guessing more than half are broken when coming out of the fab; Physically broken.

A 3nm semiconductor fab costs about 15 - 20 Billion Dollars. The fab has equipment pieces costing $145–$150 million USD each (EUV steppers).

All very physical to me ....
 
Last edited:
Hmm.

The typical American football field has 150 million to 200 million blades of grass.

Apple’s M3 Ultra SoC (3nm) has 184 billion transistors, the number of grass blades on about a thousand football fields.

Apple doesn't disclose the yield; personally I'm guessing more than half are broken when coming out of the fab; Physically broken.

A 3nm semiconductor fab costs about 15 - 20 Billion Dollars. The fab has equipment pieces cost $145–$150 million USD each (EUV steppers).

All very physical to me ....
This is extremely cool, seriously. Like, with all of this, in an abstract sense and in my imagination, I can kind of picture it, visualize it, or at least try to wrap my head around it.

But I literally don’t understand any of this at all, seriously.

I know how insanely complex this whole process is — compared to literally everything else, really.

And how many opportunities it opens up.

That’s exactly why, whether I like it or not, I’m learning AI hands-on, in practice.

I think all of this actually doesn’t impress me that much yet… only for now.

In any case, like it or not, I’m mastering it and using it so that in 10 years I won’t be left behind on all these major trends.

I sometimes think maybe I’m just not smart enough for all this stuff.

But it’s still really cool to watch it all happen, and overall, to see and interact with genuinely smart people.

You know, in practice, if we step away from all the numbers and analogies — which honestly might not even be real arguments for anything anyway.

To reach that actual, deep understanding and awareness. Like when you don’t just learn something, but you genuinely get it, you internalize it.

I don’t have that with AI and the entire industry — data centers, processors, GPUs, and everything else.

I don’t truly understand a huge number of basic things connected to it.

I’m basically just learning it like a dog learning commands, lol.

Personally, I don’t understand a massive amount of stuff related to all this.

I know the complexity, I know the process, and I probably understand maybe 10% at best — the technical side of 3-nanometer chips, how the actual concept of AI works.

But when it comes to these new neurons that can process information — literal, real biological neurons, damn — human neurons that they transplant into mice… I don’t get it at all.

I don’t understand this the way I understand a small number of other things.

I’m talking about that basic, fundamental level of understanding — like when you look at the blue sky and you just know it’s blue. No effort.

Something intuitive, not rational, not through analogies or numbers.

Really smart people understand this stuff intuitively.

There are many things a person just knows on an intuitive level.

Even if I spent 10 years trying to understand something, I still wouldn’t come close to someone who grasped it intuitively in just a couple of weeks.

Anyway, I'm off to sleep. I overthink way too much about the things.


P.s

By the way, I'm sure I said a lot of controversial things.

But seriously, I'm convinced that many things — especially at a deep level — simply cannot be taught. At best, you can only get somewhere very close to true understanding.

I think the greatest breakthroughs that moved progress forward usually came from people who had a kind of intuitive, almost innate sense of where they were going. They possessed a real concept and deep understanding — not just a formula written down somewhere and memorized, but an actual visualization and genuine comprehension of how that formula worked. I believe this ability is simply given to some people from the start.

No matter how hard you try, if you don't have it on an intuitive level, you'll only remember the formula or something like that, but you won't truly understand it — you won't see it the way you see colors or any physical object.
 
Last edited:
This is extremely cool, seriously. Like, with all of this, in an abstract sense and in my imagination, I can kind of picture it, visualize it, or at least try to wrap my head around it.

But I literally don’t understand any of this at all, seriously.

I know how insanely complex this whole process is — compared to literally everything else, really.

And how many opportunities it opens up.

That’s exactly why, whether I like it or not, I’m learning AI hands-on, in practice.

I think all of this actually doesn’t impress me that much yet… only for now.

In any case, like it or not, I’m mastering it and using it so that in 10 years I won’t be left behind on all these major trends.

I sometimes think maybe I’m just not smart enough for all this stuff.

But it’s still really cool to watch it all happen, and overall, to see and interact with genuinely smart people.

You know, in practice, if we step away from all the numbers and analogies — which honestly might not even be real arguments for anything anyway.

To reach that actual, deep understanding and awareness. Like when you don’t just learn something, but you genuinely get it, you internalize it.

I don’t have that with AI and the entire industry — data centers, processors, GPUs, and everything else.

I don’t truly understand a huge number of basic things connected to it.

I’m basically just learning it like a dog learning commands, lol.

Personally, I don’t understand a massive amount of stuff related to all this.

I know the complexity, I know the process, and I probably understand maybe 10% at best — the technical side of 3-nanometer chips, how the actual concept of AI works.

But when it comes to these new neurons that can process information — literal, real biological neurons, damn — human neurons that they transplant into mice… I don’t get it at all.

I don’t understand this the way I understand a small number of other things.

I’m talking about that basic, fundamental level of understanding — like when you look at the blue sky and you just know it’s blue. No effort.

Something intuitive, not rational, not through analogies or numbers.

Really smart people understand this stuff intuitively.

There are many things a person just knows on an intuitive level.

Even if I spent 10 years trying to understand something, I still wouldn’t come close to someone who grasped it intuitively in just a couple of weeks.

Anyway, I'm off to sleep. I overthink way too much about the things.


P.s

By the way, I'm sure I said a lot of controversial things.

But seriously, I'm convinced that many things — especially at a deep level — simply cannot be taught. At best, you can only get somewhere very close to true understanding.

I think the greatest breakthroughs that moved progress forward usually came from people who had a kind of intuitive, almost innate sense of where they were going. They possessed a real concept and deep understanding — not just a formula written down somewhere and memorized, but an actual visualization and genuine comprehension of how that formula worked. I believe this ability is simply given to some people from the start.

