Are you for or against Data Centers?

[Owen Lucas]

I'm seeing a future where foreign AI actually has the unintended consequence of driving almost all financial traffic off the internet because of nonstop cyber attacks on bank customer accounts. My local independent bank was bought by a small regional chain a few years ago and they have a web portal like most banks do now. I don't even remotely believe their security measures could withstand even a basic attack from a Chinese or Russian AI hacking algorithm.

If I could find a small air-gaped bank somewhere close I would move everything except a small checking account there.

Who knows. Anthropics Mythos model is supposedly so powerful that they are only releasing it to enterprise clients specifically to try to hack and harden their own systems. This new model was able to find exploits in software no one has come across yet supposedly. SOmethign an adversary will certainly do, so it's up to us to harden before the attack.

We've been doing this manually since the internet began. AI allows you to do this at scale, pen test your systems until they break so no one else can do it as well. We are certainly doing this to the Chinese, Russians, likely all of the EU. Any and all data vacuumed up, run through a manufacturing plant sized AI with amazing analysis coming out the other end.

 

[pbm]

With the building explosion of data centers are you for or against them?

I am against them. They are noisy, use lots of water and too much power.

I am also afraid of the A.I that comes with it.

Your thoughts?

My thoughts are the same as yours.

 

[NibbanaBanana]

Every major company is using it to supplement and partially power research, development, and as a force multiplier where an employee can now be more productive.

AI results are mostly non-tangible but I can assure you, many small companies already have 6 figure + monthly bills using AI for internal operations like coding.

A website won't be able to mow your lawn either but it has existed for almost 40 years now. But AI will be powering the robots mowing your lawn eventually and today's data centers will be working on the programming, models, chip designs, and everything else that will go into that robot and anything else you will use in the future.

AI is the future, resistance is futile, AI will be everywhere, it will cure cancer and cause social unrest.

Assimilate now or be left behind. AI is the new competitive advantage.

I agree that companies are spending big $ on AI. But where's the payoff? No proven productivity gains so far. Companies are rehiring.

"AI will cure cancer and cause social unrest", "Assimilate now or be left behind". Same old AI hype from the venture boys. Cashing in the stock options. The old "pump and dump" wall street scam. Every twenty years or so there's a stock bubble. Why? That's how long it takes to raise a new generation of suckers.

 

[Trav]

look for a new SSD drive and check the prices! The reason is these data centers are buying up the memory creating a shortage, add that to the amount of electricity they consume pushing everyone's electric bill through the roof and water consumption. Stop building them and limit the use of the ones already in use.

 

[alarmguy]

I agree that companies are spending big $ on AI. But where's the payoff? No proven productivity gains so far. Companies are rehiring.

"AI will cure cancer and cause social unrest", "Assimilate now or be left behind". Same old AI hype from the venture boys. Cashing in the stock options. The old "pump and dump" wall street scam. Every twenty years or so there's a stock bubble. Why? That's how long it takes to raise a new generation of suckers.

AI is far superior to detecting cancer on imaging scans vs the human eye. Massive productivity gains however we are only at the cusp evidenced by the need for data centers.
AI is not a scam, it's already replaced 10's of thousands of workers and will continue. That doesn't mean there will not be opportunities elsewhere. (yet) HUGE break through is CT Angiograms instead of sending wires through your groin up to your heart to check for blockages and can be done with a scan.

Are we there 100% yet. Not even close, one the "cusp" we can't be there yet, because we dont have the computing power yet. But we know what an be done and there is already proof because its in use on limited scales, over time it will be greatly scaled up.

https://openmedscience.com/sharper-...nsforming-early-detection-in-medical-imaging/

https://pmc.ncbi.nlm.nih.gov/articles/PMC11795265/

https://www.cancer.gov/about-nci/or...gence-ai-versus-radiologists-assessing-tumors

Keep in mind the "car" was just invented. We are only at the cusp of what is to be.
"77,999 tech jobs were cut in the first half of 2025 due to AI adoption.
Reference: https://www.wearetenet.com/blog/ai-job-replacing-statistics"

7.5 million data entry and administrative jobs could disappear by 2027 due to AI tools.
Reference: https://www.wearetenet.com/blog/ai-job-replacing-statistics
....

