Jump to content

Recommended Posts

Posted
2 minutes ago, Hank Kingsley said:

I, for one, welcome the guillotines. 

 

At this point, an increasingly larger percentage of Americans -- from various spots on the political spectrum -- feels similarly.

The hubris of the billionaire class will be their undoing.  The problem is, as if often the case in such situations, their well-being is too wrapped up in the functioning of the rest of society, so the collateral damage will be significant when it happens.

Plenty of folks have been shouting these warnings for years, including some intelligent and self-aware members of the billionaire class, but human ego, narcissism, and hubris are unstoppable forces.  They will stop only when they meet their brutal destruction, by which point it is way too late.

  • Hook 'Em 2
  • Fuck Around and Find Out 1
Posted
5 minutes ago, Brisketexan said:

At this point, an increasingly larger percentage of Americans -- from various spots on the political spectrum -- feels similarly.

The hubris of the billionaire class will be their undoing.  The problem is, as if often the case in such situations, their well-being is too wrapped up in the functioning of the rest of society, so the collateral damage will be significant when it happens.

Plenty of folks have been shouting these warnings for years, including some intelligent and self-aware members of the billionaire class, but human ego, narcissism, and hubris are unstoppable forces.  They will stop only when they meet their brutal destruction, by which point it is way too late.

4th turning. I mean this shit has all happened before already and it's not even that far off from the last time it happened. 

Teddy Roosevelt and trust busting. We are living in similar times, with similar technologies coming to change everything etc etc. 

 

https://en.wikipedia.org/wiki/Robber_baron_(industrialist)

 

  • Hook 'Em 5
Posted
21 minutes ago, immamac said:

4th turning. I mean this shit has all happened before already and it's not even that far off from the last time it happened. 

Teddy Roosevelt and trust busting. We are living in similar times, with similar technologies coming to change everything etc etc. 

 

https://en.wikipedia.org/wiki/Robber_baron_(industrialist)

 

My great hope is that all of them seem to be pouring all of their energy and wealth into AI.....so when it collapses, it will take them down with it.

  • Drool 1
Posted
Just now, Brisketexan said:

My great hope is that all of them seem to be pouring all of their energy and wealth into AI.....so when it collapses, it will take them down with it.

Mine is that we don't have to have a global scale war. 

  • Hook 'Em 1
  • Like 1
  • Rage+1 1
Posted
Is this water really not recycled over and over? Please tell me it’s not being pumped out of our aquifer 


FWIW, this generation of data centers are mostly closed loop cooling with little water usage. Up to about 2 or 3 years ago, most were evaporative cooling, which are the water hogs. The big techs switched to closed loop due to community pushback. One big example was in Uruguay, which was developing a Google dc while citizens were drinking brackish water due to a drought. But closed loop is much less efficient, resulting in more power usage.
  • Hook 'Em 2
Posted
31 minutes ago, immamac said:

Mine is that we don't have to have a global scale war. 

 

Oh, my friend.  Hope is for suckers.  Despair following the rules of (1) it only gets worse, and (2) there is no bottom, is where it's at these days.  You should join us out on the ledge.  I think there's even room for you to lie down.

 

dce632d783f4ded571f83d24472dc3f9.jpg

  • Haha 1
Posted

But here's the thing with recycling that water... you can only recycle it so many times and then you have to get rid of it...

https://ketos.co/closed-loop-cooling-water-saver-or-chemical-time-bomb

Operators bleed a small fraction of water each month to keep these levels in check, a practice known as blow-down. Even a 2% monthly bleed sounds proactive, yet contaminants can still rise much faster than in open towers that take on fresh water continuously. Higher cycles push minerals out of solution, feed microbial slime, and shorten equipment life.

The bigger worry sits outside the fence. When blow-down finally leaves the site, it can carry nitrite, glycol, and heavy metals at thousands of times the limits set for surface waters. In other words, the water saved today may return tomorrow as a very different kind of liability.