No matter how hard you try, if you don't have it on an intuitive level, you'll only remember the formula or something like that, but you won't truly understand it — you won't see it the way you see colors or any physical object.
What just happened


I read the first and second paragraph while enjoying a beer....looked down seen there was more..... then i think I had a stroke lol
 
This is extremely cool, seriously. Like, with all of this, in an abstract sense and in my imagination, I can kind of picture it, visualize it, or at least try to wrap my head around it.

But I literally don’t understand any of this at all, seriously.

I know how insanely complex this whole process is — compared to literally everything else, really.

And how many opportunities it opens up.

That’s exactly why, whether I like it or not, I’m learning AI hands-on, in practice.

I think all of this actually doesn’t impress me that much yet… only for now.

In any case, like it or not, I’m mastering it and using it so that in 10 years I won’t be left behind on all these major trends.

I sometimes think maybe I’m just not smart enough for all this stuff.

But it’s still really cool to watch it all happen, and overall, to see and interact with genuinely smart people.

You know, in practice, if we step away from all the numbers and analogies — which honestly might not even be real arguments for anything anyway.

To reach that actual, deep understanding and awareness. Like when you don’t just learn something, but you genuinely get it, you internalize it.

I don’t have that with AI and the entire industry — data centers, processors, GPUs, and everything else.

I don’t truly understand a huge number of basic things connected to it.

I’m basically just learning it like a dog learning commands, lol.

Personally, I don’t understand a massive amount of stuff related to all this.

I know the complexity, I know the process, and I probably understand maybe 10% at best — the technical side of 3-nanometer chips, how the actual concept of AI works.

But when it comes to these new neurons that can process information — literal, real biological neurons, damn — human neurons that they transplant into mice… I don’t get it at all.

I don’t understand this the way I understand a small number of other things.

I’m talking about that basic, fundamental level of understanding — like when you look at the blue sky and you just know it’s blue. No effort.

Something intuitive, not rational, not through analogies or numbers.

Really smart people understand this stuff intuitively.

There are many things a person just knows on an intuitive level.

Even if I spent 10 years trying to understand something, I still wouldn’t come close to someone who grasped it intuitively in just a couple of weeks.

Anyway, I'm off to sleep. I overthink way too much about the things.


P.s

By the way, I'm sure I said a lot of controversial things.

But seriously, I'm convinced that many things — especially at a deep level — simply cannot be taught. At best, you can only get somewhere very close to true understanding.

I think the greatest breakthroughs that moved progress forward usually came from people who had a kind of intuitive, almost innate sense of where they were going. They possessed a real concept and deep understanding — not just a formula written down somewhere and memorized, but an actual visualization and genuine comprehension of how that formula worked. I believe this ability is simply given to some people from the start.

No matter how hard you try, if you don't have it on an intuitive level, you'll only remember the formula or something like that, but you won't truly understand it — you won't see it the way you see colors or any physical object.

This is the Random Thought Thread, but Damn…..
 
This is extremely cool, seriously. Like, with all of this, in an abstract sense and in my imagination, I can kind of picture it, visualize it, or at least try to wrap my head around it.

But I literally don’t understand any of this at all, seriously.

I know how insanely complex this whole process is — compared to literally everything else, really.

And how many opportunities it opens up.

That’s exactly why, whether I like it or not, I’m learning AI hands-on, in practice.

I think all of this actually doesn’t impress me that much yet… only for now.

In any case, like it or not, I’m mastering it and using it so that in 10 years I won’t be left behind on all these major trends.

I sometimes think maybe I’m just not smart enough for all this stuff.

But it’s still really cool to watch it all happen, and overall, to see and interact with genuinely smart people.

You know, in practice, if we step away from all the numbers and analogies — which honestly might not even be real arguments for anything anyway.

To reach that actual, deep understanding and awareness. Like when you don’t just learn something, but you genuinely get it, you internalize it.

I don’t have that with AI and the entire industry — data centers, processors, GPUs, and everything else.

I don’t truly understand a huge number of basic things connected to it.

I’m basically just learning it like a dog learning commands, lol.

Personally, I don’t understand a massive amount of stuff related to all this.

I know the complexity, I know the process, and I probably understand maybe 10% at best — the technical side of 3-nanometer chips, how the actual concept of AI works.

But when it comes to these new neurons that can process information — literal, real biological neurons, damn — human neurons that they transplant into mice… I don’t get it at all.

I don’t understand this the way I understand a small number of other things.

I’m talking about that basic, fundamental level of understanding — like when you look at the blue sky and you just know it’s blue. No effort.

Something intuitive, not rational, not through analogies or numbers.

Really smart people understand this stuff intuitively.

There are many things a person just knows on an intuitive level.

Even if I spent 10 years trying to understand something, I still wouldn’t come close to someone who grasped it intuitively in just a couple of weeks.

Anyway, I'm off to sleep. I overthink way too much about the things.


P.s

By the way, I'm sure I said a lot of controversial things.

But seriously, I'm convinced that many things — especially at a deep level — simply cannot be taught. At best, you can only get somewhere very close to true understanding.

I think the greatest breakthroughs that moved progress forward usually came from people who had a kind of intuitive, almost innate sense of where they were going. They possessed a real concept and deep understanding — not just a formula written down somewhere and memorized, but an actual visualization and genuine comprehension of how that formula worked. I believe this ability is simply given to some people from the start.

No matter how hard you try, if you don't have it on an intuitive level, you'll only remember the formula or something like that, but you won't truly understand it — you won't see it the way you see colors or any physical object.
The sky is blue due to Rayleigh scattering which scales with the fourth power of the frequency.
Also provides the color of sunsets.
 
Back
Top