 

[Trav]

Fine use it for medical but why does a web browser need AI mode? I have tried it and found it almost useless, it is wrong as much as correct. It appears to be nothing more than a glorified search engine. My fear is people will take whatever it throws out as gospel, in some ways it is a scam.

 

[Pew]

7.5 million data entry and administrative jobs could disappear by 2027 due to AI tools.
Reference: https://www.wearetenet.com/blog/ai-job-replacing-statistics
....

This is really my only concern with AI (I'm not knowledgeable enough in the construction nor operations of 'AI DCs'.) We end up losing a lot of jobs where a human could do as a start-of-career-learning-basics/part time/retired/no degree required.

We can see examples of what happens after this in other industries have seen offspring or automation (ie: factory jobs.)

But stifling progress and adapting to new technological developments has always been a balance. What do we do.... Do we continue with advancements in technology (robots, automation, AI, etc) in order to continue running manufacturing as efficient as possible with more automation and less payroll for more profit...or do we try to keep humans running those jobs so we can make sure folks can stay employed, pay their bills, and keep our middle class at the expense of efficiency and employee compensation? Heck if I know the answer, I'm no economist.

 

[NibbanaBanana]

AI is far superior to detecting cancer on imaging scans vs the human eye. Massive productivity gains however we are only at the cusp evidenced by the need for data centers.
AI is not a scam, it's already replaced 10's of thousands of workers and will continue. That doesn't mean there will not be opportunities elsewhere. (yet) HUGE break through is CT Angiograms instead of sending wires through your groin up to your heart to check for blockages and can be done with a scan.

Are we there 100% yet. Not even close, one the "cusp" we can't be there yet, because we dont have the computing power yet. But we know what an be done and there is already proof because its in use on limited scales, over time it will be greatly scaled up.

https://openmedscience.com/sharper-...nsforming-early-detection-in-medical-imaging/

https://pmc.ncbi.nlm.nih.gov/articles/PMC11795265/

https://www.cancer.gov/about-nci/or...gence-ai-versus-radiologists-assessing-tumors

Keep in mind the "car" was just invented. We are only at the cusp of what is to be.
"77,999 tech jobs were cut in the first half of 2025 due to AI adoption.
Reference: https://www.wearetenet.com/blog/ai-job-replacing-statistics"

7.5 million data entry and administrative jobs could disappear by 2027 due to AI tools.
Reference: https://www.wearetenet.com/blog/ai-job-replacing-statistics
....

Yes, I agree. It does improve upon medical imaging. I saw a report on that. But I would say that the jobs cuts we've seen are due to mass over hiring during the pandemic insanity.

 

[Owen Lucas]

"AI will cure cancer

100% it will, and already is. If you're not in the loop, then you just aren't aware of the implications.

AI is not a bubble, it is fundamentally changing the future.

Do you think there is no benefit of a manufacturing plant sized building with mega-watts of compute tackling our hardest problems and not providing any benefit?

This isn't a hype craze like the meta-verse or crypto, its actual machinery producing results. Many of us don't see these results tangibly other than public AI tools because this compute is being used for very high level operations.

and cause social unrest",

Amazon, Microsoft, Meta, and many other companies have let go thousands of developers and employees because AI can supplement or replace their work.

Some CEOs are saying that if you are hiring now, you are doing something wrong because you can't figure out a way to optimize with AI.

In 10 years the developed world will look completely different due to many jobs being eliminated,

Social unrest will definitely occur because of layoffs and lack of jobs. The financial impacts will be unavoidable. There will be a social and financial paradigm shift and a lot more of society will be on some state supported welfare. Where will the money come from? We are handing out benefits to illegals and have lots of waste and abuse, something will break.

"Assimilate now or be left behind". Same old AI hype from the venture boys.

No one had to assimilate to anything before, not the general public at least. In the world of business, yes. If your business can benefit from AI in some fashion, and you don't, your competition will eat your lunch.

 

[Blagoje]

100% against data centers but I also realize that there is little to nothing that I can do stop their trend so I choose to not lose my peace over it.

 

[Owen Lucas]

or do we try to keep humans running those jobs so we can make sure folks can stay employed, pay their bills, and keep our middle class at the expense of efficiency and employee compensation?