  • Hook 'Em 1
  • Rage+1 1
Posted
4 hours ago, Brisketexan said:

At this point, an increasingly larger percentage of Americans -- from various spots on the political spectrum -- feels similarly.

The hubris of the billionaire class will be their undoing.  The problem is, as if often the case in such situations, their well-being is too wrapped up in the functioning of the rest of society, so the collateral damage will be significant when it happens.

Plenty of folks have been shouting these warnings for years, including some intelligent and self-aware members of the billionaire class, but human ego, narcissism, and hubris are unstoppable forces.  They will stop only when they meet their brutal destruction, by which point it is way too late.

I read a great quote in a history book i just read "The History of History" which is a good, light read.

something like:

"the rich either agree to redistribute wealth and take a hit, and go on still being rich, or the poor cause a revolution and redistribute poverty to everyone".

  • Hook 'Em 1
Posted
1 hour ago, Parliament said:

Why not put these data centers in the north?  So when the servers get hot all you gotta do is literally open the windows?

Serious question.

Or you could always go south...

Antarctica Eyed by Microsoft and Amazon for Future Cloud Data Centers

  • Hook 'Em 1
Posted
5 hours ago, Parliament said:

Why not put these data centers in the north?  So when the servers get hot all you gotta do is literally open the windows?

Serious question.

like the northern United States? Mainly because they need to be generally close to the population they are serving. There's not exactly a bunch of internet run to the low population north, there's also not a lot of energy infrastructure.

Posted
6 hours ago, Todd Gack said:

I have a dumb question.  Why are these data centers low, flat buildings spread out over large areas instead of building up in high-rises?

It's kinda difficult to modify high-rise floorplates to accomodate subfloor water cooling loops, DC switchgear and battery rooms.  Then you need a place for an entire different chilled water plant (seperate from comfort air), along with a massive bank of diesel and/or turbine generators, multiple utility feeds and multiple service entries for private fiber and diverse telecom circuits.

So much utility infrastructure is involved that it only makes sense to build everything on a low, flat plane that starts life as slab-on-grade, reinforced tilt-wall structure.

  • Hook 'Em 1
Posted
On 12/15/2025 at 9:00 PM, TwiceHorn said:

Yeah, what I'm saying is, it's not truly creative and therefore not an extension for creators.

But yeah, your point about expectation is excellent.  AI by its very nature delivers expectation.

In patent law, we adjudge the patentable to be that which is non-obvious, which basically means beyond rote expectation.  Not very far beyond in many cases, but at least somewhat.  And most patentable inventions are not terribly creative, or even at all.

And, interestingly, some of the better inventions come from recognizing a problem, and, once the problem is recognized, the solution is pretty rote.

What Cuban misses in his post, I think, is that he and people like him are incapable of thinking like a true creative, because all he is able to think in is sellable units. How can I move more widgets, how can I sell tickets at a higher profit margin, how can I get more eyes, more clicks, more conversions? How do I get more iteration? 

And he completely fails to understand that the joy of creating, genuine creativity, comes in part through loving the process. The musician, the artist, the writer, they do the thing because they love doing the thing, and they’d do it if there was money to be made or not. The brilliance in the true creativity is born out of a love of the process. But all he can consider is “if I offload the process to AI, I can sell more creativity.”

  • Hook 'Em 5
  • Like 1
Posted
Why not put these data centers in the north?  So when the servers get hot all you gotta do is literally open the windows?

Serious question.

As someone else mentioned, fiber connections-large and redundant- are critical for siting. Although, it’s less critical for an AI learning facility than most data centers.

Also, this generation of chips are mostly liquid cooled because their energy density is so high. It’s literally impossible to blow enough air to cool them. Some older data centers did just have outside air for cooling in colder climates, but that’s not practical now.
  • Like 1
Posted
31 minutes ago, Tuco said:

Also, this generation of chips are mostly liquid cooled because their energy density is so high. It’s literally impossible to blow enough air to cool them. Some older data centers did just have outside air for cooling in colder climates, but that’s not practical now.