Business will immediately get rid of employees as soon as possible. It's the bottom line and quarterly profits at stake.

Maybe a tax can be initiated eventually to reflect lost employment capacity due to AI that will feed into the general welfare fund. Possibly an attempt to incentivise human hiring, but that's just one idea.

 

[alarmguy]

This is really my only concern with AI (I'm not knowledgeable enough in the construction nor operations of 'AI DCs'.) We end up losing a lot of jobs where a human could do as a start-of-career-learning-basics/part time/retired/no degree required.

We can see examples of what happens after this in other industries have seen offspring or automation (ie: factory jobs.)

But stifling progress and adapting to new technological developments has always been a balance. What do we do.... Do we continue with advancements in technology (robots, automation, AI, etc) in order to continue running manufacturing as efficient as possible with more automation and less payroll for more profit...or do we try to keep humans running those jobs so we can make sure folks can stay employed, pay their bills, and keep our middle class at the expense of efficiency and employee compensation? Heck if I know the answer, I'm no economist.

Well said, and I can’t disagree
However, we are nation of 350 million people out of eight or 9 billion
Whoever stays ahead in the technology world will always be the ones that can defend their freedom versus countries like China. Should they get the upper hand on us well ….

Many jobs have been lost over the decades two technology. Just think at one time, thousands of people were employed routing telephone calls in bell system operation centers.

Anyway, I don’t know the answer. Either all I know is as a nation. We need to stay at the forefront of technology because if we don’t, there’s somebody knocking at our door every day of the week looking to take that lead.

 

[PandaBear]

You never know what is inside the data center. If AI is more profitable and they have customers then other hardware gets moved out.

Fact is the “other hardware” is necessary to access the AI.

The spec for AI data center is a lot more heat dense and typically need more weight support for liquid cooling than air cooling.

Most if not all the recent data center build are for AI. Chips like DDR and NAND, even HDD now are going up so high people don't suddenly build a bunch of web servers unless they are for AI.

When things cool down probably the other way around due to pent up demand but at the moment I would assume 80%+ of all data center builds are for AI otherwise cost won't justify the build.

 

[Pew]

Business will immediately get rid of employees as soon as possible. It's the bottom line and quarterly profits at stake.

Maybe a tax can be initiated eventually to reflect lost employment capacity due to AI that will feed into the general welfare fund. Possibly an attempt to incentivise human hiring, but that's just one idea.

Yea my sister's BF works for United Healthcare and if they have a $$ goal to meet at the end of the year, they get rid of workers.

Many jobs have been lost over the decades two technology. Just think at one time, thousands of people were employed routing telephone calls in bell system operation centers.

Funny you mention that, I was thinking the exact same thing too. One of the lawyers at my side gig was telling me about those in the 50s.

 

[OVERKILL]

If AI is all super duper powerful, how come they do not direct the AI to make ultra efficient solar panels ending the worlds power "issue".

Some say "Because we have not figured it out yet".......well with the help of super computers that can make billion of decisions per minute, and access to all of the worlds data......coming up with something as simple as a 100% efficient solar panel should not be that far away.

You've used this example because you don't understand the underlying mechanics. That's not a dig, so please don't take offense, but a lot of people ask "why hasn't X been solved yet?" because they don't understand the subject.

The amount of solar irradiance at the top of the earth's atmosphere is ~1,361 watts per square meter.

The amount that hits the earth varies, but we can average it to around 1,000 watts (1kW) per square meter with no clouds, snow...etc.

The average amount the earth receives over a 24hr period is 6 kWh/m2, or an average of 250 watts per square meter.

So, to generate 1.5MW (a number being tossed around in this thread), which is 1,500,000 watts, you'd need 6,000 square meters of panels.

You can see how this doesn't scale well. Now, factor in northern latitudes with cloud cover, snow, rain...etc. And your 250W per square meter shrinks massively.

Excellent solar panels are 25% efficient. So that 1,000W of sun turns into 250W of electricity.

But, let's take a real life example here. In January, solar in my area averages less than 4% capacity factor. So, for a 24hr period, 2 MW of solar would produce at an average of 80 kW. Even if we made that 100% efficient, that's still only 320 kW for the same size array. Your 2 MW of solar producing 80 kW simply becomes 8 MW of solar producing 320 kW.

So a 200 acre solar farm that's 20 MW become an 80 MW solar farm.

So, let's say you wanted to match the output of a nuclear plant (popular comparison up here). Plant proper takes up 180 acres for 3,512MW and can produce 84,288 MWh/day.

You would need 87,800 MW of solar, which at 0.4 MW per acre, would take up 219,500 acres. 343 square miles of panels that are 100% efficient to produce the same amount of electricity.

 

[gfh77665]

I did not read through all this...but I know communist China is vehemently against the USA building data centers. Kevin O'Leary (shark tank guy) spoke about it. Communist China floods our social media trying the lather up gullible Americans to oppose any new data centers.

 

[burbguy82]

You've used this example because you don't understand the underlying mechanics. That's not a dig, so please don't take offense, but a lot of people ask "why hasn't X been solved yet?" because they don't understand the subject.

The amount of solar irradiance at the top of the earth's atmosphere is ~1,361 watts per square meter.

The amount that hits the earth varies, but we can average it to around 1,000 watts (1kW) per square meter with no clouds, snow...etc.

The average amount the earth receives over a 24hr period is 6 kWh/m2, or an average of 250 watts per square meter.

So, to generate 1.5MW (a number being tossed around in this thread), which is 1,500,000 watts, you'd need 6,000 square meters of panels.

You can see how this doesn't scale well. Now, factor in northern latitudes with cloud cover, snow, rain...etc. And your 250W per square meter shrinks massively.

Excellent solar panels are 25% efficient. So that 1,000W of sun turns into 250W of electricity.

But, let's take a real life example here. In January, solar in my area averages less than 4% capacity factor. So, for a 24hr period, 2 MW of solar would produce at an average of 80 kW. Even if we made that 100% efficient, that's still only 320 kW for the same size array. Your 2 MW of solar producing 80 kW simply becomes 8 MW of solar producing 320 kW.

So a 200 acre solar farm that's 20 MW become an 80 MW solar farm.

So, let's say you wanted to match the output of a nuclear plant (popular comparison up here). Plant proper takes up 180 acres for 3,512MW and can produce 84,288 MWh/day.

You would need 87,800 MW of solar, which at 0.4 MW per acre, would take up 219,500 acres. 343 square miles of panels that are 100% efficient to produce the same amount of electricity.

Very helpful and informative. Not taken as a dig.

I was simply using solar panels as an example of an issue that the "great" AI could solve, maybe I used the wrong example.

 

[kenny91]

The strain they put on the local power grid and water supply is a valid concern, especially with how fast they are expanding. It's definitely a tough balance between technological growth and environmental impact.

 

[KrisZ]

I did not read through all this...but I know communist China is vehemently against the USA building data centers. Kevin O'Leary (shark tank guy) spoke about it. Communist China floods our social media trying the lather up gullible Americans to oppose any new data centers.

You mean the same guy that shut down pretty much all candidates that wanted to manufacture their product in USA and would only invest if they agreed to move production to China?

The hypocrisy is thick, that’s for sure. When China floods the media with glowing reviews of their EVs, pretty much everybody wants them to sell in US cause they want that cheap car, or because they want to somehow punish the US manufacturers for being greedy. When China surpassed US in car manufacturing and sales, most on here cheered.

Now the AI scare is here and we cannot allow China to surpass us. You cannot make this stuff up lol!

 

[SC Maintenance]

You mean the same guy that shut down pretty much all candidates that wanted to manufacture their product in USA and would only invest if they agreed to move production to China?

The hypocrisy is thick, that’s for sure. When China floods the media with glowing reviews of their EVs, pretty much everybody wants them to sell in US cause they want that cheap car, or because they want to somehow punish the US manufacturers for being greedy. When China surpassed US in car manufacturing and sales, most on here cheered.

Now the AI scare is here and we cannot allow China to surpass us. You cannot make this stuff up lol!

We need a boogeyman so China is the AI boogeyman. If AI is a National Security issue, then the military can build the datacenters, and we the people will own them. We can sell tokens back to big tech at a profit in the meantime. Between water, power and infrastructure we the people are mostly paying for it already.