The density and power consumption of these things is unreal. 600 kilowatts through a single server rack is mind boggling, and then when you throw in closed loop cooling and heat pumps to move all that heat away from the chips... It's insane how much power is being burned on these things that don't actually make a profit

  • Rage+1 1
Posted
9 hours ago, heso said:

What Cuban misses in his post, I think, is that he and people like him are incapable of thinking like a true creative, because all he is able to think in is sellable units. How can I move more widgets, how can I sell tickets at a higher profit margin, how can I get more eyes, more clicks, more conversions? How do I get more iteration? 

And he completely fails to understand that the joy of creating, genuine creativity, comes in part through loving the process. The musician, the artist, the writer, they do the thing because they love doing the thing, and they’d do it if there was money to be made or not. The brilliance in the true creativity is born out of a love of the process. But all he can consider is “if I offload the process to AI, I can sell more creativity.”

Excellent post.  I meant to extend that out to a criticism of Cuban, but neglected to.

I think you're absolutely correct.

  • Hook 'Em 1
  • Like 1
Posted
10 hours ago, heso said:

What Cuban misses in his post, I think, is that he and people like him are incapable of thinking like a true creative, because all he is able to think in is sellable units. How can I move more widgets, how can I sell tickets at a higher profit margin, how can I get more eyes, more clicks, more conversions? How do I get more iteration? 

And he completely fails to understand that the joy of creating, genuine creativity, comes in part through loving the process. The musician, the artist, the writer, they do the thing because they love doing the thing, and they’d do it if there was money to be made or not. The brilliance in the true creativity is born out of a love of the process. But all he can consider is “if I offload the process to AI, I can sell more creativity.”

This is a good post. But, art has long been commoditized by capitalists.  There's a reason "starving artist" was a stereotype long before AI came around.  AI is just a leap in that exploitation.  Maybe a final gambit?

I don't think Cuban is 100% incorrect.  AI can be a great tool for the artist.  I don't mean oh, go prompt your way to "Citizen Kane."  But using AI to storyboard a concept?  To make a demo to help sell an idea?  To rapidly iterate through ideas to get where to where there is a clear vision of what you want to produce?  As a means of continuing education to hone your craft?  Those are all ways AI is going to help legitimate artists create and/or be better artists without replacing them as an artist.

I know a guy who is a singer songwriter in Nashville who is pretty enamored with AI.  He's not a starving artist, he's had some tunes cut by people with names you would recognize.  He's not using AI to come up with lyrics and chords.  He does that.  But at that point, he's using it to produce quality demos to shop to people looking for songs.  He can go from a song he wrote that he can play on guitar and sing to pseudo-professional and shoppable demo in hours.  He loves AI and doesn't fear it in the least.

I use it a lot to "hone my craft".  One of the better ways to learn music theory right now is to just chat with ChatGPT about it while sitting at a piano validating what it tells you.  I say that as someone with a mid-level understanding of music theory that is using ChatGPT to increase that understanding.  Is it as good as a Berkeley school of music education?  No, but its better than static youtube video content because I can ask questions, delve into what its saying and direct it where I want the lesson to go.  Ultimately, it helps me be better at what I love doing at a commitment level I can make.

If this is all AI was going to be for the arts, it would be a great and empowering tool.  But that's not going to be all it will be.  People like Cuban are going to use it as a way to further commoditize the arts.  Streamline efficiency and cost optimize its production so the capitalist can capitalize.  But, that's not really anything new.  AI is just the newest method.

As always, its not the thing, its the humans directing the thing that are the fucking problem.

 

  • Fuck Around and Find Out 1
Posted
23 minutes ago, Goredho said:

As always, its not the thing, its the humans directing the thing that are the fucking problem.

Goddamnit, I hate this fucking talking point. 

AI, just like guns, wouldn't exist if humans didn't create it. If civilization can't handle it, then we have to enact major restrictions or kill it altogether. 

  • Hook 'Em 1
  • Like 1
Posted
27 minutes ago, Goredho said:

I use it a lot to "hone my craft".  One of the better ways to learn music theory right now is to just chat with ChatGPT about it while sitting at a piano validating what it tells you.  I say that as someone with a mid-level understanding of music theory that is using ChatGPT to increase that understanding.  Is it as good as a Berkeley school of music education?  No, but its better than static youtube video content because I can ask questions, delve into what its saying and direct it where I want the lesson to go.  Ultimately, it helps me be better at what I love doing at a commitment level I can make.

This is a really fascinating hands-on example, since LLM's literally don't know what music sounds like. Your ear is able to easily acid test the output and advice given to you, and those models have almost certainly ingested enough music theory information to actually give good guidance. It's like having a music tutor on-demand to ask questions to. It can't listen to you, but you can get some useful information from it.

29 minutes ago, Goredho said:

If this is all AI was going to be for the arts, it would be a great and empowering tool.  But that's not going to be all it will be.  People like Cuban are going to use it as a way to further commoditize the arts.  Streamline efficiency and cost optimize its production so the capitalist can capitalize.  But, that's not really anything new.  AI is just the newest method.

As always, its not the thing, its the humans directing the thing that are the fucking problem.

My job is technical at its core, but involves alot of creativity in terms of picking tradeoffs and putting together workarounds for deficciencies. I've been able to use genAI to really help speed up some of my solutioning and troubleshooting, but it's really only something I use every once in a while. But it is extremely effective when I choose to use it as a tool.

The problem (IMO) is when it's not just another tool in your toolbox, and is instead the beginning and end of your workflow. That wrong leap is often made by management that doesn't really understand or care what their people do, and are just KPI jockeys. If your only abstraction to what is being produced is some dashboard or KPI or P/L line, then it's easy to think that AI can just do everything. 

The most successful genAI implementations i've seen so far have been when it's been used to augment an existing process or workflow to remove annoying bullshit or workaround a specific blocker. Nearly all of the failed projects have been when someone just writes a really fancy prompt and burns a billion tokens to do the same thing that a simple API could have serviced for way cheaper

  • Hook 'Em 1
Posted
32 minutes ago, Captainant said:

This is a really fascinating hands-on example, since LLM's literally don't know what music sounds like. Your ear is able to easily acid test the output and advice given to you, and those models have almost certainly ingested enough music theory information to actually give good guidance. It's like having a music tutor on-demand to ask questions to. It can't listen to you, but you can get some useful information from it.

Yeah, it's honestly great.  If I didn't already have a foundation in theory, and I needed an instructor to really direct a lesson, it might not be as usable.  Meaning, an utter beginner might not be able to call bullshit and/or find it as useful.  But with at least a working knowledge of theory and the ability to test its answers at a piano and validate with my ears, I can call bullshit and question the answer if it seems/sounds incongruent.  I think the big thing AI products are missing right now are domain specific interfaces.  You just have the empty chat entry.  It's too open-ended and abstract and people have to use their creativity in figuring out how to use it.

If instead an app had been built to give a musician-specific interface into ChatGPT being a music theory tutor at learnmusictheory.com, it would get a lot more use in this regard.  Like an onscreen keyboard to "play" with a mouse to exercise what is being discussed, maybe midi integration to check what notes you played on a synth and give feedback, maybe a Spotify integration to pull up a song at a specific point that illustrates a point of theory, maybe some "learning paths" within theory as starting points to seed the AI discussion for the complete neophyte.  Domain specificity with prescriptive workflows.  Most people won't even think of ways it can help them because they just see a chat bar.

That abstraction is also part of why it's scary.  "Here's Hal 9000, he can help you in every way, will learn from you as it does and may make you so efficient that the workforce doesn't need that many of you."  That's a lot more unsettling than, "Here's eMozart, he's going to help you learn music theory."

Posted
12 minutes ago, Goredho said:

If instead an app had been built to give a musician-specific interface into ChatGPT being a music theory tutor at learnmusictheory.com, it would get a lot more use in this regard.  Like an onscreen keyboard to "play" with a mouse to exercise what is being discussed, maybe midi integration to check what notes you played on a synth and give feedback, maybe a Spotify integration to pull up a song at a specific point that illustrates a point of theory, maybe some "learning paths" within theory as starting points to seed the AI discussion for the complete neophyte.  Domain specificity with prescriptive workflows.  Most people won't even think of ways it can help them because they just see a chat bar.

TBH DeepComposer with AWS was a pretty good implementation of that idea - it converted a .midi file into a "piano roll" which is basically a visual format of a midi file, and they would apply custom transformer models to the piano roll to change the style. Stuff like "make this melody jazzy" or "add accompany parts to this song". I think it's been since deprecated, but it was a really interesting swing at the idea. And it was before LLM's were really a thing, but still using the Transformer architecture.

The github is still up with Jupyter notebooks that explain the methodology though, both for setting up the GAN architecture to train a new producer, and for actually augmenting a song. 

https://github.com/aws-samples/aws-deepcomposer-samples/blob/master/ar-cnn/AutoRegressiveCNN.ipynb
https://github.com/aws-samples/aws-deepcomposer-samples/blob/master/gan/GAN.ipynb

I think this stuff is fascinating since it's blurring the line on input modality and communicating ideas between parties

  • Hook 'Em 1
Posted
1 hour ago, Hank Kingsley said:

Goddamnit, I hate this fucking talking point. 

AI, just like guns, wouldn't exist if humans didn't create it. If civilization can't handle it, then we have to enact major restrictions or kill it altogether. 

I hate to tell you this, but AI -- like guns -- is not going away.  The genie isn't going back in the bottle.  There may be segments of the population that eschew it for varying reasons, but they will be relatively small, independently wealthy and not dependent on corporate America for a paycheck.  Regulation and protections are a possibility, but I wouldn't expect vast effects in this regard either given current trends in leadership, governance and voting.  I have all sorts of socioeconomic concerns about AI.  I could easily see it winding up being a complete net negative to the world we live in like social media has been.  But alas, I still have to collect a paycheck to afford a roof over my head and put food on the table.

So I dance for my corporate overlords like the monkey I am. If my bosses want dispassionate, commoditized and ultimately disposable workers, you got it.  Here I am, and here are my concerns so that you know that I told you so when you come back around to wanting passionate, indispensable workers.

Posted
38 minutes ago, Goredho said:

I hate to tell you this, but AI -- like guns -- is not going away.  The genie isn't going back in the bottle.  There may be segments of the population that eschew it for varying reasons, but they will be relatively small, independently wealthy and not dependent on corporate America for a paycheck.  Regulation and protections are a possibility, but I wouldn't expect vast effects in this regard either given current trends in leadership, governance and voting.  I have all sorts of socioeconomic concerns about AI.  I could easily see it winding up being a complete net negative to the world we live in like social media has been.  But alas, I still have to collect a paycheck to afford a roof over my head and put food on the table.

So I dance for my corporate overlords like the monkey I am. If my bosses want dispassionate, commoditized and ultimately disposable workers, you got it.  Here I am, and here are my concerns so that you know that I told you so when you come back around to wanting passionate, indispensable workers.

I'm clearly talking about a different scale here. That the current iteration of "AI" is useful to you personally in your job is fine. I see the benefits of it in certain aspects.

Where the rich and powerful want to take AI hasn't even begun yet. My hope is they never get there. Bigger picture, we don't have to be compliant. It doesn't have to be inevitable. AI is useful or profitable or beneficial until it isn't. 

Compliance on a small scale like your example is one thing. But as a civilization, we still have agency. Let's try to keep it. 



×
×
  • Create New...